A LEARNING ZONE OF ONE'S OWN
This page intentionally left blank
A Learning Zone of One's Own Sharing Representations and Flow in Collaborative Learning Environments
Edited by Mario Tokoro Sony Computer Science Labs Inc., Tokyo, Japan and
Luc Steels Sony Computer Science Laboratory, Paris, France and VUB AI Laboratory, Brussels, Belgium
/OS
Press Amsterdam • Berlin • Oxford • Tokyo • Washington, DC
© 2004, The authors mentioned in the table of contents All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior written permission from the publisher. ISBN 1 58603 4103 Library of Congress Control Number: 2004101285
Publisher IOS Press Nieuwe Hemweg 6B 1013 BG Amsterdam The Netherlands fax:+3120 620 3419 e-mail:
[email protected]
Distributor in the UK and Ireland IOS Press/Lavis Marketing 73 Lime Walk Headington Oxford OX3 7AD England fax: +44 1865 75 0079
Distributor in the USA and Canada IOS Press, Inc. 5795-G Burke Centre Parkway Burke, VA 22015 USA fax: +1 703 323 3668 e-mail:
[email protected]
LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS
Preface This book is part of a long-term effort, initiated by Mario Tokoro, to 'weave a web' of researchers and practitioners that are dedicated to advance the quality of education in the 21st century. It involves workshops, books, large-scale experiments, international projects and web resources. The initiative intends to stimulate multidisciplinary discussions which contribute to improving education and learning for children and adults of all ages everywhere in the world, based on an improved understanding, methodology and policy for education and human learning. In our view, information and communication technology can contribute to the creation of powerful support tools, but should not be seen as an end in itself. The present book follows up on a first introductory volume: 'The Future of Learning - Issues and Prospects' (also published by IOS Press and edited by M. Tokoro and L. Steels), which has set the scene for the discussion. Both books are to a large extent based on workshops that have involved participants from the various sciences investigating learning (educational psychology, AI, brain science), technology developers, especially in computing and networking, as well as practitioners involved in education at different age levels. The material is as much as possible presented in a way that might be accessible to a broad audience. The discussions reported in this book do not claim to be the final answer to the enormously complex and extremely important question of the future of learning, far from it. Many issues remain to be explored and educational practice keeps evolving. This book covers free and open discussions, and we hope that they will be enjoyable as well as inspiring to everybody who takes an interest in education or learning either for the short or the longer term. Mario Tokoro and Luc Steels Editors
A Learning Zone of One's Own Sharing Representations and Flow in Collaborative Learning Environments Editors: Mario Tokoro and Luc Steels Copy Editor: Marleen Wynants
Acknowledgements The present book is to a large extent based on two workshops, with additional authors included. The first workshop was organised by Luc Steels in April 2002 in Calheiros, near Ponte di Lima (Portugal). It was financed by Sony Computer Science Laboratories (Tokyo and Paris) and co-funded by the Swedish Foundation for Strategic Research, thanks to Peter Gardenfors, Olle Edqvist and Maria Larsson. Workshop participants included Edith Ackermann, Walter Aprile, Olivier Coenen, Olle Edqvist, Peter Gardenfors, Celia Hoyles, Bunya Kasai, Maria Larsson, Sarat Maharaj, Yvonne Rogers, Takahiro Sasaki, John Sivell, Luc Steels, Mario Tokoro, Marleen Wynants and Jan-Christope Zoels. We thank the Count and Countess of the Pa9o de Calheiros and their staff for the warm welcome and their hospitality, and Nicole Bastien for her help in the practical organisation of the workshop. The second workshop was organised by Luc Steels in April 2003 and took place in the Casa di Carmona, in Carmona near Sevilla (Spain). It was funded by Sony Computer Science Laboratories. The participants included Antonella Delle Fave, Patrick De Muynck, Mark Eisenstadt, Ulrich Hoppe, Michinori Kawachi, Ken Mogi, Francois Pachet, Takahiro Sasaki, Daniel Schneider, Chris Sinha, Luc Steels, Mario Tokoro, Colwyn Trevarthen, and Marleen Wynants. We thank the staff of the Casa de Carmona for the great care they took in supporting the workshop activities, and Nicole Bastien for her help in the practical organisation. The present book would not have come into existence without the enormous, much appreciated efforts of each author. Special thanks go to Marleen Wynants who acted as copy editor. She convinced busy authors to work on their texts and cooperated with each of them to shape and reshape the texts and illustrations. Pictures of authors have also been contributed by her. Mario Tokoro also thanks Takahiro Sasaki and Michinori Kawachi for their help in editing this volume, and gives special thanks to Joseph Goguen for his insightful comments. Finally we thank the staff of IOS Press, particularly Einar Fredriksson, Anne Marie de Rover and other members of the IOS Press team for their efficient production of the book.
Introduction The articles in this volume have been divided into three parts. The first part, entitled "Play and Grounding", looks at play as a context likely to reveal the essence of grounding. Grounding is the embodiment of understanding things/actions in relation to and/or integrated with their environments. The second part, entitled "Optimal Experience and Emotion", shows the close association between grounding and emotion. The third part, entitled "Pedagogy and Technology", elaborates new technologies, such as the computer and internet, and concepts and pedagogical methodologies supported by such technologies. Part I. Play and Grounding In Grounding, Emotion and Learning, Mario Tokoro, a leader in the field of computer and internet technologies, who is currently much interested in learning and brain science, and Takahiro Sasaki, a researcher on computer science and learning science, focus on the significance of emotion for grounding, arguing that emotion plays an essential role in learning and human development. Grounding is the embodiment of understanding things/actions through our own real experience, and emotion facilitates embodiment and leads to the development of "self. The authors try to interpret observations in developmental psychology from the viewpoint of brain science, especially the emotion system. In Constructing Knowledge and Transforming the World, Edith K. Ackermann, a developmental psychologist, reviews Piaget, Papert and Vygotsky, noting similarities and differences among the three. While Piaget described the developmental process of cognition from the viewpoint of an individual, Vygotsky regarded social context as primary. Papert complemented Piaget by stressing the importance of communication, and noting the significance of externalizing inner feelings and ideas through media. The author considers each view important, and tries to give a unified interpretation. She then refers to the variety of representations made by individuals, one of the forms of which is "play", especially for children. Simulacre and simulation as activities are contrasted in that light. Grounding takes place in play, and it is significant that the child can reconfirm through his/her own performance what has been learned. In How Infants Learn How to Mean, Colwyn Trevathen, a child psychologist, describes precise developmental stages from newborn to infant. He cites observations showing that newborn babies have all the essentials of a whole "self. He also states that the power of a child's brain to find motivation and confidence in sensitive communication with others makes possible lifelong compassion and sympathy. By constructing narratives, mother and infant come to share history and invoke community, which gives the basis for shared meaning. He states that the principles of the productive communication are reciprocity, mutuality, attunement, regular timing, and turn-taking behavior. In Pretend Play as Learning: Case Studies from the Home, Marleen Wynants, an independent journalist and mother of two daughters, highlights the significance of children's pretend play, seeing it as simulation of self and communication. She describes
her experience, arguing the importance for individuals to fully engage in play at home, where safety is assured. She also raises some important open questions.
Part II. Optimal Experience and Emotion In A Feeling of Well-Being in Learning and Teaching, Antonella Delle Fave, a psychologist, explains the essence of the notion of "optimal experience", also called "flow". Optimal experience is a complex and highly structured state of consciousness, preferentially associated with highly structured tasks and with activities supporting autonomy and creativity. For personal development, it fosters the growth of complexity in individual behavior, and at the cultural level, it promotes survival and eventually the diffusion of information. She also stresses the importance of optimal experience not only for learning, but also for teaching. In On the Design of a Musical Flow Machine, Fra^ois Pachet, a researcher on computer science and music, with co-author Anna Rita Addessi, presents the happy results of experiments using an automatic piano-like instrument named "Continuator". It responds to what you played, in a similar style with similar duration. Both professional musicians and children were excited by it. The reflective property of the system, in which users can play with virtual copies of themselves, may have been a cause for entraining them into a "flow"-like state. Another result is that children engage in turn-taking behaviors in a spontaneous way, while playing with Continuator. In The Architecture of Flow, Luc Steels, a researcher on artificial intelligence and complex dynamic system, discusses the need to integrate notions of flow in thinking about education in general and learning environments in particular. He firstly explained the theory of "flow" devised by Csikszentmihalyi in his own words as well as referring to what Csikszentmihalyi described, focusing on how a person can be self-motivated. He then discusses the importance of the collaboration of pupils and teachers/parents in learning environments at school or at home. Lastly, from the cognitive scientist view, he proposes an operational model of agents that simulate aspects of autotelic behavior.
Part III. Pedagogy and Technology In Tools for Embodied Teaching: Celestin Freinet and the Learner-Centered Classroom, John Sivell, an educational practitioner, introduces the Freinet approach, in which real tools that have only been size-adjusted are given to children, so that they are encouraged to create. Freinet teachers select and use tools such as PC's and the web, in the light of this pedagogical philosophy. Through such genuine experiences, a learner should undergo the stage of grounding, and grow to be an individual who can handle new experiences on his own. In Playing and Learning in Digitally-Augmented Physical Worlds, Yvonne Rogers, a leader in Human-Computer Interaction and Computer-Supported Collaborative Work, and Sara Price, a researcher on information science, propose digitally augmented physical space to support reflection and interpretation in children. They illustrate this proposal with three examples. The first, "Chromarium", is a mixed reality environment for young children to explore different ways of mixing color, using various physical and digital actions, and combinations of the two. The second, "The Hunting of the Snark", is a digitally-augmented adventure game whose goal is to look for an elusive virtual creature hidden in virtual space, and appearing digitally in physical space. The last one, "Ambient Wood", is a learning experience in the form of real field trip, utilizing wireless networking with mobile and
handheld technologies. In these activities, children experience physical feeling and motion through a digital extension. In Learning Together Through Collaborative Portal Sites, Daniel K. Schneider, an educational practitioner and promoter of innovative pedagogical and technological projects, demonstrates use of the internet in education, noting that positive results are attained only for pedagogically well designed projects. The importance of the teacher's roles in learnercentered learning is also stressed; these roles are facilitator, manager and orchestrator. To support these roles effectively, the author proposes a framework and modules for building an ICT (Information and Communication Technologies) educational environment as a portal, where teachers and students can carry out their activities. In Collaborative Mind Tools, H. Ulrich Hoppe, a researcher on intelligent support in educational systems and distributed collaborative environments, first reviews various tools in the light of Computer-Supported Collaborative Learning, and examines their effectiveness. Then he introduces his own collaborative tool named "Cool Modes" (for Collaborative Open Learning, Modelling and DEsigning System). It has been shown to be a good instrument for supporting collaborative modeling activities in science and mathematics. He also finds it important for learners to use visual representations. In Playgrounds for a New Mathematics, Celia Hoyles, an educational practitioner in mathematics, points out the importance of representational infrastructure for reasoning, with two examples. One of them, the Playground project, uses Ken Kahn's ToonTalk to explore new ways for children to develop a sense of mechanism. The other example uses geometric modeling to illustrate the interaction of knowledge and representation. In ToonTalk - Powerful Building Blocks for Computer-Based Learning Environments, Ken Kahn, a specialist in computer programming, introduces an enjoyable educational software environment named ToonTalk, through which kids can program pieces of algorithms, logics, or even entire game-like programs. Logo and Smalltalk were previously proposed for educational purposes, but mastering these languages is not necessarily easy for kids. On the other hand, in ToonTalk, they can create programs easily by manipulating characters and objects in an animated world. The tool lets kids structure and externalize their concepts logically in a formal way, although it feels like playing a game. Further Vision The contributions in this volume demonstrate the great potential for the further development of the vision that inspired the meetings in which they were presented. First, we are now more confident that emotion is indispensable for human learning, and that learning requires grounding. Emotion accompanies play activities and play activities accompany emotion. This relation provides a rich context for grounding. The communication and play between an infant and its parent(s) can be interpreted as the earliest context for "emotional grounding". Play activities in which learners can become deeply involved require a safe environment. The notion of optimal experience, or flow, seems an important factor for grounding, and it is suggested that optimal experience occurs when the emotion and neuro systems come into harmony. To further examine the possible association of flow with grounding, we might analyze emotion in the light of neurophysiology; this could be a future focus for our discussions. Moreover, the reports and proposals made here on pedagogy and tools assure us of the feasibility of learning environments that are safe, effective and innovative. Technologies, if appropriately applied, should greatly benefit learning in both classroom
and home. We should remain fully aware of the potential of environments where learners and teachers are supported by easy-to-handle, pedagogically sophisticated tools of the future. We hope that the contributions in this volume will be enjoyable and inspiring to a broad audience interested in education and learning, either in the short or the longer term. Although work in the classroom context (group teaching) seems to prevail in many contributions, our attention is directed equally to settings of individual learning, and of group learning. Since this book is intended to promote open discussion of the future of learning, any comments or inquiries on the articles are most welcome.
Mario Tokoro
Contents Preface Mario Tokoro and Luc Steels
v
Acknowledgements
vi
Introduction Mario Tokoro
vii
Part I. Play and Grounding Chapter 1. Grounding, Emotion and Learning Mario Tokoro and Takahiro Sasaki
3
Chapter 2. Constructing Knowledge and Transforming the World Edith K. Ackermann
15
Chapter 3. How Infants Learn How to Mean Colwyn Trevarthen
37
Chapter 4. Pretend Play as Learning: Case Studies from the Home Marleen Wynants
71
Part II. Optimal Experience and Emotion Chapter 5. A Feeling of Well-Being in Learning and Teaching Antonella Delle Fave
97
Chapter 6. On the Design of a Musical Flow Machine Francois Packet
111
Chapter 7. The Architecture of Flow Luc Steels
13 5
Part HI. Pedagogy and Technology Chapter 8. Tools for Embodied Teaching: Celestin Freinet and the Learner-Centered Classroom John Sivell Chapter 9. Playing and Learning in Digitally-Augmented Physical Worlds Yvonne Rogers and Sara Price
153
171
Chapter 10. Learning Together Through Collaborative Portal Sites Daniel K. Schneider
193
Chapter 11. Collaborative Mind Tools H. Ulrich Hoppe
221
Chapter 12. Playgrounds for a New Mathematics Celia Hoyles
237
Chapter 13. ToonTalk - Powerful Building Blocks for Computer-Based Learning Environments KenKahn
253
List of Authors
271
Subject Index
273
Parti. Play and Grounding
This page intentionally left blank
1
"Education is very difficult but fortunately humans are very flexible. Even if they receive a bad education, they try to come up with the best in themselves. That doesn't mean we should resign to bad education! We have to provide a good atmosphere, a good environment during the time that individuals grow and we should try not to kill their motivation for doing things, for trying things out." Mario Tokoro
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
1.
Grounding, Emotion, and Learning Mario Tokoro and Takahiro Sasaki As knowledge is what should be finally constructed inside each individual, education must be carried out from the point of view of facilitation for "learnercentered learning." In this article, we review some issues of "grounding," which we believe are important for the process of learning. These are discussed particularly in connection with "emotion." Through the insights, we would like to envisage what human learning and the way of its support should be.
It is said that we human beings, or higher primates, possess instinctive motivations such as: "sensuous motivation" or "inquisitive motivation", which are desires for seeking stimuli in the environment, "cognitive motivation" which is a desire for understanding our relationship to the environment, and "motivation to act" which is a desire for being active in the world. Thus, human beings leam instinctively, which is a manifestation of our innate desire. It is a starting point for all kind of learning activities to consider how we can satisfy our desire, and how we can "enhance" this motivation, which is naturally autonomous, and move it toward "the right direction." "Grounding" is one of the important processes through which the desire for learning or for knowing is satisfied. In our previous article [8], we gave a definition of "grounding" as the process in which mere information is transformed into knowledge by having a connection with the "place" of contexts or situations; those being the ground. More precisely, grounding is the embodiment of understanding things/actions in relation to and/or integrated with their environments. In the following pages, we would like to further discuss grounding in which we consider the essence of human learning, and review it particularly from the point of view of its connection to emotions. The discussion will include a few arguments which may be a bit premature, but we present them in the hope that they invite exciting discussions about reconsidering what learning and the way of its support are and what they should be. Grounding in the Physical World Not only human beings, but dogs, cats, and almost every living thing can easily move around with a certain degree of rationality in the real world. Why is it possible? The key idea is "grounding." Since the stage of infancy, or as unborn children, we keep moving our hands and legs, and by acquiring associations with the perception of sights and sounds, we become able to control our body without thinking. Moreover, we physically acquire the basic concept of causality by operating with objects in the real world and repeat experiences of basic physical phenomena such as "an object to be falling", "a ball to be rolling", "water to be flowing", and so on. Knowledge for living in the real world cannot be
1. Grounding, Emotion and Learning
represented symbolically or logically, let alone separated from the real world, but somehow should be embedded inside the interactive loop established between the body and environment. One aspect of "grounding" is the process of embodiment which integrates the sense of "body" and "physical world", and establishes the relationship between the two by repeating physical interactions. M. Montessori discovered what may be exactly related to this process. The fact that many infants and preschoolers concentrating on tenaciously repeating simple actions such as opening and closing doors, piling up and breaking down blocks, and so on [4]. The primary knowledge structure formed during this stage will be a ground for surviving thereafter. Meanwhile, as we grow up, we come to understand things which have neither a direct connection with the physical ground nor real experience. We develop a structure of knowledge that is seemingly ungrounded from the real world and extend it further. For example, we can imagine and understand historical issues or the theory of a huge universe that we cannot directly see; we can formalize the theory of abstract mathematics; we can conduct logical thinking which is purely symbolic. These form the knowledge structure of a more abstract level which could be placed upon the primary ground mentioned above. These higher knowledge structures can also be a part of the ground if they are acquired with consistency. Thus, we regard the process of cognitive development as hierarchical grounding, which may correspond to the theory of J. Piaget [5]. According to his theory, the development of human cognition passes through the following four phases (Figure 1): 1.
Sensory-motor period (postnatal to 2 years old). Movement is directly connected with sensation, and children capture the outside world sensuously through actions, such as "pulling," "clapping," "catching," and so on. 2. Preoperational period (2 to 7 years old). The period when symbolic functions appear clearly and develop. The thinking style of children during this period is marked by "centralization," "egocentrism" and "animism." They hardly see things from another's situation or viewpoint. Many make-believe plays are observed during this period. 3. Concrete-operational period (7 to 11 years old). Various logical operations are made possible during this period, but their objects are limited to concrete targets and realityrelated matters. It is this period when they acquire the concept of conservation (even if a substance changes in its appearance, there is no change in its essential number, length, weight, and quantity.). 4. Formal-operational period (11 to 15 years old). They can perform logical operation and abstract reasoning for not only realistic and concrete events on the site but also imaginary matters and abstract events. They can understand the concepts of "combination" and "proportionality", and can conduct systematic analysis. Each knowledge structure of a certain level is founded on a lower structure, and the most primitive structure of knowledge should be attached to the physical ground. That is, all knowledge must finally, somehow, connect with the sense of the physical world.
/. Grounding, Emotion and Learning
concept
deduction image word index signal
intuitive thinking
symbolic thinking
word 4 o.
ill
I scheme birth
= real
8 mo D 9 mo
preconcep (im age)
1yrSmo D 2yis
sensory-motor period
concept = nonreal (proposition)
concept = real
word f
thinking word
1v rv
_ ll motor
logical thinking
proposition hypothesis abstraction
!,
= real
4yrs
preoperational period
7yrs D 8yrs
11 yis D 12yis
concrete - operational period
age
formal -operational period
Figure 1. The process of cognitive development.
Emotions, Feelings, and Grounding So far, we have discussed especially the importance of physical aspect of "grounding" for knowledge construction. Now we will discuss the issue of "emotions" and "feelings". We consider that "emotions" and "feelings" play important roles in the process of grounding. Both "emotions" and "feelings" are very similar words, and there is much confusion in their common use. Therefore, we should sort out their relationship in accordance with the ideas of A. R. Damasio [1]. When we encounter some danger, we are startled and run away. When we fill our stomachs, we relax in a mood of happiness. These reactions come from the following mechanisms (Figure 2). 1. First, a sensory stimulus enters the limbic system, where the amygdala and frontal cyngulate gyrus play an important role. Here the type of the stimulus is judged to be either comfortable or uncomfortable. 2. These stimuli activate several autonomic nuclei in the hypothalamus, and then the internal organs are induced in a certain type of condition which activates the internal secretion system and peptide system. Thus the chemical reaction accelerates, for example, to expand and contract the intestines, changing the heart rate, and increasing the blood flow volume which causes sweating, and so on. 3. Simultaneously, through the motor system, some expressions in the face, posture or behavior are found that are caused by skeletal muscle movements. As described thus far, in response to stimuli from external situations or from internal images, a rather structured innate reaction of body occurs, which is defined as an "emotion." The next step continues further.
8
/. Grounding, Emotion and Learning
4. The "emotional somatic condition", which arises due to the above steps, returns signals via nervous system and chemical messengers via blood vessels to the brain, and creates an image of the "somatic landscape", mainly through the network in the somatosensory region. At this point, a "feeling" is defined as an experience of "emotion." It is experienced by continuously monitoring images of the somatic landscape, i.e., the condition of the "body," represented in the brain.
Emotional Somatic States
Figure 2. How "feelings" arise from "emotions."
Although emotions and feelings have an important relationship, it does not mean that emotion is a feeling in itself. In a famous psychological experiment by S. Shachter and J. E. Singer, a particular somatic condition artificially stimulated by adrenalin produced quite different feelings among test subjects, such as a feeling of happiness or anger, according to their surrounding environment [7]. After experiencing an emotional somatic condition, the relationship between the emotion and external or internal condition which was considered to cause the emotion is then labeled as a feeling. That is to say, feelings involve higher cognitive functions than emotions. At an early stage of development, such as in the case of infants, their emotions and feelings are brought about by direct stimuli. However, by repeating the above-mentioned loop and enforcing the systematic association among feelings and characteristic conditions at those times, it becomes possible to express happiness or anger even in some more complex social contexts. Furthermore, it also becomes possible to be emotive by only imagining certain stories without having direct experiences of them. Cooperative work
1. Grounding, Emotion and Learning
9
between the prefrontal region and somatosensory region develops a complex and rich emotion and feeling system. This can be envisaged as the grounding of feeling on emotion. The proof of the relationship between the prefrontal region (especially the ventromedial prefrontal region) and somatosensory region, in connection to the functions of emotions and feelings, has been demonstrated in many cases of patients who have suffered from brain damage. Figure 3 depicts those regions. It is widely known that their feelings become extraordinarily flat, especially in expressing happiness, sadness, and anger. For patients who suffered from damage in the ventromedial prefrontal region, it seems to be difficult to appropriately label the emotional somatic conditions. For patients who suffered damage in the somatosensory region, it seems to be impossible to update the somatic landscape and/or to monitor it.
somatosensory area
ventromedial prefrontal area
Figure 3. Cortical regions which play an important role for emotions and feelings.
Regarding the relationship between emotional and cognitive structures, some patients described above are known to show normal intelligence or rationality, when they are tested within a laboratory. Nevertheless, in actual life, their reasoning and decision making severely lack appropriateness in personal and social matters, and therefore they could not make the proper decision about how to act. A. R. Damasio has argued that whether we can reason suitably in most cases in our daily life, which is made up of a series of decisions at every moment, depends on our possessing of emotions and feelings [1] (Figure 4, left). That is, even if a person possesses normal social knowledge, he/she cannot produce appropriate behaviors from that knowledge when his/her feeling system has not functioned properly. If so, why can't we imagine, vice versa, that how the emotion and feeling system functions is also important for the process of constructing and expanding his/her knowledge structure by acquiring various information and experience through grounding (Figure 4, right)? Although we do not have any scientific and concrete evidence for our conjecture at
10
/. Grounding, Emotion and Learning
this moment, it seems feasible to consider emotions and feelings also as important factors for learning.
Figure 4. Emotions and feelings as a bridge between physical world and knowledge structure.
Learner Support in Learner-Centered Learning Based on the discussion so far, it is of primary importance in learning to provide appropriate support according to the developmental stages, not only exposing abundant knowledge, but also nourishing a sound and rich physical and emotional system in the first place. It is important for infants to form a rich ground by playing enthusiastically under circumstances with an abundant variation of physical stimulation because any high-level knowledge is grounded on the primary ground which is acquired in infancy. Basically, children have a great chance of experiencing various kinds of physical stimulation and acquiring a physical ground because they are curious about everything. At the same time, they need stimulations which nourish their rich emotion and feeling system. Of course, the stimulation involves tactile contact with their parents. Learning is ultimately mediated by synaptic change in the neocortex. However, the neocortex cannot function by itself. It requires the functionality of the subcortical regions, most notably the limbic system, which mediates information processing related to emotions and feelings. In this sense, learning should be grounded on the emotion-related subcortical systems in the brain, as well as on the physical world. In other words, learning should be emotionally as well as physically grounded. This is particularly true in the case of infants and small children.
1. Grounding, Emotion and Learning
11
Figure 5. Environmental support for learning.
It is important to support forming an appropriate connection between emotions and feelings, which then makes communication with others and makes adaptation to society as well as cultural learning possible. This corresponds to the process of mind construction or personality building, so to speak, which will form one's way of thinking, way of living, motivation and self-confidence. For a sound construction of mind, it is indispensable for children to be encouraged, praised and sometimes scolded by parents (or other people) with sincerity. If the mind is constructed well, the learning motivation will naturally flow out. Upon this, if we can provide opportunities for children to come in touch with rich sources of knowledge, they could steadily expand their worlds from the viewpoint of each child's interest (Figure 5). Learning in a group is also very important and indispensable. Children learn a lot with each other by mimicking each other and/or stimulating each other. Moreover, children are always observing grownups. It is also important that we grownups show them our lifestyles as we enjoy ourselves enthusiastically concentrating on something of interest. After all, "conveying" emotion and feeling, rather than "telling" knowledge, will also open the entrance for children to expand their worlds. It is also important to bring out their intrinsic motivation. Many observations indicate that motivation brought about extrinsically, by compelling force or external rewards cannot last and that deteriorates children's activities [2, 3]. The manner of support should be natural and in learner-centered way. Nevertheless, except for a few examples such as Freinet or Montessori's school, education today is too institutional or systematic, and lacks the point of view of the individual. In addition, the manner of learning tends to be one of cramming knowledge without grounding it. If such is true, it is hard to say that the education of today sufficiently satisfies the learning desire which we naturally have. This results in a decline of the learning motivation. Following the nature of human beings, we believe in the necessity of establishing a methodology for learner-centered learning and of preparing a comprehensive environment; curricula, teaching materials and tools as they ought to be. For its achievement, we could exploit new technologies such as computers and the Internet. They make it possible for us to have a more complex and larger amount of experience in the virtual environment beyond economic or spatial limitation, and thus, they
12
I. Grounding, Emotion and Learning
will support the learning process. These technologies would be a great use for extending the knowledge and conceptions which have already been grounded. On the other hand, however, they are of no use or even harmful for knowledge and concepts that have not been sufficiently grounded. How could we truly understand the meaning of "broken", if we have never been shocked by dropping a toy and breaking it? As is often pointed out nowadays, people who thoroughly enjoy looking at violent scenes on TV or playing virtual fighting games, while not having had any experience of being hurt by someone or hurting someone, may have difficulty in forming sound sociability. We must pay close attention to the direction of using learning support and tools which appropriately meet the developmental stages of human beings. Conclusion: Hope for the Future Having touched upon how learning and the way of its support should be, we must admit that the discussion of such a nature ought to be based on what might be termed the grand vision for knowledge acquisition in general. That is, how should we perceive the process of human learning in its entirety? Or even, what should such learning mean to an individual as well as to human society as a whole in a new era? A vast amount of studies and experiments must be carried out before we are confident of an epoch-making methodology. But in the meantime many people will remain uncertain of their future. It is our sincere wish that a step forward be taken by those related in any sense to education so that a new method of learning can be proposed and shared worldwide before long. We have a hope that if such a new learning method is designed with a comprehensive understanding of human development and behavior, it should bring about learner-centered learning as we have envisaged. It would make everyone feel that he/she can find his/her way of self-enhancement, thus getting relieved from despair. The individual with hope for his/her future would live positively, which would generate positive waves to his/her neighbors and might eventually contribute to human society around the world. References [1] [2] [3] [4] [5] [6] [7] [8]
A.R.Damasio, 1994, Descartes' Error: Emotion, Reason, and the Human Brain., William Morris. E.L.Deci & R.Flaste, 1996, Why We Do and What We Do: Understanding Self-Motivation, Penguin Books. M.R.Lepper et al., 1973, Undermining Children's Interest with Extrinsic Rewards: A Test of the "Over Justification" Hypothesis, Journal of Personality and Social Psychology, 28, 129-137. P. P. Lillard, 1996, Montessori Today: A Comprehensive Approach to Education from Birth to Adulthood, Shocken Books. J. Piaget, 1936, La naissance de 1'enfant., Delachaux et Niestle. J. J. Ratey, 2001, A User's Guide to the Brain, Pantheon Books. S. Schacter & J. E. Singer, 1962, Cognitive, social and physiological determinants of emotional state, Psychological Review, 69, 379-399. M. Tokoro, 2003, The Knowledge Revolution. In: M. Tokoro and L. Steels (eds.) The Future of Learning. Issues and Prospects. IOS Press, 15-27
1. Grounding, Emotion and Learning
13
MARIO TOKORO is a Corporate Executive Vice President and Chief Technology Officer of Information Technology at Sony Corporation, and the President and the founding Director of Research of the Sony Computer Science Laboratories in Tokyo (1988) and Paris (1996). Until 1997 he was a professor of computer science at Keio University, His research interests range over a wide spectrum of Computer Science, from Computational Models, Programming Languages, Operating Systems to Artificial Intelligence. He was particularly interested in Object-Oriented Computing, Concurrent Programming, Distributed and Open Systems, and Cooperative Problem Solving. He developed object-oriented concurrent programming languages: Orient84/K is designed for describing problems which are solved by multiple knowledge agents and ConcurrentSmalltalk is based on Smalltalk-80 and being widely used for various applications and for teaching concurrent programming; Paradise is based on Lisp and has been used for describing behavioral simulation arid cooperative problem solving. His largest interest is the development of computational models/paradigms for open distributed environments. He is a leader in the field of Computer and Internet Technologies and known as a loving educator. Some books: "Concepts and Characteristics of Knowledge-based Systems," (Tokoro, M., Anzai, Y., and Yonezawa, A., eds, 1989, North Holland), "Object Oriented Concurrent Programming," (Yonezawa, A., and Tokoro, M., eds, 1987, MIT Press) TAKAHIRO SASAKI received a Ph.D. in computer science from Keio University, Japan. He is an associate researcher of Sony Computer Science Laboratories, Inc. His fundamental interests have been how the spread of personal computer, Internet, and its surrounding technologies promotes empowering individual and change the lifestyle of people. Recently he is seeking a new concept of education, as well as a framework of methodology for realising it. His research interests are Multiple Agents System (Distributed Artificial Intelligence) The Search Algorithms for Moving Target. Relationship between Learning and Evolution and Complex Adaptive Systems.
14
I. Grounding, Emotion and Learning
"We did some experiments in the use of a virtual environment with 20 kids from 6 to 12 years old. They really enjoyed the virtual environment we had built although these were kids that were not at all familiar with using computers. They quickly found out who was good at what and they spontaneously exchanged their skills. What struck us even more, and I don't know whether this is relevant or not, is that children do like to communicate very much. So after a while they tended to spend most of the time chatting to each other and would rather not go to any of the virtual pavilions to find out what was going on there..." Takahiro Sasaki
2
16
"Call it powers of pretence, seductions of sim, virtues of virtual... In all cases we have to rethink the relationship between virtual and real. Irony has it that a so called "escape" in a virtual environment can be a rich way to gain deeper understanding about the world in which we live. The difference between micro-worlds and simulation is that a micro-world is a simulation that doesn't mimic the real world but creates a new world, a displaced reality in which the dynamic properties AT play are different. What counts in the idea of simulation is an "instead of and not an "equal as"... So if you meddle around with the properties of the displaced reality, you might understand some aspects oj reality "better". We should also rethink the role of the mediator, the leader, the facilitator and allow this balance between self-determination, self-directed learning, and the ability to actually work together with others." Edith Ackermann
A Learning Zone of One's Own M. Tokoro andL. Steels (Eds.) IOS Press, 2004
17
2.
Constructing Knowledge and Transforming the World Edith K. Ackermann
The first part of this chapter examines the differences between Piaget's constructivism, what Papert refers to as "constructionism," and the socio-constructivist approach as portrayed by Vygotsky. All these views are developmental, and they share the notion that people actively contribute to the construction of their knowledge, by transforming their world. Yet the views also differ, each highlighting on some aspects of how children learn, while leaving other questions unanswered. Attempts at integrating these views [learning through experience, through media, and through others] helps shed light on how people of different ages and venues come to make sense of their experience, and find their place in the world. Tools, media, and cultural artefacts are the tangible forms, through which we make sense of our world and negotiate meaning with others. In the second part of this chapter, I speak to the articulations between make-believe activities and creative symboluse as a guiding connection to rethink the aims of representations. Simulacrum and simulation, I show, play a key role besides language in helping children ground and mediate their experience in new ways. From computer-based micro-worlds for constructive learning (Papert's turtle geometry, TERC's body-syntonic graphing), to social virtual environments (MUDing). In each case, I discuss the roles of symbolic recreation, and imaginary projection (people's abilities to build and dwell in their creations) as two powerful heuristic to keep in touch with situations, to bring what's unknown to mind's reach, and to explore risky ideas on safe grounds. I draw implications for education.
1. Constructivism, one or many? The beliefs we held about children's learning are deeply grounded in our own convictions on what it means to be knowledgeable, intelligent, experienced, and what it takes to become so. Whether implicit or explicitly stated, these convictions drive our attitudes and practices as educators, parents, teachers, and researchers.
18
2. Constructing Knowledge and Transforming the World
If we think, for example, that intelligence is innate and that talents are given, we are likely to gear our interventions at helping others unfold their existing potentials. We may do so at the cost of not giving a chance to those we think of as being "ungifted". If we believe, on the other hand, that knowledge or intelligence are a reflection of a child's surrounds, then we are more likely to "pass on" our own solutions and values. And we sometimes do so at the cost of ignoring a person's own ways of doing, of thinking, and of relating to the world. And if we believe, as constructivists do, that knowledge is actively constructed through relating to others and acting in the world, then we are tempted to step aside and just set the stage for kids to engage in hands-on explorations that fuel the constructive process. We may do so at the cost of letting them "rediscover the wheel" or drift away endlessly when shortcuts may be welcome. Obviously, there is nothing wrong in showing youngsters the right ways of doing things, in helping them unravel their natural gifts, or in creating opportunities to let them discover things by themselves. Yet, the believe in either extreme "fixity" or extreme malleability of mind can become a formula for disaster especially when worldviews are at odds, when value systems clash, or when some "unpopular views" stubbornly persist within a community. My own life-long interest in constructivism and socio-constructivism grows out of a personal belief that wherever diversity reigns, the mere transmission of traditional values just won't do. That is when people(s), young and old, need to become their own path-finders, speak their own voices, bring their own personal and collective experience to the world, and negotiate their differences with others. Constructivism, in a nutshell, states that children are the builders of their own cognitive tools, as well as of their external realities. In other words, knowledge and the world are both construed and interpreted through action, and mediated through symbol use. Each gains existence and form through the construction of the other. Knowledge, to a constructivist, is not a commodity to be transmitted—delivered at one end, encoded, retained, and re-applied at the other— but an experience to be actively built, both individually and collectively. Similarly, the world is not just sitting out there waiting to be to be uncovered, but gets progressively shaped and formed through people's interactions / transactions. Psychologists and pedagogues like Piaget, Bruner, Papert, Vygotsky, Bakt'in, but also Dewey, Freynet, Freire, Malaguzzi and many others', remind us that indeed, learning is less about acquiring information or transmitting existing ideas or values, than it is about collectively designing a world in which it is worth living. What's more, this process of negotiating views with others requires the co-construction of [taken as] "shared" forms (Reddy, 1993). In what follows, I present some aspects of Piaget's constructivist theory, and I contrast them with Papert's constructionism, and Vygotsky's socio-constructivism. I flesh out what each captures and leaves out, thus setting the stage for my own attempt at integrating the two.
1 An extended verion of this paper appears in French in Ackermann, E. (2001) Constructivism et Constructionism: Quelle difference?. Constructivismes: usages et perspectives en education. (Volumes 1 et 2). Actes du Colloque "Constructivismes". Geneva: Service de la recherche en edication / Cahier 8 / September 01. pp. 85-94.
2. Constructing Knowledge and Transforming the World
19
Eloge a 1'abstraction - Piaget, the rationalist Piaget is best known for his stages, which offer a window into what children are interested in and capable o/at different levels of their cognitive development. While this is an important contribution, there is more to Piaget than his stage. Piaget has forcefully shown that children have their own views of the world, which differ from those of adults, and that these views are extremely coherent and robust. They are stubborn, if you wish, i.e., not very easy to shake. Children, to Piaget are not incomplete adults. Instead, their ways of thinking have a reason to be, mostly well suited to their current needs and possibilities. This is not to say that children's views of the world, as well as of themselves, do not change through contact with others and with things. The views are continually evolving. Yet, to Piaget, knowledge grows according to complex laws of self-organisation, which operate in the background according to some "logic" of their own. Thus, for a child—or an adult—to abandon a current theory, or believe system, requires more than just being exposed to a better theory. Conceptual changes in children, like theory changes in scientists (Kuhn, 1970), emerge as a result of people's action-in-the-world (their living experience) in conjunction with many "hidden" regulatory processes at play behind the scene2. The function of these processes is to maintain the livelihood of the cognitive system as a whole, and to compensate for surface perturbations (regulatory mechanisms). Piaget's developmental theory emphasises how children become progressively detached from the world of concrete objects and local contingencies, and gradually able to mentally manipulate symbolic objects, within a realm of hypothetical worlds. The focus is on the construction of cognitive invariance as a means to interpret and organise the world. Piaget's empirical studies shed light on the conditions under which learners are likely to maintain or change their views of a phenomenon when interacting with it during a significant period of time. The child that Piaget portrays in his theory is an idealised child. Often referred to as an epistemic subject, s/he is a representative of the most common way of thinking at a given level of development. And this "common way of thinking" is similar to that of a scientist driven by the urge to impose stability and order over an ever-changing natural world. Piaget's child, one may say, is like a young Robinson in the conquest of an unexplored territory. Robinson's conquest is solitary yet exciting since the explorer himself is very active. Piaget's child is an inner-driven, very curious, and independent character. The ultimate goal of his adventure may not be the navigation per se, but the joy of mastering the territory under exploration. In essence, Piaget the rationalist portrays children's intellectual development as a progressive move away from intuitive towards rational thinking, from everyday cognition towards scientific reasoning. In his view, the path leading to higher forms of reasoning, or 'formal operations', proceeds from local to general, from context-bound to context-free, from externally-supported to internally-driven. Accordingly, cognitive achievements are gauged according to three major acts of distancing. 1. The ability to emerge from here-andnow contingencies (characteristic of practical intelligence); 2. the ability to extract knowledge from its substrate (i.e. from contexts of use and personal goals); and 3. the ability to act mentally on virtual worlds, carrying out operations in the head instead of carrying them out externally. For more on this cf. Piaget (1975) "1'Equilibration des structures cognitives. PUF, Paris.
20
2. Constructing Knowledge and Transforming the World
The implications of Piaget's theory for education are profound, even if Piaget himself didn't think of his work as being "educational". Let me mention three aspects that have captured my attention as a researcher and educator, or the main lessons I learned from working with Piaget: 1. Teaching can't ever be direct. Children don't just take-in what is being said. Instead, they interpret, or translate, what they hear in the light of their knowledge and experience. Willingly or unwillingly, that is, they transform the input to fit their level of understanding. This occurs whether we like it or not. A more radical formulation of lesson 1 would be to say that learning does not occur as a result of teaching or, in Piaget's own provocative terms 'whatever you tell a child, you won't allow her to discover it by herself. 2. Knowledge is not information to be delivered at one end, and encoded, stored, retrieved, and re-applied at the other end. Instead, knowledge is experience to be constructed through interactions with the world (people and things). To equate knowledge with information—and knowledge construction with information processing—confuses matters when it comes to human learning or teaching. 3. A theory of learning that ignores resistances to learning misses the point. One of Piaget's main teachings is that children have extremely good reasons not to abandon their current worldviews in the light of external surface perturbations. And this is so no matter how relevant the suggestions. A good teacher, in this sense, is one that helps learners explore, express, exchange—and ultimately expand— their views, from within [ not a sage on the stage, but a guide on the side] To conclude, while capturing what is common in children's ways of thinking at different developmental stages—and describing how this commonality evolves over time— Piaget's theory tends to overlook the role of context, uses, and media, as well as the importance of individual preferences, or styles, in human learning and development. That's where Papert's "constructionism" comes in handy! Media Matters - Papert, the intuitionist If Piaget did not see himself as an educator, Papert, on the other hand, used what Piaget learned about children as a basis for rethinking education in the digital age. He coined his theory "constructionism". In his words, "Constructionism—the N word as opposed to the V word— shares contructivism 's view of learning as "building knowledge structures "through progressive internalisation of actions... It then adds the idea that this happens especially felicitously in a context where the learner is consciously engaged in constructing a public entity, whether it's a sand castle on the beach or a theory of the universe (Papert, 1991, p.l) To Papert, projecting out—or externalising— inner feelings and ideas is as important as internalising our actions. In expressing ideas, or giving them form, we make them tangible and shareable which, in turn, shape and sharpen these ideas. Externalising ideas is also a key to communicating with others. We can only negotiate meaning through tangible forms: our own expressions or existing cultural mediations (language, tools, and toys). The cycle of self-directed learning is, to Papert, an iterative process by which learners invent for themselves the very tools and mediations that best support the exploration of intriguing
2. Constructing Knowledge and Transforming the World
21
ideas. Because of his focus on learning through making (on could say learning as design) Papert's "constructionism" sheds light on how people's ideas get formed and transformed when expressed through different media, when actualised in particular contexts, when worked out by individual minds. The emphasis has shifted from general laws of development to individuals' conversation with their own representations, artefacts, or objects-to-think with. Stressing the importance of external supports as a means to augment the unaided mind is not new. As will become clear in the next section, Vygotsky has spent his entire life studying the role of cultural artefacts—tools, language—as a resource for drawing the best out of every child's potential. So have many other researchers in the socio-constructivist tradition. The difference, as I see it, lays in: • The role such external aids are meant to play at higher levels of a person's development. • The types of external aid, or media studied (Papert focuses on digital media and computer-based technologies) and more important, • The type of initiative the learner takes in the design of her own "objects to think with". Papert's constructionism is more situated & pragmatic than Piaget's. This is so even if Papert himself doesn't make explicit use of the terms when describing his enterprise. One of its main contributions is to remind us that intelligence should be defined and studied insitu; alas, that being intelligent means being grounded, connected, and sensitive to variations in the environment. To Papert, abstract or formal thinking may well be a powerful tool. Yet, it is not necessarily the most appropriate in all situations. Unlike Piaget, Papert thinks that "diving into" situations rather than looking at them from a distance, that connectedness rather than distanciation are powerful means of gaining understanding. Becoming one with the phenomenon under study, in other words, is a key to learning. The child that Papert studies is more relational than Piaget's Robinson. S/he likes to get in tune with others and situations. S/he resembles what Sherry Turkic described as a "soft" master (Turkic, 1984). Like Piaget's Robinson, s/he enjoys discovering novelties, yet more than him, she wants to be in the flow of things, and in tune with people. S/he likes to feel at one with them.3 Like Robinson, she likes to try out things rather than being told. Unlike him, S/he is more of a conversationalist than a builder. She may prefer sharing what s/he understands while in context, rather than telling what s/he experienced in retrospect. To conclude, while Piaget best described the genesis of internal mental stability in terms of successive levels of equilibrium, Papert is interested in the dynamics of change. He stresses the fragility of thought during transitional periods. His great contribution, as an educator, is to focus our attention on how people think once their convictions break down, once alternative views sink in, once adjusting, stretching, and expanding their current view of the world becomes necessary. Papert always points toward this fragility, contextuality, and flexibility of knowledge under construction. A strong believer in the ideas that momentary losses are a key to learning, and that people are good at using what they don't know as a lever to grow, Papert has spent much of his life creating technology-enhanced environments, or micro-worlds, in which learners can mess around with otherwise risky ideas, on safe ground.
22
2. Constructing Knowledge and Transforming the World
It takes a whole village to raise a child - Vygotsky: the socio-culturalist At the heart of Vygotsky's socio-constructivism lays a simple idea. From the day they are born, people learn, thrive, and grow in relation with others. We "are" because of others. The theory in stresses the importance of caring and knowledgeable adults in helping a child thrive. Vygotsky also emphasises the role of language—and other cultural artefacts—in mediating human transactions. In spite of his focus on culture as a teaching machine, Vygotsky saw a child's intellectual development as constructive process. This is why his socio-constructivist approach cannot be put at odds with the theories of Piaget, Papert, Bruner, and others. To Vygotsky, and socio-cultural theorists, the "social" has a primacy over the "individual" in a very special sense: Society is the bearer of a cultural heritage without which the development of an individual is simply impossible. Parents and other members of a community create a developmental niche for the newcomer, which embodies the adults' cultural past and impacts the new generations' future. It is at once a habitat and a cultural medium, or mediation. It is at once a "terrain," or stage, for human experience and a lens, or interpretative frame, at the disposal of the terrain's inhabitants. Vygotsky's theory of cultural appropriation is not so different from Piaget's notion that children learn through acting in the world—i.e., through relating to people and things. This being said, Vygotsky puts greater emphasis on how the presence of adults with greater expertise can "speed up" and enhance a child's self-directed learning, and how shared cultural artefacts are used to help mediate this process. More than Piaget and Papert, Vygotsky stresses the role of adults as teachers, and cultural artefacts as teaching tools. One of the key concepts in Vygotsky's theory is the notion of zone of proximal development. Much quoted and often misunderstood, the "ZPD" has become a buzzword among many educators. The ZPD defines a potential area of expansion that each individual has at their disposal to overcome their limits, provided the social environment in which the learning takes place "pitches in". In other words, the zone of proximal development tells us "how far" we can push the envelope of what we know, when helped by others. It is, again, through social interaction, that learners can mobilise, and best use, the psychological tools available to them. To Vygotsky, a person's cognitive development proceeds outside-in, i.e., from other to self: "Every function in the child's development appears twice: first, on the social level, and later on the individual level; first, between people, and then inside the child" (Vygotsky, 1978:57 in Lock, 1989). Inter-personal relations are the precursors, and necessary conditions, for the emergence of individual/intra-mental processes: Youngsters first share their experience with others, before they become able to master and understand them, for themselves. Their development proceeds from socio-centric to egocentric.
3
Vygotsky introduced the concept of psychological tool to capture the idea that the cultural artefacts that surround us, once appropriated, become part of our own "psychology". Psychological tools include: various systems for counting; mnemonic techniques; algebraic symbol systems; works of art; writing; schemes, diagrams, maps, and technical drawings; all sorts of conventional signs, and so on. (Vygotsky, 1982:137, cited in Cole & Wertsch, 1996). Note that of the psychological tools that mediate our thoughts, feelings, and behaviours, language was the most important to Vygotsky.
2. Constructing Knowledge and Transforming the World
23
Vygotsky's child, as I see it, is more of a trusting disciple than an autonomous agent, in Harris's sense is (Harris, 2002). While curious, active, inner-driven, and autonomous, s/he also trusts that others, more experienced, can tell her things that she cannot directly experience. In other words, s/he knows that she can learn vicariously by listening to what others say about what interests her. The autonomous agent, in contrast, is not comfortable if he cannot check out for himself what others propose, at the cost - sometimes - of reinventing the wheel! Integrating the views: People as World-Makers, Dwellers in the World, and Social Creatures. In The Evolving Self, Kegan portrays human development as a lifelong attempt to resolve the unsolvable tension between getting embedded and emerging from embeddedness (Kegan, 1982). In a similar way, 1 think of cognitive or affective growth as a lifelong attempt on the part of people, young or old, to find a viable balance between fusion and separation, openness and closure, or in Piaget's own words, between assimilation and accommodation. Imposing one's order upon things [building cognitive invariants as selforienting devices] goes hand in hand with being sensitive to variations, and letting go of one's obsolete believes—should this not jeopardise previously attained balance, or equilibrium. Along with Piaget, I view separateness through progressive decentralisation as a necessary step toward relating even more intimately and sensitively to people and things. In any situation, no matter how engaging, there are moments when we need to remove ourselves and reconsider what we did, from afar. This view of separateness does not preclude the value of being embedded in one's experience. I also share Papert's view that diving into the unknown, at the cost of experiencing a momentary sense of loss, is a crucial part of learning. Only when a learner actually travels in a world, by adopting different perspectives, or putting on different "glasses," can a dialogue begin between local and initially incompatible experiences. What Vygotsky adds to this equation is the notion that "it takes a whole village to raise a child". In other words, no human can "be" or "grow' without the help of many people, peers or adults. Belonging to a caring community, and knowing how to relate to others are needed to build a sense of self. And since people relate to one another through cultural mediations—tools, language and artefacts —these, in turn, get woven into—and become an integral part of— the social transactions. To conclude, both "dwelling in" and "stepping back" are equally important in getting the cognitive dance going. Both individuation and socialisation are needed for us to grow as people. How could anyone learn from experience as long as they are totally immersed in it. There comes a time when viewing things from afar, or adopting a 'god's eyes view', is a must (Ackermann, 1996). From then on, a new cycle can begin, and the stage is set for new and deeper grounding and understanding. How could anyone get to know who they are—and what's they are worth—of they are not "held" by others. In other words, to get the cycle of self-directed learning going, learners need to exist as persons. And to exist as a person —or know who they are—to need to belong: Any child stops to speak if her words are not heard.
24
2. Constructing Knowledge and Transforming the World
2. Powers of Pretense, Seductions of SIM, Virtues of Virtual 4 "Imagine a child playing with other children, and using a stick as a horse: the child jumps around his friends, goes places, feeds the horse, claims that the horse is lazy. In creating this make-believe play, the child is making present the horse, a horse that otherwise would be absent in this child's life. Furthermore, she is not only making the horse present but doing things with it. We say that the horse is ready at hand to convey this idea that the horse is made to participate in the child's playful activities. This scene exemplifies what we call symbolising: a creation of a lived-in space in which the absent is made present and ready at hand". (Nemirovsky and Monk, 1998) The formulation by Nemirovsky and Monk frames the act of symbolising as a means to sustain a dialog between what is [believed to be] and what could be, between fact and fancy. It highlights that to represent is not merely to describe what exists but to make tangible what doesn't. The authors also remind us that beyond replicating, young pretenders often modify outcomes, and subvert the meaning of things. As in improvisational theatre, they recast unfolding events, opening up new paths as they play along. Meaning and coherence both emerge as a result of this creative process. In what follows, I challenge the prevailing theory of representation, often referred to as correspondence theory (Lakoff, 1993), suggesting that there is an a-priori object out there (a territory), that the act of representation duplicates one way or another (map). I show that representations are better thought of as performing acts, or fictionalising techniques in Iser's sense5 (Iser, 1987) The enactive/generative aspects of representations are particularly relevant in design activities where an artefact to be built doesn't exist before the process comes to an end. In design, it becomes clear that the representations needed to generate new forms couldn't possibly be conceived as descriptions of what's out there—since not much is out there yet! Designers are left with envisioning and engaging forms in the becoming. They build sketches, prototypes, and simulations as intermediary objects to generate these forms. What is true of design is also true of other constructive processes. Most striking in this respect is children's natural tendency to invent for themselves the supports and mediations they need to reach their goals, whenever the tasks they face lay beyond their mastery. Children's extraordinary talent as learners comes in great part from their ability to set the stage that allows them to safely project themselves in the unknown. Doing as if and playing what //are the techniques they use to achieve this balance. Nemirovsky and Monk's notion of "ready at hand" (above citation) further suggests that the props used in pretense play need not be [treated by the child as] arbitrary tokens, nor do they have to be at the image of what it stands for. In other words, the stick that the child "rides and feeds" in her play is a double (ersatz) in that it acts on the imaginary horse's behalf. Yet, this doesn't imply, 4
An extended German version of this paper appears in Ackermann, E. (1999) "Sich einrichten in Fantasie Raumen: Untersuchungen zum Gebrauch von Symbolen" (E. Renk Ed.) Lernen und Leben am der Welt im Kopf: Konstruktivismus in der Schule. Neuwied, Kiftel: Luchterhand, pp. 79-99. ' To Iser , the English term 'representation' causes problems because it suggests a given which the act of representation duplicates. This conceals the performative qualities through which the act of representation brings about something that hitherto doesn't exist as a given object (Iser, 1987,p.217). Iser proposes to replace the English term with the German Darsellung, which does not drag this mimetic connotation.
2. Constructing Knowledge and Transforming the World
25
again, that the double mimics its behaviour or mirrors its appearance. Symbols often take on a life of their own, and it is their ability to do so—both be and not to be what they stand for—in the pretender's mind, that enables their creative use. We know from research on early pretense play that children's abilities to treat a stick as j/it were a horse requires a de-coupling between signifier and signified (Piaget, 1962, Perner, 1993). In other words, a child who uses a stick "as if it were a horse also knows that it is not "really" a horse. What is less obvious is the notion that de-coupling has to go hand in hand with its opposite, fusion, for the symbolic transform to be complete. Along with Nemirovsky and Monk, I suggest that a child's ability to engage an "ersatz" as if it were the thing itself, i.e., to fuse signifier and signified, is a necessary condition for creative symbol-use. Fusion is what ultimately gives symbols their dramatising power. Without empathic projection—engaging the double as is—no "lived" experience would be possible. Working out intriguing materials, fictional or real, requires both the creation of make-believe ground and an occasion for "true" identification. Engaging in symbolic activities, in this sense, is not just a matter of giving form to ideas, making them tangible and shareable. It is also a matter of bringing ideas and forms to life. Treating doubles as if they were [as vivid and vibrant as] the ideas they stand for, is what brings the materials engaged in pretense closer to mind's reach. Like a mythical character, the make-believe horse-companion that the child plays with in her pretense is more like a unicorn than a real horse: a fictional creature that embodies hidden fears, desires, and purposes. And its appearance, the stick, once made to participate in the child's activities, helps reshape her original ideas about unicorns. It is, again, the ambiguous nature of the stick in the child's eyes, at once double (de-coupling), object in its own right (separation), and extension of self (fusion) that lends it its evocative and dramatising powers. To conclude, making the absent present, giving form to ideas, and bringing form and ideas to life are 3 important functions of the act of symbolising. Not just a kid's thing! Pretense or symbolic play is not just a kid's matter. Nor is it a privilege reserved to artists and poets alone. People of all ages, stages, and styles engage in symbolic recreations. And they do so in ever more sophisticated ways as they grow older (Ackermann, 1999). As Sayeki points out in his paper "Anthropomorphic epistemology," adults, from lay people to scientists, use their imagination to project themselves into situations (Sayeki, 1989). They too dwell into their mental constructs to reach deeper understanding, and they do so, according to Sayeki, by literally dispatching little pieces of self, that he calls "kobitos" to inhabit their object of interest (little people in Japanese). Once "in there" via their imaginary doubles, they can act out and feel for what their kobitos experience, while remaining physically removed (Sayeki, 1989). Obviously, diving into situations and putting oneself in other people's shoes, or minds, won't suffice to learn, or grow. Being well-grounded, also requires its share of distancing and calls for achieving a balance between getting embedded and emerging from embeddedness (Kegan, 1982). In other words, every so often, people need to re-emerge by extracting themselves from the deep waters. They need to step back and look at things from afar. In their imagination, they generally achieve this by changing their stance in the world, by putting themselves in other people's shoes, or by adopting a god's eye's view, an
26
2. Constructing Knowledge and Transforming the World
altogether removed and all-encompassing view, that miniaturises the worlds they just inhabited (Ackermann, 1996). To conclude, people are both world makers and beings-inthe-world: they at once create their habitats, inhabit their creations, and become "inhabited" by them. In the world of their imagination, fusion (becoming one) and separation (removing oneself) coexist. Both contribute to their personal and cognitive growths. Creative Playpens for Constructive Learning The uses of projective imagination are at play in many forms of symbolic activities, from drawing to scientific modelling, from remote chats in social virtual environments (VE), to reading and writing. So are our attempts at anthropomorphizing and role-play. In the following sections, I focus on two specific aspects of pretense and symbolic play: people's abilities to dwell into their creations, and to fuse signifiers and signified as ways of becoming mindfully engaged. Their role in the constructive process is discussed in different contexts: from architects' drawing, to children's exploration of mathematical ideas, to digital kids' love affair with social virtual environments. The chosen contexts, or learning stories, are of two kinds: 1. Handling tools and driving machines 2. Exploring conversational writing in digital media. In both cases, the interactivity afforded by responsive artefacts (computers) is used to tap into people's tacit body smarts and situational wisdom. I show, through examples, that the apparently most primitive side of symbol use, empathic projection, is not just a key for natural learning but can be promoted by design to help children learn better. To conclude, I draw some implications for developmental psychology and education.
Dwelling into the Drawing A few years ago, Bonne Smith, a former student at the School of Architecture, MIT, designed a simple and compelling experiment. She asked some of her fellow students to sketch a floor plan of the house in which they lived when they were 5 years old. She encouraged her subjects to think aloud as they drew, and she videotaped the process (Smith, 1991). What this experiment revealed, is that the act of drawing was in itself a worldmaking technique. Moreover, the draftsman's engagement in the represented "site under construction" was quite anthropomorphic, surely more than one may expect from sophisticated architectural students. Alternatively becoming dwellers and creators, or giants and dwarfs, Bonne's subjects mentally moved in and out of the situation, seamlessly. They projected themselves into the pen-ball "as if it were a prosthetic device, and driving the pen around made it possible for them to travel along in their mind. The pen became a vehicle of mental teleportation. Dwelling in the drawing is what allowed Bonne's subjects to evoke, revisit, and reconstruct their lost memories. The most surprising aspect of this experiment is that the subjects' use of projected movement to bring back the place of their childhood increased with their level of sophistication as architects. It was much less prevailing among young children and nonarchitects. This came as no surprise to Bonnie, an architect herself, who reminded me that designers often imagine themselves and set themselves in motion in a space to be. They do so proactively to envision what that space may be. In her eyes, the experiment confirmed
2. Constructing Knowledge and Transforming the World
27
her intuition that people's ability to dwell into their drawing, or use drawing as a trailmaking technique (Nemirovsky and Monk, 1998) is one of the expertise that architects develop in the course of their studies and work. Here are 3 vignettes by Roy, Emily, and Andrew, architectural students whose protocols were rich with imagined movement (Smith, 1991): Roy: (thinks aloud) "/ am starting from the exterior and I'll be moving in. Here's the car (draws a car), the sidewalk moves perpendicularly from the driveway, past three shrubs, and up to the porch and then the front door. Then you move into the front hall like that..." (traces gesture of moving in and completes by drawing front door and entrance). What's remarkable in this account is that Roy is not the only one to move about. The sidewalk "moves" too, perpendicular to the driveway and past shrubs! Andrew reconstructed the lived space around the concept of "boxiness" —rectangular container —the shape and content of which he adjusted and refined as he moved through the virtual house. "This house was a breadbox. Just a good old American colonial [draws rectangley, brick box. Do you enter in the middle? OK [draws entry]. So you enter and there is this staircase [draws stairs middle of rectangle/ Yeah, that's pretty much the main focus when you come in [Andrew then proceeds to locate different spaces around the stairs and adjusts sizes by invoking action in and around them]. As he mentally moves into the salon "Wait? Can you walk behind the couch? the door? [He reaches out to grab an imaginary doorknob to determine the door swing]. Emily's use of imaginary projection was different yet. She spoke about the visual fields, or "perspectives," that unfold before her eyes as she walked through space in her mind: the view down the main street, the view of the facade. Holding these perspectives in mind helped her re-institute otherwise forgotten adjacencies and directions. Emily: "...whenyou go up the stairs, on each side you have... two regular doors that you can open, that you can push into...first thing you see is the reception desk. You 'II have a lot of, I think there's an old sofa here... " Note that all the subjects used the present tense in their accounts, which reinforces the idea that, in their minds, they were "really in there", as they were when they were kids.
Drawing Shapes by Driving Turtles Our bodies hold quite a bit of knowledge about space in their movement. Yet, much of this knowledge remains tacit, hidden in the beholder's habitual activity and experience. It needs to be brought to the mind's reach. One of Papert's greatest insights in designing Logobased Turtle Geometry, a software environment for building geometric shapes, was to tap children's knowledge about their own movement in space, and to use this knowledge as a lever to help them explore spatial relations and transformations. In turtle geometry, children "instruct" a computational creature to draw shapes by moving in prescribed directions by prescribed amounts. The turtle can be represented by cursor on the screen or, better, embodied as a mechanical toy-robot. The children communicate with the turtle using a language that it can "understand" (Logo programming language). Using
28
2. Constructing Knowledge and Transforming the World
Logo, a turtle can be made to move by typing commands at the keyboard. FORWARD 100 makes the turtle move in a straight line a distance of 100 turtle steps of about a millimetre each. Typing RIGHT 90 causes the turtle to pivot in place through 90 degrees. Typing PENDOWN causes the turtle to lower a pen so as to leave a visible trace of its path while PENUP instructs it to raise the pen. The commands and procedures available to drive the turtle are fairly intuitive to the child. They are also carefully chosen to enable the generation of many mathematically relevant and intriguing figures in space. The guiding principles behind Turtle Geometry are simple and much in tune with our views: Papert's turtles become extensions of self that the child controls using words. Giving directions — remote driving —encourages the child to reflect upon her own know-how and to express it precisely enough so that the machine can carry it out. "In teaching the computer how to think, children embark on an exploration about how they themselves think"(Papert, 1980, p. 19) . More important, Papert's turtles are designed to be "egocentric". Directions are given in reference to a turtle's position and heading and not as a function of some external reference system (xy co-ordinates). This requires that users put themselves in the turtle's shoes, literally, to figure out where it wants to go next. The syntax of Logo further provides a rich toolkit to assemble basic available operations (like rotations and translations) in interesting and surprising ways. Using computational tools and object responsiveness offers instant feedback, which helps sustain the interaction. In Mindstorms, Papert (1980) explicitly states the role of what we call mental teleportation: "A turtle has a position and a heading. In this, it is like a person or an animal or a boat (p.55). Children can identify with the turtle and are thus able to bring their knowledge about their bodies and how they move into the work of formal geometry (...) Drawing a circle in turtle geometry is body syntonic in that the circle is firmly related to children's sense of and knowledge about their own bodies. It is ego syntonic in that it is coherent with children's sense of themselves (one could say children's point of view or stance in the worlds" (p.63). Swinging a Graph Other learning environments have been designed to facilitate the articulation between world-making and world-dwelling. A case in point is the use of a motion detector by researchers at TERC (Technical Education Research Center), Cambridge MA, to help children learn about graphs. The display was designed by Nemirovsky and his team to augment children's control and understanding of graphical representations of mathematical variations over time (Nemirovsky, 1998; Tierney, Nemirovsky, Wright, Ackermann, 1993). I call the micro-world "swinging a graph" because, like Papert's turtles, it uses body motion as a vehicle to generate and control shapes. This time the activity is mediated by a motion detector, and the shape to be "drawn" is a time / graph on a computer screen. The motion detector used in these studies consists of a small button, the position of which is measured, of a sensor or electronic eye (also referred to as tower), and a computer. In interacting with the device, children hold the button or pin it on their shirt and move their bodies. They can also place the button on a moving object such as an electric train. The electronic eye (tower) measures the distance that separates it from the button at each moment in time, and outputs a graph that plots positions over time on the computer screen. Thus, by moving the sensitive button back and forth in front of the "eye," children can
2. Constructing Knowledge and Transforming the World
29
impact the graph's shape in real time: shapes vary as a function of the direction and speed of their, i.e., the button's, movement. Kids' first encounters with the motion detector are almost exclusively experiential, or sensorimotor. As they move back and forth with their button, they realise that the shape of the graph varies in reliable and somewhat principled ways. Very soon, though, the children learn to identify and to describe some of the changes they provoke. They tell us, for example, that as they move closer (to the tower) the graph goes up, and as they move away it drops; that if they move faster it becomes steeper, and if they slow down it flattens out. Sooner or later, kids also become interested in "swinging" very specific graph shapes. This requires that they understand, at least in action, what causes a graph's specific response. In doing so, they come to learn, for example, that they can't draw a circle or a square because the graph on the screen never goes backwards. As in Turtle Geometry, mediations have been introduced to help children move away from regulation-in-action to reflection. One of the mediations proposed was to remove the distance-sensitive button from the child's body, and to place it on the "face" of an electric train. The train was placed on a straight track in front of the motion detector. The child has now to move aside and to drive the train using a rotating knob, or dial. A next step in the mediating process, which was not explored at the time, would be to let the kids instruct or program the train, digitally. This would complete the cycle between moving one's own body, driving the train by hand using an analogical dial, and programming the train or give it a set of instructions. Switching back and forth between doing it oneself (engaging one's body) and giving instructions to "other" (instructing some responsive artefact) is what brings about deeper understanding (either about geometric or arithmetic operations). In both cases, the dynamic properties of interactive tools are used to tap into learners' knowledge-in-action, while mediations are offered to favour the passage from reflectionIN-action to reflection-ON-action. In both cases, "the idea is to give children a way of thinking of themselves as "doing science" when they are doing something pleasurable with their bodies" (Papert, 1980, p 68). Children learn because they are offered an occasion to use their own experience as a lever to actively explore mathematical ideas. Virtual virtues Social virtual environments (SVE) like chat rooms, Alphaworld, MUDS, offer yet another rich ground to explore how children and adults project themselves into fantasy worlds, as a way to revisit, enact and work through "real" issues. SVE can be thought of as digital stages for improvisational theatre, or psychodrama. They are fictionalising devices in Iser's sense. In MUDS, 6 "players encounter other players as well as objects that have been built for the virtual environment. MUD players can communicate with each other in real time, by typing messages that are seen by other players. Some of these messages are seen by all players in the same "room", but messages can also be designated to flash on the screen of Dungeons and Dragons was popular game in which a master created a world in which people take on fictional personae and play out complex adventures. The term "dungeon" persisted in high-tech culture to connote a virtual place. So when virtual places were created that many users could share and collaborate within, they were deemed multi-user dungeons, or MUDS, a new kind of social virtual reality, and the term MUD and the verb MUDding have come to refer to all of the multi-user environments. Some MUDs use screen graphics or icons to communicate place, characters, and action. Others rely entirely on plain text.
30
2. Constructing Knowledge and Transforming the World
only one specific player" (Turkic, 1995, p. 181). VE inhabitants, or avatars, are extensions of the human players. Their appearances and modes of interaction are mostly created and staged by the players themselves, in dialog with others. What's particular about Social Virtual Environments, as compared with other playgrounds for pretense, is the intricacy of the connection between users and their avatars, the immediacy and unpredictability of other player's response to one's virtual appearance, and the hybrid nature of the world itself, neither representation nor reality. As Turkic points out, VE-mediated exchanges deeply change the nature of our commitment to others, as well as our sense of selves. MUDs provide a stage for anonymous interaction in which players can choose a role as close to or as far from their "out of MUD self(ves)." (Turkle, 1995, p. 180) In social VR, as in good improvisational theatre, players do not recite scripts that are written by someone else. Instead, they are their own playwrights, choreographers, and actors. As in pretense play, staged events are both lived in and acted out. Players unfold scenarios and make drama come to life. Dwelling in social VE allows them to mediate their experience—live their lives on the screen—while remaining mentally engaged. It is the make-believe nature of the virtual space created, in conjunction with the truthfulness of the thoughts and feelings experienced through dialog with others, that make for the power of VE enactment. Attached to their avatars like a puppeteer to her string puppets, players act and feel through them. Virtual string puppets are both built by the puppeteer and brought to life by her. They are masks for idealised identities, allowing players to appear in a desirable light and hide those aspects of self that are not thought of too highly. Like Sayeki's kobitos, digital avatars are extensions of self that can be launched into the VE and made to act on one's behalf. It is the creator's strong connection / identification with their avatars that allows them to vicariously experience what they "go through". More easily than traditional puppettheatre, players can endorse multiple personae and launch them into different habitats at the same time. People's ability to put on the hats of multiple personae is not new in itself, and has its offline equivalents in adult psychodrama and face-to-face role playing games. What's different in VE, is the ubiquitous quality of self-appearances. It's like being in two "bal masques" at once or maintaining parallel streams of conversation. Along with Turkle, I think that digital fictionalising tools, enriched MUDS of sorts, can be used to help people, young and old, work out intriguing mental events, foster projective imagination, and construct their inner and outer worlds. To summarise, in VE, players can live things at a distance and get in touch with them at the same time. They can take risks on relatively safe ground. Using avatars allows them to remain anonymous, filter their appearance and control their level of engagement. Last but not least, the opportunity to come back again and again, changing face, and reconfiguring habitats (changing props) allows them to work out different versions of intriguing scenarios over extended periods of time. As in pretense, MUDers vary outcomes and rearrange story elements. Yet, as in psychodrama, they interact with others for good. What's unique in VE is that players can engage multiple dramas at once, or take on multiple hats in a same drama.
2. Constructing Knowledge and Transforming the World
31
3. Conclusions Fusion and separation are two poles of a continuum that are too readily opposed or placed in a developmental sequence. It has been our view, in this paper, that the abilities to put ourselves in another person's shoes, or mind, i.e. to change perspective and switch roles requires both fusion and de-coupling, being simultaneously "there" and "not there," embedded and disengaged. Fusion doesn't precede de-coupling, it accompanies it. Playing "what if or the ability to pretend (establishing a dialog between what is and what could be) is the means by which children as well as adults achieve the difficult a balance between getting immersed and emerging from embeddedness. Play is an important aspect in human learning, from identity building to constructing knowledge about the world. Erick Erickson defined play as a toy situation that allows us to reveal and commit ourselves in its unreality. Play operates within a transitional space (Winnicott, 1989), halfway between self and world, distinct from self yet under its control and, above all, more resilient that the world, in which the child can take safe risks. Throughout this paper, the articulation between make-believe and symbol-use has been a guiding connection to rethink the aims of representation. I explored the ways in which doing as if and playing what if inform people's conversations with—and through—artefacts. I discussed the benefits of children and adults' abilities to dwell into their symbolic creations and to treat symbols as objects in their own right. To situate my argument, I presented a series of learning stories or learning environments that support both world-making and dwelling into one's world. By way of conclusion, let me offer two suggestions that I wish were taken more seriously by developmental theorists and educators alike. The first suggestion is that people's abilities to fuse signifiers and signified and to treat signifiers as interesting objects by themselves are two powerful heuristics in creative symbol use. Their role in knowledge construction and scientific activities has been generally underscored as being primitive and generative of confusion. A second suggestion is that we seriously consider the significance of enactive forms of representations, from pretense play or simulacre, to simulations. As mentioned before, the French word simulacre and simulation sound very much alike. In both cases, a scenario or sequence of actions is being played out, which has been decoupled from its usually associated contexts. What's more, scenarios are not just described, as in writing or drawing, but they are actually run, or executed, as by a calculator. From objects-to-think-with (Papert, 1980) they become operations embodied, and people tend to relate to them as partners, with whom they share a task (Ackermann, In press). The difference between the two is the medium through which the performance is run. In simulacres and rituals, the medium is a human actor, or an actor's extension. In simulations, the medium is a human-made artefact, machine or program that runs a sequence of operations on your behalf. Simulations need not mimic something that exists. Their particularity is to execute operations that are only posed in language or notations. At a time when computational objects make it easy to run programs, model dynamic interactions, and simulate behaviours, people's ideas on what modelling is all about are deeply changing. So are their ways of relating to existing modelling tools. More than in the
32
2. Constructing Knowledge and Transforming the World
past, performance and simulation are granted a new place alongside language. It's time for us, researchers in cognitive development and educators, to catch up and revisit our own views.
Acknowledgements I thank Luc Steels and Mario Tokoro for having given me a chance to participate in the Future of Learning Workshop in Paco de Calheiros. It was a deeply engaging experience in many respects. I am grateful to my mentors, Jean Piaget, Barbel Inhelder and Seymour Papert, for their inspiration and support. I thank Ricardo Nemirovsky, for having given me an opportunity to work with them, and Sherry Turkic for her insights into MUDers' minds. I thank Herta Renk for encouraging me to write an earlier version of this paper, and Marleen Wynants for editing and improving it.
References Ackermann, E. (2001) Constructivism et Constructionism: Quelle difference?. In Constructivismes: usages et perspectives en education. (Volumes 1 et 2). Actes de Colloque. Geneva: Service de la recherche en edication / Cahier 8 / September 01. Pp. 8594 Ackermann, E. (1999) "Sich einrichten in Fantasie Raumen: Untersuchungen zum Gebrauch von Symbolen" (E. Renk Ed.) Lernen und Leben aus der Welt im Kopf: Konstruktivismus in der Schule. Neuwied, Kiftel: Luchterhand, pp. 79-99. Ackermann, E. (1999) "Enactive representations in learning: pretense, models, machines". In Learning Sites: Social and technological Resources for Learning (J. Bliss, P. Light, & R. Saljo, Eds) Elsevier Serie: Advances in Learning and Instruction, 144-155 Ackermann, E. (1996) "Perspective-taking and object construction: Two keys to learning". In Constructionism in Practice: Designing, Thinking, and learning in a Digital World (J. Kafai, & M. Resnick, Eds.). Mahwah, NJ: Lawrence Erlbaun, Publishers, pp. 25-37. Ackermann, E. (1993). "Systemes de notations chez 1'enfant: Leur place dans la genese de 1'ecrit". In Entretiens 3: Parole, Ecrit, Image. Paris: Editions Nathan, 51-69 Brown, J.S., Collins, A., & Duguid, P. (1989). Situated knowledge and the culture of learning. Educational Researcher. Vol. 18 (1). pp. 32-42. Carey, S. (1983) Cognitive Development: The Descriptive Problem. In Gazzaniga (Ed.). Handbook for Cognitive Neurology. Hillsdale, NJ: Lawrence & Erlbaum. Carey, S. (1987). Conceptual Change in Childhood. Cambridge, MA: MIT Press. Cole, M. (1996) Culture in mind. Cambridge, MA: Harvard University Press.
2. Constructing Knowledge and Transforming the World
Fox-Keller, E. (1985). Reflections on Gender and Science. New Haven. Yale University Press. Harris, P. (2003). Paper presented at the Jean Piaget Society Conference On Play and Development. Chicago, IL. Iser, W. (1987). "Representation: A performative act". In (Krieger, Ed.) The aims of Representation: Subject/text/History. New York: Columbia University Press. Reissued (1993) Stanford, CA. Stanford University Press. Kegan, R. (1982). The Evolving Self. Cambridge, MA: Harvard University Press. Kuhn, T. (1970) The structure of scientific revolutions. SecondEd. . University of Chicago Press. Lakoff, G. (1993) "The contemporary theory of metaphor" In Metaphors and Thought (Ed. A. Ortony). Cambridge University Press, pp. 202 -252. Lock, A., Service, V., Brito, A. & Chandler, P (1989) The social structuring of infant cognition. In A. Slater and G. Bremner (Eds) Infant Development Chapter 10. Pp 24372. Nemirovsky R., and Monk, S., (1998) "If you look at it the other way: An exploration into the nature of Symbolizing". In Cobb, Yackel, and Me Clain (Eds.,) Symbolizing and Communicating in Mathematics Classrooms: Perspectives on Discourse, Tools, and Instructional Design. Mahwah, NJ: Lawrence Erlbaum. Papert, S. (1980). Mindstorms. Children, Computers and Powerful Ideas. New York: Basic books. Perner, J. (1993). Understanding the Representational Mind. Cambridge, MA: MIT Press. A Bradford Book. Piaget, J. (1962) Play, Dreams, and Imitation in Childhood. New York: W.W. Norton & Co. Piaget, J. & Inhelder, B. (1967). The Child's Conception of Space. See especially "Systems of Reference and Horizontal-Vertical Coordinates." p. 375-418. New York: W. W. Norton &Co. Piaget, J. (1975). L 'equilibration des structires cognitives. Paris: PUF Reddy, M. (1993). The conduit metaphor: A case of frame conflict in our language about language" In: Metaphor and thought (Ortony Andrew, Ed.). Cambridge University Press, Chapter 2 p. 164-202. Rogoff, B., Lave, J. (Ed.) (1984). Everyday Cognition: Its Development in Social Context. Cambridge, MA: Harvard University Press. Sayeki, Y.(1989) Anthropomorhic Epistemology. Unpublished Paper. Laboratory of
33
34
2. Constructing Knowledge and Transforming the World
Comparative Human Cognition. University of California. San Diego Smith B. (1991) Dwelling in the Drawing: An inquiry into actual movement, imagined movement, and spatial representation. Unpublished Paper. Cambridge, MA.: MIT School of Architecture. Suchman, L; (1987). Plans and Situated Actions. Cambridge, MA: Cambridge University Press. Tierney, C., Nemirovsky, R., Wright, T., and Ackermann, E. (1993) "Body motion and children's understanding of graphs. In (J.R. Becker & B.J. Pence Eds.) Proceedings of the XV Annual Meeting of the North American Conference for Mathematics Education, pp. 192-198 Turkle, S. (1984). The Second Self: Computers and the Human Spirit. New York: Simon and Schuster. Turkle, S. (1995) Life on the Screen. New York: Simon and Schuster Vygotsky, L.S. (1962). Thought and Language. Cambridge, MA: MIT Press. Vygotsky, L.S. (1978). Mind in Society. Cambridge, MA: Harvard University Press. Wertsch, J. (1991) Voices of the Mind: A Socio-culotural approach to mediated action. Cambridge, MA: Harvard University Press. Winnicott, D.W. (1989) Playing and Reality. London, New York: Routledge. Witkin, H. & Goodenough, D. (1981). Cognitive Styles: Essence and Origins. International University Press.
Suggested Readings Aries, P. (1962) Centuries of childhood: A social history of family life. New york: Vintage Books. Ginsburg, H., an Opper, S. (1969) Piaget's theory of intellectual development. Englewood Cliffs. Prentice Hall, Inc. Harel, I. and Papert, S. (Eds) (1991). Constructionism. Ablex Publishing Corporation. Norwood, NJ. Hutchins, E. (1995) Cognition in the wild. Cambridge, MA: MIT Press. Leont'ev, A.N. (1932) Studies on the cultural development of the child. Journal of genetic psychology, 40, 52-83. Luria, A.R. (1928) The problem of the cultural development of the child. Journal of Genetic Psychology, 35, 493-506.
2. Constructing Knowledge and Transforming the World
35
Piaget, J. The Essential Piaget (Gruber, H. and Voneche, J. Eds) Basic Books, Inc. New York: 1977
EDITH ACKERMANN is professor of developmental psychology at the University of AixMarseille, France, on absence of leave. She is interested in the interaction between learning, teaching, children and the media, in collaborative learning, constructive play, and creative work / design, in technology-mediated environments. She studies how the conjunction of virtual and physical spaces support human learning, and how people-mostly childrendevelop senses of identity and community as they meet in actual, symbolic, or virtual worlds. She has pursued these interests in working with technologists, students, designers, and researchers in milieus concerned with learning and education. Ackermann teaches at the M.I.T. School of Architecture, Design Technologies, where she is appointed Visiting Professor, and consults for LEGO and MIT Media Laboratory. Previously, she was a Senior Research Scientist at MERL - A Mitsubishi Research Laboratory; a Junior Faculty at the MIT Media Lab and University of Geneva; and a Research Collaborator at the Centre International d' Epistemologie Genetique, Geneva, Switzerland, under the direction of Jean Piaget.
This page intentionally left blank
3
38
"In teaching and learning, one kind of event that arises naturally between children and their friends, or between parents and children in play, is affectionate, mutually enjoyed joking and teasing, the sharing of surprises inside familiar rituals. Learning should have surprises, and great pleasure can be had from taking part in unexpected events, or events that come too soon, ahead of the usual plan. That is why Jerome Bruner, a famous professor of psychology and education, became interested in how a baby can enjoy 'peek-a-boo' over and over again. He thought that the 'format' of a 'build up' that creates an expectancy in playful communication has an important message for educational theory, as well as for a theory of the grammar of language." Colwyn Trevarthen
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
39
3.
How Infants Learn How to Mean Colwyn Trevarthen
About 30 years ago developmental psychologists could not believe it when they were offered proof that newborn infants express states of mind and respond like persons. They had to accept a new explanation for behaviours that the famous modern developmental psychologists Freud, Piaget and Skinner had thought were impossible in such immature and inexperienced humans. Newborns actively seeking for experience and for communication must be doing something that is important for cognitive development and for the learning. Charles Darwin (1872) proved with photographs that infants show a variety of emotional expressions that communicate. He believed infants are sensitive to the emotions of other people. It seems clear now that babies have the essentials of a whole 'self with conscious awareness; they have a capacity for coherent purposefulness, all parts moving efficiently together, seeking experience (Figure 1 A). Within minutes of birth an awake baby can listen to changing sounds and turn to locate them in space, can feel the difference between her/his own parts moving and a different object that touches or moves against her/him. A newborn can see patterns in light, is curious to track a moving object, and is especially responsive to the odour and sound of the mother, identifying her as a person different from others. And all these innate abilities, now proved in many careful studies of babies' reactions, are adapted to learning through communication with a known person. Newborns can also imitate simple expressions of face, hands or voice, and expect to get a response from the person they are attending to (Figure 1 B). It seems that they are ready for picking up the motives of other persons, and this is now confirmed by a very original study carried out by a young Hungarian doctor and psychologist, Emese Nagy. She not only invited newborns to imitate attitudes, expressions and gestures, she patiently tested the infant's communicative initiative by withholding her presentations of expression after she had gained imitation, tempting the infant to 'provoke' an imitation from her. This was successful (Nagy and Molnar, 2003). Recordings of the infant's heart beat proved that he or she was in two quite different states of expectancy when imitating or when 'provocating'. Just before the time of imitating, the heart accelerated, indicating an intention to be active in a new way. With provocation there was an heart rate deceleration starting just before the baby moved, signalling a receptive focussing of attention, waiting for a response (Figure 1
3. How Infants Learn How to Mean
40
C). Thus Nagy showed, in agreement with observations of the Greek psychologist Giannis Kugiumutzakis, that a newborn baby can seek two-way imitative communication with a 'respectful' partner, exhibiting intentional preparation of complementary expressive and receptive conscious states. Kugiumutzakis (1993) had shown that neonatal imitations have two characteristics that prove they really are communicative:a) they are voluntary, in the sense that a goal-directed effort by the baby shapes them towards the form of the 'model1 by successively improved approximations. b) they are selective, matching 'special' forms of conversational expression that can be part of a communicative exchange with invention in it.
Figure 1
1. Intuitive Parenting Any person who wants a close and affectionate contact with a newborn infant has to display behaviours of gentle, playfully happy kinds that are unconsciously controlled and cannot be learned (Murray and Andrews, 2000). The similarities that appear in mothers' vocalisations to very young babies in different cultures, like the features observed when men or children attempt to talk with a baby, too, are evidence both for the universal needs of the newborn
3. How Infants Learn How to Mean
41
and for what Mechthild Papousek (1996) called 'intuitive parenting' motivation to meet these needs. The emotional 'codes' in infant and adult express the same kinds of impulse. Their affective expressions are adaptive to one another as sympathetic complements that confirm mutual awareness. Papousek thought this is the foundation for cultural learning. Who it is that addresses an infant is always important, as well as how they do so. Experiments that measure the preferential orienting of newborns to voices have proved that some foetuses have already learned in utero to identify their mother's voice. A newborn human can be alert to the face of a sympathetic caregiver speaking, drawing comfort from the expression of affection carried by the eyes and the loving voice. A mother greets her newborn with ecstatic cries in falling pitch, and gentle fondling, unable to look away from her baby. She touches hands, face, body with rhythmic care, and holds the infant close to her so they can greet one another. Her speech is a kind of singing, with high gliding gestures of pitch and repetition of regular gentle phrases on a graceful beat, leaving places for the infant to join in with coos, smiles and gestures of the hands or whole body (Fernald, 1989). Every move of the baby elicits a reaction from her. These exchanges are intricately co-ordinated with a subdued choreography that brings out matching rhythms and phrasing in mother and infant. And not only mothers are affected in this way. The example of a 'conversation' of coos between a father and a two-month premature newborn that has been subjected to acoustic analysis by an Australian musician, Stephen Malloch (1999) illustrates the essential features of a syllabic beat, phrasing and sympathetic co-ordination of emotion. In this case it is likely that the father's voice was transmitted to his daughter by vibration as much as by air-born sound as she rested one ear against his chest. The beautiful quiet games of expression that engage an affectionate parent with an alert young baby start the process of sharing experience that will carry the child's curiosity and eagerness to learn into a world of meaning that other people have created, and by which they direct their lives (Figure 2).
2. Sympathy Neurons A few years ago, physiologists recording from cells in the cortex of awake monkeys busily grasping pieces of food found neuronal "mirroring" elements. These appear to anticipate the evolution of imitations that make learning of human speech and language possible (Rizzolatti and Arbib, 1998). Cells in prefrontal cortex were active both when the monkey carried out a particular hand movement, and when the monkey sees or hears someone else making a similar movement for the same purpose. Apparently the monkey's 'picking up' nerve system 'resonates' with the action produced by an equivalent part of the human motor brain. It has become clear that the cells from which 'mirroring' effects have been recorded are part of widely distributed systems through the brain, that both move and feel with another subject. It might be better to call these the neural mechanisms of sympathy, which is a Greek word meaning 'moving with'. They reflect states of intention, awareness and feeling, not just shapes of movement, and they require a radical re-conception of the 'social brain' as an organ for sympathetic engagement between motives in physically separate moving bodies. Functional brain imaging with human subjects is beginning to explore the neural basis of sympathetic emotions that enable us to share the quality and vitality of consciousness as it comes to life in our separate bodies (Decety and Chaminade, 2003).
42
3. How Infants Learn How to Mean
More than in any other species, brain and behaviour development in humans makes no sense if the individual is considered in isolation, or if the impress of stimulation on 'plastic' neural circuitry, or the influx of information into a computational mechanism, is taken as the principle process driving development. We intuitively get into other persons' minds by actively sensing the impulses to action in their brains that enable them move the way they do (Trevarthen, 200la). One finding is particularly startling. When a 6 week old baby looks at a picture of a woman's face, this excites all the brain parts that, nearly two years later, will begin to acquire the skills of recognising a partner who is talking to us and for making and hearing speech in a particular language — including a 'face recognition' area, and areas for 'talking' and for 'hearing speech' (Tzourio-Mazoyer, et al., 2002). The baby is already trying to find 'common sense1 with the other person.
3. Human Companions Support Both Growth of Intelligence, and a Child's Emotional Health Babies communicate with caregivers to receive assistance with the regulation of their state of physical well-being. But they also do so to share motivation for learning. Older and younger brains are linked by emotional systems in the processes of care, and also in the quest for skill and understanding (Schore, 1994). Clever work by Jaak Panksepp and colleagues on 'laughter' in rats, attempting to identify its neurochemistry, has brought the wonderful realisation that 'joy' from tickling by playful companions may be good for the brain and for friendship between the playmates (Panksepp and Burgdorf, 2003). This study is part of a growing body of research that encourages belief in the importance of shared exuberance and pleasure in cognitive functions of the mammalian brain, in development of social collaboration, and in learning. The power of a child's brain to find motivation and confidence from sensitive communication with other persons gives a lifetime opportunity for compassion, and for assisting those in whom feelings and thoughts have become dysregulated by inherited or acquired damage to motives for companionship — for therapy. Activating beneficial states in brainstem and the right limbic cortex, which have been identified as the most important components in the regulation of states of self-awareness and of sympathetic communication, must be the principle effect of therapy aimed to restore healthy psychosocial life. The emotions brought about offer the means for recovery from dysregulation of the self (Trevarthen, 200 Ib; Schore, 2003a, b).
43
3. How Infants Learn How to Mean
Figure 2 Protoconversational Give and Take: Telling and Thinking in Turn, Before Language From birth to the beginning of language the child makes expressive moves to show impulses of thought that gain in meaning by being shared. These developments in the child's eagerness for communicating have profound effects on the behaviour of an affectionate, firmly 'attached' parent. The child is 'educating' the adult how to discover meanings that make sense, and joy, for both of them. Memories and ideas are built in communication, in increasingly rich narratives and games of imaginative 'mimesis' - the telling of imagined or remembered experiences by moving the body in a dramatisation (Goldin-Meadow and McNeill, 1999; Donald, 2001). All parts of the body, and all the modalities of sense, can play a part in this 'mind and memory sharing'. Visual life before birth must be almost zero, but, as the imitation studies show, newborns can see, and in a few weeks a baby is watching the other's eyes with clear focus, and obviously reacting to their direct regard. By 2 months baby and parent create a lively 'protoconversational' form of communication, the most obvious developments being a marked increase in the accuracy of the baby's eye-to-eye contact and a quickening of all responses (Trevarthen, 1977, 1980). Now the baby can be attracted to play a part in an exchange of expressions that resembles the body movements, gestures and vocal intonations of adult face-to-face conversation - it is conversation stripped of words (Figure 2 A). Sight of others, and how they express themselves, is not the only way a baby can get into communication, of course. Well supported, a totally blind baby, or one that is both deaf and blind, can develop happily and well, seizing other ways of sensing a partner and engaging with them. The communication is carried by any awareness of the impulse by which a person can express their mind (Trevarthen, 1993; Trevarthen and Aitken, 2001). The behaviours of protoconversation have been analysed in great detail now, and researchers are impressed with the infant's sensitive responses and fine appreciation of
44
3. How Infants Learn How to Mean
timing (Trevarthen, 1993, 1999; Trevarthen, Kokkinaki and Fiamenghi, 1999). The attention of the infant to a partner's voice and face expression, shown by a 'knit brow and jaw drop' expression of fixed orientation, is followed promptly by recognition of positive, affectionate elements in the sight and sound of a partner's feelings. The rhythms of a parent's inviting vocalisations, touching and movements of the head, eyebrows and mouth evoke first a smile of recognition then an animated 'utterance1. At the climax of expression the 6- to 8-week-old assumes an attitude of declamation, vocalising, gesturing and changing head posture in a co-ordinated assertion of communicative purpose, frequently removing orientation from the partner, who has been closely attended to until this moment, as if 'carried away' by the ideas being expressed. The adult, seeing this, is stimulated to give an encouraging, praising kind of reply that matches the level of affect of the infant's 'utterance', interpreting it with a parallel emotional form or 'affect attunement' that gives back or complements the feeling (Stern, 2000). Then the infant re-orients to the adult and observes 'thoughtfully' what they are expressing, before being excited again to smile and make another utterance (Figure 2 B). For some moments the baby really seems to be 'thinking', watching the partner with a quiet unsmiling face. Each phase of the 'chat' is characterised by particular initiatives and emotional responses in interaction with the other person's mind. Shifts in interest are growing in the baby's consciousness that will lead to shared exploration and use of places and things. After 3 months, the baby will develop increasingly adventurous plans, making more vigorous use of their senses and limbs, seeking to explore and to form concepts of objects, negotiating purposes and the tempo of experience more vigorously with others (Figure 3). Then the infant starts also to become interested, not only to look about for himself or herself, but also to follow the shifts of gaze of the other person who is occupied in seeking and acting on objects. The mother is likely to be the first companion and teacher in these games with objects, but soon father, siblings and others can join in the infant's expressions of curiosity and anticipation as they seek information together. The developments in expressive body signs before speech — from protoconversations of 'primary intersubjectivity' with two-month-olds, through games of the person and person-personobject games in the middle of the first year, to 'secondary intersubjectivity' or 'co-operative awareness' and protolanguage at the end of the first year — show that communication of intentions, experiences and feelings is the foundation on which co-conscious use of experience and the precise references and recollections of language are built (Bruner, 1983; Trevarthen, 1980, 1988; Tomasello, 2003) (See Figures 4 to 7). The Sadness of Lost Contact The emotions in protoconversation have been tested by observing what happens when the human response to a baby's interest is blocked or fails. If a mother holds her face still for a minute in the middle of face-to-face play with her two-month-old, this causes the infant to turn away and show distress (Tronick et al., 1978; Murray and Trevarthen, 1985). A similar pattern of anxiety and sadness appears when the mother presents the uncommunicative manner of simulated depression. Real post-natal depression interferes with the infant's communication and cognition, and, if it persists, is accompanied by limited cognitive development in later months (Murray and Cooper, 1997; Tronick and Weinberg, 1997). An unhappy, unresponsive adult cannot be a good companion and teacher. A Double TV apparatus, in which a young baby and the mother are communicating via a video and sound link while in separate rooms, seeing and hearing each other face-to-face
3. How Infants Learn How to Mean
45
and life size on two monitors, allows a critical experiment to be performed. A lively, positive portion of the recording from the mother is replayed a few minutes later to the infant (Murray and Trevarthen, 1985; Trevarthen, 1993). The distress this produces shows that a 2-month-old is extremely sensitive to the contingent responsiveness of the live mother's expressions, which is lost when a physical recording is offered of what was a cheerful live conversation. If, conversely, a portion of the recording of a communicating infant is replayed to the mother, she experiences an uneasy loss of contact, and she may conclude that the infant is avoiding her or that she is somehow giving the wrong signals. This makes her confused and unhappy. Live communication has to be just that, a real time engagement of feelings and impulses to communicate. A delay or an inappropriate response proves that the other is 'out of touch'.
Figure 3
4. Playing Games and Tricks: Using Rules of Moving to Make Surprises and Share Ideas About Acting with Friends In the games infants play with their mothers we observe first signs of an intelligence that wants to share ideas and fix meanings in a conventional, 'made up', code. We can trace the stages by which the baby's mind grows to think in prepositional narratives (stories) and in metaphorical representations (imaginary connections between feelings of being involved with things, happenings and people) (Figure 7). These are the kind of events in the mind that can be put into language. The infant, we find, is much more than the self-sufficient, solitary explorer and problem-solver that Jean Piaget studied, more than a private mover and thinker. A baby's mind has an appetite to learn by picking up ideas from a community
46
3. How Infants Learn How to Mean
of people, accepting what they have invented (Figures 5 and 6). The baby's consciousness is adapted to 'being moved' or being taught. It is ready for 'cultural learning' (Vygotsky, 1978; Bruner, 1996; Rogoff, 1990; Trevarthen, 1988). After 3 months, a baby's games show the infant leading the mother's attention out of conversation to explore the shared world. The baby is curious about things that invite sensory tracking and manipulation (Figure 3 A and B). Developing strength in body and brain confer a power to support the head in free rotation and project the arms out to grasp objects. Objects are now discovered at an accelerating pace, by the combined application of hands, eyes, ears and mouth to pick up useful information. Because the baby often shifts attention from her face to an object, the mother has to change her tactics to keep the dialogue alive. Often fathers, with a more vigorously playful approach, find they have a special appeal as playmates, too, now (Figure 2 B). Either parent adopts a challenging, quick-changing way of making fun. Soon play routines are discovered that facilitate lively and enthusiastic participation and make the baby laugh when half-expected surprises occur, or when the baby knows they are coming (Figures 3 B, 4 and 7). The force and rhythms of movement in games of vocalisation and dance get stronger. Patterns of expression are created with a person that are more complex than any play an infant can have with an object. The parent's game can be a bit frightening, but great fun. It is like the rough and tumble play that many young mammals (such as kittens, puppies and young rats and monkeys), and some sociable birds (e. g. parrots), develop with their peers (Beckoff and Byers, 1998). But humans begin this experimenting with communicative action early, long before they can move about on their own, and in the company of exceptionally playful parents. Baby and parent are exploring ways of negotiating plans and projects in ways that will lead before the end of the first year to the learning of highly conventional 'acts of meaning', and then to a sense that words people say are of special importance as signs (Halliday, 1975). These 'dispositions to learn', and the matching parental 'dispositions to teach', are uniquely human (Figures 5, 6 and 7). In the teasing and joking games and musical/dancing entertainment of infants approaching the middle of the first year, communication is complex and highly productive, but the topic remains the communication itself, i.e. it remains 'metacommunicative' or 'communication about communication' (Bateson, 1956). The baby anticipates and responds with increasing skill, and seems to be learning rules that students of the grammar of language recognise: such as 'entailment', 'qualification', 'contrast', 'repetition with changed emphasis', 'subordination', 'opposition', 'release' and so on. The infant senses different poetic or dramatic forms or 'melodies', and enjoys their repeated presentation. This provides the 'text' for awareness of causal relationships between the many objects that now can be discriminated, identified and recognised in the infant's awareness. It all depends on a very special human imagination for planning how the body will act. Child games, and pedagogy, show what Merlin Donald calls an 'executive suite' of 'domain general skill clusters' (including metacognition or cognition about cognition), 'self reminding', 'self-triggering of memory', 'whole body imitation', 'symbolic invention', 'complex skill hierarchies') that are absent or poorly developed in apes (Donald, 2001). The mind of the baby is well on the way to sharing the thinking of and invented common sense.
3. How Infants Learn How to Mean
47
5. Narratives of Experience with Increasing Meaning The 3- to 6-month-old baby can now share the drama of a favourite game routine over many seconds (Bruner and Sherwood, 1975; Trevarthen, 1999). A 'narrative sequence1 of feelings and expressions is created that has a beginning, a development, a climax and a resolution or denouement, the last being carefully paced to be startling and provocative of laughter, is what catches the infant's interest in a classical baby song or action chant (Ratner and Bruner, 1978; Trevarthen and Malloch, 2002). These songs and rhymes have the same prosodic or musical features in different languages, which testify to the complexity of infants' communicative motives everywhere at this age, motives that will later serve for catching the meanings in an historical heritage of symbolic forms.
Figure 4 'Person-person games' become transformed into 'person-person-object games' (Figure 3) and the baby is given reactive 'toys' that make noises or roll or bounce when handled, shaken or thrown (Trevarthen and Hubley, 1978). Again, the incorporation of such objectexploring behaviours into communication games or negotiated 'formats' (Bruner and Sherwood, 1975) is peculiarly human. The narrative 'plots' of such play have no equal in the fighting and chasing games of young mammals, even though some animal play may incorporate objects that are chased-for or fought-over. In a modern home there are electronic toys that are programmed to play tunes or carry out spontaneous simulations of human or animal action. Such 'robots' are exciting, indeed - but we have to wonder if they
48
3. How Infants Learn How to Mean
threaten to eliminate the creative interaction such as a baby can only have with a live mind in a live body. The infant can gain interest in experiences through the presentations that others make to change these experiences in momentarily surprising but quasi-predictable and often repeated playful ways. Identities and relationships or operations are 'dramatised' by the emotions coupled to their demonstration (Trevarthen and Hubley, 1978; Papousek and Papousek, 1981). The infant's imitative awareness appreciates the excitement of a partner's attempts to 'attune with' what the infant is doing and the changes of action and experience that the infant is showing (Stern, 2000). Soon the infant is expert at seeking for these feelings of the other and for the habitual forms of display (Figure 3). Upon meeting a new situation or object, or after performing some deliberate 'act', the infant's learning and skill is evaluated or 'tagged' by the other person's emotions. Hopefully that person is amused and impressed by the baby's 'cleverness' when he or she 'shows off (Reddy, 2001 a, 2003). Usually parents are very proud of what their children can do (Figures 1 to 7) The infant about 7 or 8 months old is about to crawl. Now he or she can also share interest in an expanded world of places and things with other persons, taking up their direction of gaze or their pointing (Scaife and Bruner, 1975). This means that by that age, at least, the other person's awareness can be linked to the infant's awareness in a common space of experience (Figures 5 and 6). Such 'joint awareness' is recognised as a key element in communication that leads to language (Tomasello, 2003).
Figure 5 Cultural learning is what sets human beings apart. All historically contrived communities of meaning and belief, including languages, depend on motives and passions of 'companionship' different from those involved in primary regulation of a child's attachments. Thus a mother's play with her infant is a 'cradle of thought' (Hobson, 2002) as well as an 'external regulator' of what the baby's body needs or a protector from stress. Importantly, her role as playmate and companion in meaning can be taken over by any other sympathetic person, even a child, whom the baby has learned to trust. A 6-month-old can negotiate interests, intentions and feelings with two same age peers, with no adult help (Selby and Bradley, 2003). Infants are sociable in the community before they walk or talk. This is the beginning of friendships in learning that form naturally between playmates long before the child is ready for classroom instruction. Play among toddlers is imaginative,
3. How Infants Learn How to Mean
49
creating invented worlds and 'myths' of adventure that range far beyond present circumstances (Nelson, 1996).
6. Reflecting on Oneself in the 'Moral' Appraisal of Others, Being 'Sensible': The 'Me' Who 'Shows Off, and is Wary of Strangers Around six months after birth, an infant can be observed imitating gestures and mannerisms, and showing them to companions (Uzgiris, 1984). Affectionate praise from family members entices the baby to perform a 'trick' that impresses other persons — he or she is acting like a performer or clown (Reddy, 200la). Sometimes the baby's 'sign' is offered to a stranger, apparently in an attempt to 'break the ice' of an awkward, anxious confrontation. But it is usually puzzling to the stranger and may be laughed at, which is a kind of mockery that distresses the infant (Trevarthen, 1990). A bold baby may act coquettish as if to impress a stranger, but remains ambivalent. Babies laugh easily now, but this is combined with a sensitivity to who laughs 'at' or 'with' the infant. What is called 'fear of strangers' seems to be linked with a tendency to try out supercilious expressions and clowning when the baby is with a partner who is either not supportive — for example, when the mother keeps a still face — or observant of the infant's self, as when a stranger is trying to 'make friends' (Trevarthen, 1990). I think the baby knows that the meaning of actions is connected with relationships, with a history of being with the persons who shares them. All self-referred, other-sensitive emotions become stronger and clearer in the second six months. The baby begins to recognise his or her self in a mirror (Figure 4) and toward the end of the first year infants prefer pictures of infants of the same sex as him or her self. When looking in a mirror the baby makes 'experiments' with babbling, face grimaces and hand gestures, and repeats imitations of the exaggerated expressions others offer in play. By 8 and 9 months boys and girls are different in this; the boys tend to posture or 'challenge' more, while the girls show a greater range of expressions and are generally more 'friendly' (Trevarthen, Kokkinaki and Fiamenghi, 1999). From 6 months both boys and girls are clearly interested in their images and entertained by them at the same time as they are very aware when they are the focus of another's attention and interest (Reddy, 2003). Self-consciousness in the presence of others' appraisal would appear to be preparatory to learning in an active 'zone of proximal development' in which the partner can give cognitive, logistical and practical support (Vygotsky, 1978), and the child can begin to learn through 'guided participation' (Rogoff, 1990) (Figures 6 and 7). The infant is getting insight into the other's states of mind, and can be said to have a more critical 'sense of intersubjective self (Stern, 2000).
50
3. How Infants Learn How to Mean
Figure 6 Humorous teasing provokes the older infant to laughter and coyness (Reddy, 200 Ic), but subtle signs of self-other -awareness are seen very early. Reddy (2000) has documented coy smiling at a mirror image in 2-month-olds, and she lists many other behaviours that signify awareness of others at this age (Reddy, 2003). In the development of selfawareness we observe a transformation of motives to share consciousness and purposeful actions with others that were evident at birth. As the baby picks up comical ways of handling things, and starts showing objects for others, seeking congratulation, holding them up as a joke with a look, a gesture, a grimace or a vocalisation that can become a coded act that is transmitted between the child and the 'audience' of the other, the object is part of an 'act of meaning' (Halliday, 1975) or 'protosign (Trevarthen, 1990)' (Figures 6 and 7). They 'make sense' in the loyal and affectionate communication of the family or with familiar playmates, but may not with strangers (Trevarthen, 1990, 2002). It is the quality of assured mutual friendship that counts. My observations lead me to believe that 'stranger fear' is an anxiety of seeming foolish with a person who can't comprehend. It is a direct and strongly felt emotion — one of the 'complex' or 'relational' emotions, expression of which makes the infant seem a sophisticated social being long before language, and before any system of beliefs or explicit 'theory of mind'. The fact that it increases as the infant gains in self-awareness and
3. How Infants Learn How to Mean
51
'showing off with familiars, suggests that the beginning of coded communication is part of a motivation for defining a cooperative group in which meaning is consolidated by daily practices that must be protected against 'strange' customs or beliefs. Subsequent developments into language confirm this idea. Around nine months it may be observed that an infant is taking a new, much more definite initiative in games, attempting to direct a partner to behave in a certain way or to repeat a playful act (Trevarthen, 1977; Hubley and Trevarthen, 1979). The infant may show considerable skill in teasing an adult, as if knowing how to manipulate feelings and predict reactions (Reddy, 200 Ib). Such behaviours appear to be part of the transition to a new constructive sharing of interest in things and tasks, the beginning of 'protolanguage' (Halliday, 1975).
7. The Musicology of Human Communication Rhythm and the Nature of Moving In the 1970s and 80s psychologists and developmental linguists found fascinating evidence of musical talents in babies (Papousek, 1996; Trehub, 1990). Hitherto unsuspected musical listening skills were proved for infants as young as 4 months. The Papouseks described the 'intuitive parenting' mode of vocal communication with infants in musical terms, stressing the modulation of affect provided by parental tones and rhythms (Papousek and Papousek, 1981). A diary study of their daughter documented the infant's enjoyment of nursery songs, and her private practice of acquired musical forms. The concept of 'attunement', by which Stern describes how the parent picks up on infant expressions reflecting their beat, emphasis and intonation, encapsulates his strong musical sensibility (Stern, 2000). By 5 or 6 months infants quickly recognise songs or recorded music often heard ~ stopping to listen, smiling in recognition, then bouncing and waving arms and legs with the tune (Figures 3 and 4). A baby's selective orientation to musical sounds, critical discrimination of musical features of sound, and vocal and gestural responses that are timed and expressed to contribute to a joint musical game confirm that music, which is clearly a cultural achievement of human society, has strong roots in human nature (Trevarthen and Malloch, 2002. Why are babies, and mothers, so musical? What ways of behaving show their musicality? Research on the temporal foundations of expression and the development of narratives of expression has been advanced by musical acoustic analysis of vocal interactions between mothers and infants, and a theory of Communicative Musicality has been developed (Malloch, 1999), which defines features of'pulse', 'quality' of sound, and 'narrative' form that underlie the innate dynamics of moving and thinking and the sympathetic transmission of mental events between subjects of any age. Clearly, the expression of'music' for an infant is to be understood in the sense of the ancient Greek word, UA)aiKe (musike), i.e. inclusive of all temporal arts — theatre, dance, poetry, as well as what we know as 'music'. Musicality manifests its fundamental features in the ways that infants behave in interaction with the expressions of motive forces in other human beings (Dissanayake, 2000). The core element of a cheerful baby song is a four line stanza lasting about 15 to 30 seconds, with a base pulse around andante , a tripping iambic rhythm, simple pitch shifts
52
3. How Infants Learn How to Mean
and rhyming syllables at specific points, and variations in the beat to regulate excitement in the last two lines (Trehub, 1990; Trevarthen, 1999). A lullaby to sooth a sleepy or unhappy infant will be slower with gentle rocking rhythms. Infants quickly learn to recognise songs and hand clapping or bouncing games, eagerly joining in when invited (Figures 3 and 4), and old people remember songs and games they learned as infants. Early musicality certainly has a powerful role in building memories. It marks with emotional signatures the identity of persons and ritual events. After very few months an infant can 'make music', making singing sounds and banging objects rhythmically. He or she seems to have found a proud performer's personality, who can share a musical 'joke1 that surprises and pleases. Taken with the infant's clear preferences for particular companions, this musical 'showing off looks like the beginnings of his or her social identity as member of a group with known habits, celebratory experiences and acting skills that are valued for the bonds that they represent and reinforce (Trevarthen, 2002). Cultivation of intrinsic musicality is a way of declaring allegiance with a friend or to a social band (Blacking, 1988). New evidence on the place of affect in intelligence, and on how emotions regulate brain development, cognition and learning, makes the infant's sensitivity to expressions of emotional narrative in musical form more comprehensible (Panksepp and Bernatsky, 2002). Musicality may be at the source of the ability to be socialised in the human way. The Prehistory of Human Narratives An infant enjoying the message of a familiar baby song is showing us how human meaning began. As Marc Turner has made clear, the whole of our consciousness and life together is made of story-telling (Turner, 1996), and the stories are made of metaphors with affective quality that describe agents going places and doing things with energy and style, conscious of how their bodies move and how experience is made by moving. John Blacking asked if dance, music and other artistic activities are not, "essential forms of knowledge which are necessary not only for a balanced personality but also for the development of cognitive capacities." (Blacking, 1988, p. 91). In arguing the case for 'affective culture', he said, "Passion is as important in scientific endeavour as is compassion in artistic vision." (loc. cit., p. 93). The ways in which infants present themselves as performers and masters of creative acts indicates that one of the principal outcomes of affective understanding with others is the development of a secure recognised and valued 'identity' — being somebody, placed in the world with others as a 'knower' and 'doer', 'making sense' of oneself. All humans revel in a freedom of gestural action that is intensely shared, and we use our whole bodies to act out meanings (Goldin-Meadow and McNeill, 1999; Trevarthen, 1999). We need this whole body sense to learn language (Trevarthen, 2003). These are the reasons why prehistorians are giving the evolution of music precedence over the emergence of language as a means of communicating experience (Cross, 1999; Morley, 2002) Contemporary human minds, and certainly the minds of our distant ancestors were not simply rational devices for categorisation of perceptions or strategic processing of instrumental tasks. Aesthetic and mythic forces are integral to the management of a human view of the world. These forces come from the impulses to act and create with anticipation, from dynamic evaluations of experience in action, and from memories of exciting contingencies of acting in the natural world. They must also have depended on enriched
3. How Infants Learn How to Mean
53
sense of being in intelligent company sharing an affective culture (Dissanayake, 2000). Hominids acquired the semantic power of words, and the rational abstractions that words facilitate, after evolving messages of 'mimesis', in rhythmic narratives turning experienced events into allegories of body posturing and stepping, gesticulations of the hands, facial expressions of eyes and mouth, and modulated cries of the voice (Donald, 2001). The important addition Merlin Donald makes to cognitive theory is that he defines a form of story making that is embodied and performed in regulated rhythmic time, that conveys narrative in transitions of feeling. In this he agrees with Blacking's insistence on the vital role of an 'affective culture'. The greatest challenge for the psychology of language is how narratives in thought are composed by blending an infinite variety of impressions with a sense of agency—of actors doing things with purpose and emotion, seeking and evaluating goals in a world of territories, places, objects and natural events, as well as animals and people (Turner, 1996). Being involved in talk or writing and reading is being involved in conceiving and executing movements in a world that has other persons. Language is not an object that has existence outside active human hopes and interests, and outside the history of friends, families and communities. It results from brains regulating bodily events in time and space, and recording the experiences (Varela et al., 1991; Damasio, 1999; Donald, 2001).
8. Education for Culture: How Common Sense Grows in the Community Knowing and Belonging: Meaningful Relationships It may be convenient for the management of a complex industrial society to plan education as a construction of skills according to curricular formulae that are 'quality tested' at each prescribed stage. It may be practical to focus on each child as an intellectual athlete in training, who strives to master facts and rational skills that society wants. But this is an artificial, one-sided, cultivation of the natural process by which children can and want to master cultural knowledge, and for which adults enjoy giving natural encouragement. The 'intent participation' of the child in mastery of meaning (Rogoff et al., 2003) must be respected, and shared (Figures 5, 6 and 7). The emotions involved in teaching and learning are often overlooked. It is perhaps a product of social organisation and planning in industrial societies, and a reaction to the abuses of child labour as Rogoff and her colleagues (loc. cit.) indicate, that mothers are seen primarily as protectors or keepers of their infants, who may or may not be substituted by sufficiently sensitive surrogates, and teachers are seen as instructors, neither being understood as available friends and collaborators who benefit from the infant's or child's instinctive companionship and playfulness. The psychology we have created to support our society and measure its effects on individuals is one that treats the emotional and intellectical success of each person separately. We have come to think of ourselves as communicating just information about what each of us perceives is real and practical, and, perhaps, what each of us thinks. Inevitably our conceptions of sympathetic and intuitive mental life have become over-cognitive and impoverished. Sigmund Freud made a powerful effort to redress this imbalance, but left the unconscious mind at the mercy of language, the vehicle of clear thought. He did not have full confidence in the intuitive communication of purposes and concerns by non-verbal means. He did not investigate how
54
3. How Infants Learn How to Mean
the process of interested, practical and joyful communication of thoughts and imagining begins in infancy, and nor did Jean Piaget. Research on the development of communication and cooperation in the first two years of a child's life indicates a different way of conceiving human teaching and learning and the propagation of cultural knowledge (Bruner, 1996; Trevarthen, 1988; Trevarthen and Aitken, 2001; Hobson, 2002). And it leads to a new view of the relationship between cultural learning, the form and movement of the human body and the anatomical peculiarities of the human brain. The rich and rapidly developing sociability of infants and toddlers indicates that this brain has evolved for sharing knowledge and skill. The essential motivating and emotional systems are laid down long before birth as generators preparing the capacity of a human being for initiative in acting and experiencing, and for intersubjective communication. Being at home in a community is essential for the confident teaching of a parent, and learning of a child. Maya Gratier (1999) has found that the musical quality of a mothers communication with her infant, which signals her intimate pleasure with the baby and confidence in herself, maybe affected if she has been taken from her home culture to a strange land. Gratier calls this the effect of emotions of 'belonging'. She believes she has shown that consciousness of meaning, begins in an intimate coordination of the motives of mother and infant, in their seeking to generate and share experience within one space and time of companionship. Her data show that the capacity of the mother to successfully share experience with her infant through dynamic negotiation of states of interest, purpose and emotion is predicated on her having her own 'sense of belonging'. If a mother cannot find a secure attachment to her adult world that gives her a coherent identity with its specific grammar and expressive signature, she may not be able to meet her infant's desire for company. A mother brings to her child both personal and cultural ways of moving, speaking and singing. These influences shape the infant's developing sense of self and agency. They may be said to constitute a person's primary sense of belonging or "core culture" (Hall, 1989), the deeply rooted sense of being in tune and in time with certain nonverbal, intuitive, communal ways of being. In happy communication mother and infant are anticipating the other's intentional motions. They balance one another on the cusp of the future, each poised to step in at exactly the right moment, that is at the moment which is most meaningful to the other and most motivated by them. This concept of 'looking ahead' to the course of agency recalls Husserl's notion of'protention'. Infants appear to have an innate "future sense", and they instantly sense meaning in the timing of the other's expressive gestures. By making joint narratives, mother and infant come to share history and invoke community (Figures 5 and 7). The narrative form contains both the security of an ending and the exciting tension of its timing. The contrasting elements of security and tension, or familiarity and novelty, or repetition and variation, constitute the crucial vectors that give impetus to the infant's developing mind, and the one-year-old has begun to find fascination for the 'topics' of this sharing. This is the 'flow of common sense' (Figure 8). The infant's future sense may loose clarity and direction if he or she is not provided the opportunity to develop these natural skills. And a mother's future sense may become perturbed in a variety of ways. A depressed mother, for instance, seems to have trouble in precisely that way: her interactive behaviour, as we know, is less contingent and thus less meaningful to the infant, she has lost her sense of time and within a dialogical framework is
3. How Infants Learn How to Mean
55
unable to share her mental space with her infant with an even, playful grace. This difficulty of maintaining hopeful 'time in the mind' has been highlighted by many researchers as being characteristic of depression; people who suffer from depression have difficulty projecting themselves into the future, making plans, envisaging possible worlds. A depressed person lacks self esteem — he or she experiences shame in company of confident others. Gratier applied Malloch's musical acoustic analysis to talk and games between mothers and infants over the first year. She identified episodes lasting between 20 and 30 seconds presenting the universal phases of a narrative -- introduction, development, climax, and 'resolution. Recordings of the spontaneous vocal interaction of mothers who recently emigrated (from India to the U.S.) and their infants aged 2 to 6 months were compared with those of non-immigrant dyads. She analysed the spontaneous interactions of 30 dyads where the mother had recently emigrated and compared them with those of 30 nonimmigrant dyads. The immigrant dyads were less expressive in their timing than the non-immigrant dyads. They tended to present more predictable or rigid timing patterns as well as less clearly defined narrative constructions. Certain of these mothers, although they did not fulfil diagnostic criteria for clinical depression, also presented lower levels of self-esteem, clear signs of withdrawal, with lack of social support and 'cultural conflict' in their representations of mothering. This study highlights the link between the mother's sense of 'being at home' or belonging, and the self-confidence and pride that go with it, and her ability to share and negotiate coherent and meaningful experience with her infant. Most immigrants are able to adjust their ways of being and thinking so as to be in tune with their new world, and they do so by negotiating feelings and behaviors with people around them. For others, this process of rebuilding a world-view and an identity is inhibited or thwarted. Motherhood itself can bring about a certain amount of stress and identity confusion, and some immigrant mothers become "trapped" between two world-views, experiencing conflict with regard to their own identity and to the representations they have of their infants. Human consciousness is communitarian. It develops through cooperative awareness, and depends on communicating a personal narrative. The essential motive for cultural learning is a sympathetic, mimetic sense of being an actor having adventures with companions, not just imitating or sharing joint attention to objects and events. It is best to feel 'at home1 (Figures 5, 6 and 7). Getting Hold of Symbols as Cultural Tools, and Starting to Talk There are impressive gains in social awareness as children begin to speak in second year. (Trevarthen, 1988, 1990). They have fluent inventive or creative fantasy, recognising objects of technical, industrial or artistic importance; roles and postures; socio-dramatic performances with pretend emotions; moral positions. These ideas and skills are imitated from others and spontaneously displayed to portray child's personality as observed by others. The way a child gains entry to language brings out the primacy of interpersonal cognitions and their emotional regulation. Emotions enable transfer of evaluations and reinforcements to the infant and the orientation of the infant to present circumstances. They enable the infant to learn by being taught.
56
3. How Infants Learn How to Mean
Figure 7 As words are learned, they attach to the known persons in communication, to their actions or to the objects that are indicated in their shared interest and actions, and that give an account of shared memories. Different toddlers with differing parental support and differing temperaments may show preference for objects or persons as topics in their first speech (Locke, 1993). Conventional use of tools, roles and rituals of performance is mastered in the second year, beginning before speech, preparing the way for rapid learning of the maternal language (Figure 7). The change from manipulating for private gain or discovery to imitation of others' directives, indications and evaluations leads to ideas that have already been coded in words in the communication of older members of the community round the child (TomasellO, 2003). The cooperative learning of language needs flexibility of imagination, which is expressed in the pretend play that flourishes among toddlers and preschool age children (Trevarthen and Logotheti, 1987; Nadel et al., 1999). Objects and actions become assimilated into shared purposes, and this can change identity or meaning. Things can stand for other things - a banana can be a telephone (Leslie, 1987). When the desired objects or events are absent and no substitute presents itself, they may be created entirely in imagination to satisfy the motive for shared play and communication. The child can invent play actions alone, too. But all play motivated by pretense is creating meanings that are ready to be shared. The development of the child's imagination and future learning are dependent the ability to exchange points of view and imitated ideas with a companion, an ability that is deficient in an autistic child. Jacqueline Nadel shows how quickly collaborative parent-infant play transfers to communication between toddlers (Nadel and Peze, 1993). She has recorded how immediate imitation of actions and utterances is used by 18-month-olds for non-verbal negotiation of purposes and for sharing meaning, and she underlines the pleasure and humour of sharing signalled by exuberant gesture and vocal prosody. Social 'self-
3. How Infants Learn How to Mean
57
confidence1 depends on a sense of security with communication of meanings and actions, and this confidence fluctuates with developmental change (Trevarthen and Aitken, 2003). Around the middle of the second year, at 15 fo 20 months, a child has a fragile social identity, and (as in a 'replay' of the sensitivity of the 7 to 8 month-old) is acutely aware of the potential difficulties of communication with strangers (Kagan, 1981). It would appear that the imagination that is reaching out to learn how other persons categorise their experiences is sensitive to the risks of imitating without understanding. The withdrawal of a shy child into a private fantasy world may have much to teach us about the pathology of symbolic thought. Developments in preschool years show how mastery of thinking is dependent on a free and flexible regulation of contact with other minds by emotions (Figure 7). Identity and Pride, and the Shame of Misunderstanding: Emotions Beyond Attachment for Care In every human relationship the pleasure of active discovery and the mastery of experience and skill are regulated by interpersonal or moral feelings. As long as essential needs are provided for and the child is not distressed, sick or exhausted, these feelings, of pride in knowing and doing, and embarrassment or shame at not understanding or 'being out of things', are asserted powerfully in every young child. They guide the growth of experience, and they do so by emotional regulation of the growth of the brain. They are manifested out of control in disorders of mania and depression. I believe that the intensely shared pleasure of pride in knowledge and skill that others applaud, as well as the feeling of shame in failure that threatens loss of relationship and hopeless isolation, are as important to the mental health of every human being as the emotions that seek comfort and care for the body (Figure 4). Indeed, I would suggest that attachment itself, if it is a friendship and not just the very asymmetric relationship between a weak and immature 'patient' and sensitive caregiver, is animated by emotions of shared discovery and the creation of inventive art. Even the most disciplined and authoritative teaching regime requires a minimal mutual respect between teacher and taught, or its purpose is totally defeated. I suggest we need a 'circle of attachments' - of emotionally charged relationships to care and comfort givers, to places and things that foster our discoveries and activities, and to friends and companions in adventure, discovery and invention, persons who share the impulses of our thinking and acting, and of play with roles and meanings (Figure 8). I believe human relationships are motivated by innate emotions that display and evaluate shared purposes and interests, and that these emotions of 'attachment for companionship' are just as important for mental health as the emotions of attachment for care. Darwin (1872) is the acknowledged pioneer of modern studies of emotions. He sought to classify human and animal emotions according to their expressions in body attitude and movement. He did not restrict his classification of human expressions to a short list of discrete 'basic' emotions (fear, anger, surprise, sadness, joy, disgust and perhaps contempt). He included such moral qualities of motivation as 'love', 'tenderness', 'sulkiness', 'hatred', 'contempt', 'guilt', 'pride', 'shame' among those he attributed to expressions of animals and children, and he called them emotional expressions. His other terms denote different states of bodily feeling or of reaction to objects ('suffering', 'anxiety', 'grief, 'despair', 'joy', 'anger', 'fear', 'disgust'), or states of experiencing and thinking ('meditation', 'determination', 'patience', 'surprise').
58
3. How Infants Learn How to Mean
Stern (2000) has insisted on the importance for communication with infants of 'dynamic' and 'relational' emotions which cannot be described by the names of'categorical' emotions. For the dynamic emotions Stern uses the descriptive terms: "crescendo," "decrescendo," "fading", "exploding," "bursting," "elongated," "fleeting," "pulsing," "wavering," "effortful," "easy". These invite comparison with the 'sentic forms' of the musician Manfred Clynes, who attempted to identify different force curves or gesture shapes that convey momentary feelings in musical sounds (Clynes and Nettheim, 1982). Human sympathy and shared consciousness is governed by powerful emotions of pride and shame, of generosity and guilt, of moral goodness or evil. A case can be made that such 'complex emotions' have primary importance in the development of human consciousness (Draghi-Lorenz, Reddy and Costall, 2000). These feelings of human relating cannot be derived from the cognitive emotions of surprise, curiosity, and pleasure in mastery, which are appropriate for regulating actions on non-sentient objects. Emotions of satisfaction, or of disappointment and annoyance, expressed by young infants solving, or failing to solve, instrumental problems, are significant to others as manifestations of knowing and discovering 'in a human way'. They 'communicate' what is going on in the infant's mind. Evidently the 'relational' emotions of companionship are by far the most elaborate and significant for human mental growth and integration of the child into society, even though the emotions implicated in the making and breaking of attachments may have greater immediate importance in psychosomatic health and well-being. A sense of beauty, and of what looks or sounds ugly, also is a vital part of human emotion (Turner, 1991). The pleasing or disturbing properties that persons feel in empathic awareness given to objects of shared interest, especially in the appreciation of rare objects and those made artificially with special care to give them high social value, are made evident in aesthetic judgements.. Thus a carefully crafted artefact, a work of poetry or art, is made part of vital common experience. The human emotions by which cultural experience is propagated, and creativity is given moving value, appear to have evolved by elaboration of'experience seeking' and 'attachment regulating' motives emotions of subcultural species. They are matters of 'taste' that are profoundly influenced by sympathetic response to the preferences and aversions of respected persons. Art is, as Ellen Dissanayake says, the product of the intimacy that brings infants to meaning in parental care (Dissanayake, 2000), and that explains why it can be a source of solace for a trouble human spirit. All human cultural achievements arise shared meaning, even when they appear to be lonely products, of creatively dreaming or of adventurous risk-taking in thought or action. New thoughts, how an individual imagines of experiences generated by actions, make sense through the thinker sharing their originality and 'truthfulness' with others, who judge their value and 'significance'. Human effort is directed to build relationships through cooperative and inventive works. These are the reasons why a cognitive, information-processing, perception-categorising, memorising approach to human cumulative intelligence is unable to comprehend its social motivation or to perceive its intersubjective psychological foundations in evolution. The theory of cognitive modules in separate heads contrasts with a psychobiological theory of culture as a product of human will to make and understand in relationships and communities. Cultural learning is not just a cognitive achievement of the human mind. It is a new development in animal social initiative, and in ways of relating intelligently (Figure 8).
3. How Infants Learn How to Mean
59
Artists, painters, sculptors and poets, actors and musicians, explore their private discoveries in consciousness and their deepest feelings, leaving a record of what they find. Their work communicates these lived experiences. When they create in authentic relation to their feelings, they make statements that can open up our consciousness and change how we value what we experience. Scientific discovery and technical invention share with the arts a foundation in human curiosity and the capacity to convey to others the motives and excitements of finding out, or making. That is why an education that is both broad in its scope and democratic in its sharing of opportunities and findings has the most lasting value - socially and practically.
Moving, Being Moved, and Meaning
Sharing Meaning With Confidence
Figure 8 9. Educating Common Sense The Comenius Principle The 17th Century Czech educator Komensky (Comenius) led the way to an enlightened view of how children learn best. He wrote, in a book that was translated into many languages, the following: "My aim is to show, although this is not generally attended to, that the roots of all sciences and arts in every instance arise as early as in the tender age, and that on these foundations it is neither impossible nor difficult for the whole superstructure to be laid; provided always that we act reasonably as with a reasonable creature." (John Amos Comenius (1592-1671) The School of Infancy. Translated by D. Benham. London, 1858. Quoted by Quick, 1910).
60
3. How Infants Learn How to Mean
Comenius believed that older persons — parents, siblings, teachers of many kinds — naturally respond to a child's vitality and eagerness to understand. They feel they want to help. They can learn how to do so from the child, who in this is their teacher. A desire to know more and to gain skill in ways that other trusted people recognise and encourage is the defining feature of young human nature. It is the instinct that makes 'cultural learning' happen (Figures 5, 6 and 7). This may be an old and obvious idea, and Comenius expresses it well. Nevertheless, intricately rational and busily occupied adult minds often deny it, forgetting how they themselves learned. The inventive curiosity and love of social attention of the young child is easily seen. But many who assume authority and expertise have difficulty accepting the innateness of human sympathy in action and knowledge, probably because there is no obvious rational explanation for it — nothing in the physical or biological world to compare it with, no computational system that can simulate it. Even in academic psychology, intentions and feelings of the young child are given less attention, simply because they do not fit scientific models of how minds work. This neglect by those who claim expertise can have inhibiting effects on the practice of applied psychology, and on the training of teachers. We do not know how an imaginative sympathy for the human-created view of the world could be born in the human mind, so it is easier to conceive it as constructed from experience, by instruction from outside. Thus is the adult world led to teach, but not to learn from the child. Lev Vygotsky (1978), Michael Halliday (1975), Jerome Bruner (1983), John Locke (1993), and Michael Tomasello (2003) have all emphasised that a child picks up words by noticing what other persons do with it, aided by shared human interest. Acts negotiating social participation with emotion come earlier in development than intention-directing 'protoimperatives1, just as 'person-person games' came before 'person-person-object games' in the middle of the first year (Trevarthen and Hubley, 1978). The early stages of'grammar' learning, getting the syntactic and functional conventions right for sentences, is not simply a matter of coordinating vocalisations with intentions and attentions - requests, pointing, showing, giving. It has concern for human feelings and sensitivities which form the backing texture of all live communication and 'experiencing together'. 'Joint attention', strongly associated with the picking up words, is not just a convergence of lines of sight and directions of instrumental action. It involves 'mutual attention' as well (Reddy, 2003) ~ subtle awareness of moods and purposes, of instantaneous shifts of interest and emotional reactions that the infant has practised with familiar playful company through the first year. Meaning and language continue to grow in personal relationships of shared pleasure and trust (Figure 7). Children and adults alike are easily caught in dramatic make-believe, identifying the roles of'characters'. Infants play with emotional narratives long before they talk, and toddlers create dramas together before they have any demonstrable 'theory of mind'. This gives both the reason and the means for language learning. From 2 to 6 children make things, tell and listen to stories, create drama, acting fantastic parts, dance and exhibit all sorts of musical skills. Their appetite for cultural forms of life is enormous and their perception of human roles is rich and penetrating.
3. How Infants Learn How to Mean
61
Teaching Well and Teaching Badly: Teachable Learners Need Learnable Teachers Human learning requires the young to develop deep insight into the thinking behind the moves that elders make, and into the expressions of approval or disapproval that signal the value of experiences. The young learn an historically established cosmology and meanings that were invented among ancestors who were intensely aware of one another's interests and purposes. Discoveries of new ideas and ways of acting are especially attractive to human minds, even to those that are very young and inexperienced. This curiosity for meaning has innate motivation, and it needs an exceptional emotional sensitivity that goes far beyond the expression of immediate bodily needs. The process can build in comfort, confidence and confiding in a loving family and community, or it can fall prey to fear and distress, loneliness and self-doubt. This is why infants crave the consistent sensitive company of an affectionate parent or other person who can be trusted to sustain the shared memories that have been discovered in their company. Research inspired by Vygotsky has shown how an expert and novice interact together in the 'Zone of Proximal Development', where, by collaborating with the expert, the novice becomes able to achieve a goal that would otherwise be impossible by his or her effort alone (Vygotsky, 1978). Wood and Bruner (1976) identified techniques of'scaffolding1 by which adults assist a child's efforts in solving a problem or completing a task (Figures 6 and 7). Rogoff and colleagues contrast 'adult-run' and 'child-run' ways of teaching and describe a 'community-of-learners' model where all share responsibility for learning (Rogoff et al., 2003). In many cultures 'intent participation' in meaningful and immediately useful activities is the way children become able to contribute to their community and culture (Rogoff et al., 2003). This contrasts with the 'instruction' model of education in industrialised and literate cultures where the value of what is taught may not be immediately evident to the learner. Although learning takes place in any, and indeed all, kinds of educational practice, the community-of-learners model has been shown to promote in the pupil greater co-ordination with others, and responsibility for his or her own learning and motivation. Where the adults are supportive and provide leadership, rather than controlling all interactions, the participants work together, with each may serve as a potential resource for the others. The degree to which all are actively trying to learn and understand determines how satisfying a learning environment will be. The teacher should be prepared to learn continuously from the learner, being 'guided, directed and inspired' by the children's understanding. Bruner (1996) conceives this kind of classroom organisation as a subcommunity of mutual learners with the teacher orchestrating the proceedings'. He highlights the crucial role of the school, as an an institution that judges a child's performance and subsequently facilitates a process of self-evaluation. "What characterises human selfhood is the construction of a conceptual system that organises, as it were, a 'record' of agentive encounters with the world, a record that is related to the past but that is also extrapolated into the future - self with history and with possibility" (Bruner, 1996, p.36). "The management of self-esteem is never simple and never settled, and its state is affected powerfully by the availability of supports provided from outside. These included...above all the chance for discourse that permits one to find out why or how things didn't work out as planned." (Bruner, 1996, p.37) As in the sharing of experience that grows between an infant and a parent, the timing and quality of expression in the communication are important in teaching and learning with
62
3. How Infants Learn How to Mean
older children. Erickson (1996) has explored how timing and 'contextualization cues' (such as volume and pitch shifts in the voice and in body motions) function in classroom discussion to help participants to anticipate impending courses of action: "Timing appears to be what holds the whole ecology of interaction together in its performance. The relative temporal location of the various actions of interlocutors is an important aspect of the ordering of the collective activity of conversation in both its reciprocal and its complementary aspects..." (Erickson, 1996, p.34). The interaction is held together by 'cadential patterns' that produce 'points of emphasis in the verbal and nonverbal behaviour stream' (loc. cit., p. 54). The members of the group collectively organise their attention and thus contribute to listening and speaking in a smooth, coherent manner that is cognitively stimulating. It can be hypothesised that, in cases where the teaching is not effective, the smooth running of turn-taking behaviour within an organised temporal framework will break down. Robb and colleagues have undertaken a study in Scotland to explore ways of analysing teacher-talk or 'teacherese' in a small class of young children (Robb et al., 2003). The aim was to identify characteristics of communication that can facilitate or impede learning and retention. Teacher-pupil interactions were recorded on video with sound and analysed in detail, to identify key characteristics of satisfying and effective communication. A target group of 8 teachers, who were selected by their colleagues and consultant Educational Psychologists as particularly skilled communicators in the classroom, were compared with a control group of 7 experienced teachers on a standardised teaching activity with groups of pupils matched for age and ability. All were teachers of Primary 4-7, with equal levels of experience and general competence. Each teacher chose 6-8 pupils from their class to take part in a group discussion. Teachers were asked to introduce the following imaginary task to the pupils and to orchestrate a discussion in their usual way: "You are to spend a week on an uninhabited island with a partner. You have to find the buried treasure. Plan everything that you will take with you for the entire week." Audio and video recordings were made of each group discussing the project. Five minute excerpts were chosen for analysis by the researcher who was blind to the categorisation of teachers. The video tape was analysed for teachers' contact, mood, verbal and non-verbal initiatives and responses to pupils. Discourse analysis was also employed to identify types of comments. The audiotapes were analysed for length and frequency of turns, timing, phrasing, pitch and tone of teacher/pupil vocalisations. Although there was no overall difference between targets and controls in their positive responses to pupils, the target group were more attentive to the pupils and also more lively and humorous. They made more supportive approving responses and higher levels of positive initiations to pupils than the controls. Target teachers also made more accomodating, reflective and metacognitive interventions. Voice spectrograph analysis demonstrated that the target group showed more reciprocal attunement with their pupils in terms of pitch plot contours and phrasing. They were also more sucessful in eliciting pupil participation.
3. How Infants Learn How to Mean
63
The study indicates that satisfying and productive communication in a classroom should embody the same principles of reciprocity, mutuality, attunement, regular timing and turntaking behaviour that have been found sustain parents' communication with infants. The same principles of intersubjectivity apply. Young children learn naturally in dynamic relationships of admiration and trust. Thus, the living emotions in the teacher's voice, language and non-verbal behaviour may be as important as the timeless facts and routine exercises of thought and skill he or she may be wanting to transmit. These rules of relating apply for an infant learning at home with family, in preschool, in classroom instruction through primary and secondary school, and in the university. They are at work in informal recreational learning, too. They also promote individual effort and the discovery of achievement through experience of'flow 1 in mastery of difficult tasks.
References Bateson, G. (1956) The message This is play'. In B. Schaffner (Ed.), Group Processes: Transactions of the Second Conference, (pp. 145-242). New York, Josiah Macy Foundation. (Republished in, Bateson, G. (1973) . Steps to an Ecology of Mind. Frogmore, St Albans: Paladin.) Bekoff, M. & Byers, J. A. (1998). Animal play: Evolutionary, Comparative and Ecological Approaches. New York: Cambridge University Press. Blacking, J. (1988). Dance and music in Venda children's cognitive development. In G. Jahoda and I. M. Lewis (eds.), Acquiring Culture: Cross Cultural Studies in Child Development., (pp. 91112). Beckenham, Kent: Croom Helm. Bruner, J. S. (1983) Child's Talk. Learning to Use Language. New York: Norton. Bruner, J. S. (1996). The Culture of Education. Cambridge, MA.: Harvard University Press. Bruner, J. S. and Sherwood, V. (1975). Early rule structure: The case of peekaboo. In J. S. Bruner, A. Jolly, K. Sylva (eds), Penguin.
Play: its Role in Evolution and Development. Harmondsworth:
Clynes, M. & Nettheim, N. (1982). The living quality of music: Neurobiologic basis of communicating feeling. In M. Clynes (ed), Music, Mind and Brain, the Neuropsychology of Music. New York/London: Plenum Press. Comenius (1592-1671) The School of Infancy. Translated by D. Benham. London, 1858. Cross, I. (1999). Is music the most important thing we ever did? Music, development and evolution. In Suk Won-Yi (ed),. Music, Mind and Science, (pp. 10-29). Seoul: Seoul National University Press. Damasio, A. R. (1999). The Feeling of What Happens: Body, Emotion and the Making of Consciousness. London: Heinemann. Darwin C. (1872) The Expression of Emotion in Man and Animals. London, Methuen.
64
3. How Infants Learn How to Mean
Decety J. & Chaminade T. (2003) Neural correlates of feeling sympathy. Neuropsychologia 41:127-138 Dissanayake, E. (2000): Art and Intimacy: How the Arts Began. University of Washington Press, Seattle and London. Donald M (2001) A Mind So Rare: The Evolution of Human Consciousness. New York, NY and London, England: Norton Draghi-Lorenz, R., Reddy, V., & Costall, A. (2001). Re-thinking the development of'non-basic' emotions: A critical review of existing theories. Developmental Review, 21(3), 263-304. Erickson, F. (1996). Going for the zone: Social and cognitive ecology of teacher-student interaction in classroom conversations. In D. Hicks (ed.), Discourse, Learning and Schooling, (pp. 29-62). New York: Cambridge University Press. Fernald, A. (1989). Intonation and communicative interest in mother's speech to infants: Is the melody the message? Child Development, 60, 1497-1510. Goldin-Meadow, S. & McNeill, D. (1999). The role of gesture and mimetic representation in making language. In, M. C. Corballis and E. G. Lea, (Eds.), The Descent of Mind: Psychological Perspectives on Hominid Evolution, (pp. 155-172). Oxford: Oxford University Press. Gratier, M. (1999) Expressions of belonging: The effect of acculturation on the rhythm and harmony of mother-infant vocal interaction. In "Rhythms, Musical Narrative, and the Origins of Human Communication". Musicae Scientiae, Special Issue, 1999-2000, pp. 93-122. Liege: European Society for the Cognitive Sciences of Music. Hall, E. T. (1989) The Dance of Life: The Other Dimension of Time. New York: Anchor Books. Halliday, M. A. K. (1975). Learning How to Mean: Explorations in the Development of Language. London: Edward Arnold. Hobson, P. Macmillan.
(2002)
The Cradle of Thought: Exploring the Origins of Thinking. London:
Hubley, P., and Trevarthen C. (1979). Sharing a task in infancy. In I. Uzgiris (Ed.), Social Interaction During Infancy, (New Directions for Child Development, Vol. 4), (pp. 57-80). San Francisco: Jossey-Bass. Kagan, J. (1981). The Second Year of Life: The Emergence of Self-Awareness. Cambridge, MA: Harvard University Press. Kugiumutzakis, G. (1993). Intersubjective vocal imitation in early mother-infant interaction. In J. Nadel and L. Camaioni (eds),. New Perspectives in Early Communicative Development, (pp. 2347). London: Routledge. Leslie, A. (1987) Pretense and representation: The origins of "theory of mind". Psychological Reviews, 94:412-426.
3. How Infants Learn How to Mean
65
Locke, J. L. (1993) The Child's Path to Spoken Language. Cambridge MA and London: Harvard U. Press Malloch, S. (1999) Mother and infants and communicative musicality. In "Rhythms, Musical Narrative, and the Origins of Human Communication". Musicae Scientiae, Special Issue, 19992000, pp. 29-57. Liege: European Society for the Cognitive Sciences of Music.. Morley, I. (2002) Evolution of the physiological and neurological capacities for music. Cambridge Archeological Journal, 12(2): 195-216. Murray, L. & Trevarthen, C. (1985). Emotional regulation of interactions between two-month-olds and their mothers. In Field, T M & Fox, N A (eds) Social Perception in Infants, (pp. 177-198) Norwood, NJ: Ablex. Murray, L., & Andrews, L. (2000). The Social Baby: Understanding Babies' Communication from Birth. Richmond, Surrey, UK: CP Publishing. Murray, L., & Cooper, P. J. (eds.) (1997). Postpartum Depression and Child Development. New York: Guilford Press. Nadel, J. & Peze, A. (1993). Immediate imitation as a basis for primary communication in toddlers and autistic children. In J. Nadel and L. Camioni (Eds.), New Perspectives in Early Communicative Development, (pp. 139-156). London: Routledge. Nadel, J., Guerini, C., Peze, A., & Rivet, C. (1999). The evolving nature of imitation as a format for communication. In J. Nadel & G. Butterworth (Eds.), Imitation in Infancy, (pp. 209-234). Cambridge: Cambridge University Press. Nagy, E. & Molnar, P. (2003) Homo imitans or Homo provocans? Human imprinting model of neonatal imitation. Infant Behavior and Development, (in press). Nelson, K. (1996) Language in cognitive Development: Emergence of the Mediated Mind. New York: Cambridge University Press. Panksepp , J. & Bernatzky , G. (2002) Emotional sounds and the brain: The neuro-affective foundations of musical appreciation. Behavioural Processes 60: 133-155 Panksepp, J. & Burgdorf, J. (2003) "Laughing" rats and the evolutionary antecedents of human joy? Physiology & Behavior, 79: 533- 547. Papousek, M. (1996). Intuitive parenting: A hidden source of musical stimulation in infancy. In I. Deliege and J. Sloboda (Eds.), Musical Beginnings: Origins and Development of Musical Competence, (pp. 88-112). Oxford: Oxford University Press. Papousek, M. & Papousek, H. (1981). Musical elements in the infant's vocalization: Their significance for communication, cognition, and creativity. In L. P. Lipsitt and C. K. Rovee-Collier (eds.), Advances in Infancy Research, Vol. 1, pp. 163-224. Norwood, NJ: Ablex. Quick, R. H. (1910) Essays on Educational Reformers. London: Longmans, Green, and Co. Ratner, N. & Bruner, J. S.(1978) Games, social exchange and the acquisition of language. Journal of Child Language, 5: 391-400.
66
3. How Infants Learn How to Mean
Reddy, V. (2000) Coyness in early infancy. Developmental Science, 3(2): 186-192. Reddy, V. (200 Ib) Mind knowledge in infancy: understanding attention and intention in the first year. In, G. J. Bremner and A. Fogel (eds.) Blackwell Handbook of Infancy Research, pp. 241-264, Oxford: Blackwell Reddy, V. (2003) On being the object of attention: implications for self-other consciousness. TRENDS in Cognitive Sciences, 7(9): 397-402 Reddy, V. (200la) Infant clowns: the interpersonal creation of humour in infancy. Enfance 3, 247-256 Rizzolatti, G., & Arbib, M. A. (1998). Language within our grasp. Trends in the Neurosciences, 21: 188-194. Robb, L., Simpson, R., Forsyth, P. & Trevarthen, C. (2003) Satisfying and effective teacher-class communication. (In preparation. Presented to the Early Child Education Research Association Conference, Glasgow, September, 2003). Rogoff, B. (1990). Apprenticeship in Thinking: Cognitive Development in Social Context. New York: Oxford University Press. Rogoff, B., Paradise, R., Arauz, R. M., Correa-Chavez, M., & Angelillo, C. (2003) Firsthand learning through intent participation. Annual Review of Psychology, 54: 175-203. Scaife, M. & Bruner, J. S. (1975) The capacity for joint visual attention in the infant. Nature, 253: 265-6. Schore, A. N. (1994) Affect Regulation and the Origin of the Self: The Neurobiology of Emotional Development. Hillsdale, NJ: Erlbaum. Selby, J. M. & Bradley, B. S. (2003). Infants in groups: A paradigm for study of early social experience. Human Development, 46:197-221 Stern, D. N. (2000): The Interpersonal World of the Infant: A View from Psychoanalysis and Development Psychology. (Originally published in 1985. Paperback Second Edition, with new Introduction) Basic Books, New York. Tomasello, M. (2003) Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press, Trehub, S. E. (1990). The perception of musical patterns by human infants: The provision of similar patterns by their parents. In M. A. Berkley and W. C. Stebbins (eds.), Comparative Perception; Vol. 1, Mechanisms, (pp. 429-459). New York: Wiley. Trevarthen, C. (1977). Descriptive analyses of infant communication behavior. In H. R. Schaffer (Ed.), Studies in Mother-Infant Interaction: The Loch Lomond Symposium, (pp. 227-270). London, Academic Press.
3. How Infants Learn How to Mean
67
Trevarthen, C. (1980) The foundations of intersubjectivity: Development of interpersonal and cooperative understanding in infants. In D. Olsen (ed.), The Social Foundations of Language and Thought: Essays in Honor of J. S. Bruner, pp , 316-342. New York: W. W. Norton.. Trevarthen, C. (1988). Universal cooperative motives: How infants begin to know language and skills of culture. In G. Jahoda and I.M. Lewis (Eds.), Acquiring Culture: Ethnographic Perspectives on Cognitive Development, (pp. 37-90). London: Croom Helm. Trevarthen, C. (1990). Signs before speech. In T. A. Sebeok and J. Umiker-Sebeok (Eds.), The Semiotic Web, 1989, (pp. 689-755). Berlin, New York, Amsterdam: Mouton de Gruyter. Trevarthen, C. (1993). The self born in intersubjectivity: An infant communicating. In U. Neisser (Ed.), The Perceived Self: Ecological and Interpersonal Sources of Self-Knowledge, (pp. 121-173). New York: Cambridge University Press. Trevarthen, C. (1999). Musicality and the Intrinsic Motive Pulse: Evidence from human psychobiology and infant communication. In "Rhythms, Musical Narrative, and the Origins of Human Communication". Musicae Scientiae, Special Issue, 1999-2000, pp. 157-213. Liege: European Society for the Cognitive Sciences of Music. Trevarthen, C. (200la). The neurobiology of early communication: Intersubjective regulations in human brain development. In A. F. Kalverboer and A. Gramsbergen (Eds.), Handbook on Brain and Behavior in Human Development. Dordrecht, The Netherlands: Kluwer Academic Publishers, (in press). Trevarthen, C. (2001b): Intrinsic motives for companionship in understanding: Their origin, development and significance for infant mental health. International Journal of Infant Mental Health, 22(1-2): 95-131. Trevarthen, C. (2002) Origins of musical identity: evidence from infancy for musical social awareness. In, MacDonald, R., David J. Hargreaves, D. J. and Dorothy Miell, D. (Eds.) Musical Identities, (pp. 21-38). Oxford: Oxford University Press. Trevarthen, C. & Aitken, K. J. (2001) Infant intersubjectivity: Research, theory, and clinical applications. Annual Research Review. The Journal of Child Psychology and Psychiatry and Allied Disciplines, 42(1) 3-48. Trevarthen, C. & Aitken, K. J. (2003) Regulation of brain development and age-related changes in infants' motives: The developmental function of "regressive" periods. In: Regression Periods in Human Infancy, Heimann M, Plooij F, eds. Mahwah, NJ: Erlbaum, pp. 107-184 Trevarthen, C. & Hubley, P. 1978. Secondary Intersubjectivity: Confidence, confiding and acts of meaning in the first year. In A. Lock (Ed.), Action, Gesture and Symbol, (pp. 183-229) London: Academic Press. Trevarthen, C. & Logotheti, K. (1987) First symbols and the nature of human knowledge. In J. Montangero, A. Tryphon & S. Dionnet (eds.), Symbolisme et Connaissance/ Symbolism and Knowledge, Cahier No. 8, Jean Piaget Archives Fondation. (pp. 65-92). Geneva: Jean Piaget Archives Fondation.
68
3. How Infants Learn How to Mean
Trevarthen, C. & Malloch, S. (2002) Musicality and music before three: Human vitality and invention shared with pride. Zero to Three, September 2002, Vol. 23, No, 1: 10-18. Trevarthen, C. (2003) Language development: Mechanisms in the brain. In: G. Adelman and B. H. Smith (eds) Encyclopedia of Neuroscience, 3rd Edition, with CD-ROM. Amsterdam: Elsevier Science. Trevarthen, C., Kokkinaki, T. & Fiamenghi, G. A. Jr. (1999). What infants' imitations communicate: With mothers, with fathers and with peers. In J. Nadel and G. Butterworth (Eds.), Imitation in Infancy, (pp. 127-185) Cambridge: Cambridge University Press. Tronick, E. Z. and Weinberg, M. K. (1997) Depressed mothers and infants: Failure to form dyadic states of consciousness. In L. Murray, P.J. Cooper (eds.) Postpartum Depression and Child Development, (pp. 54-81) Guilford Press, New York. Tronick, E. Z., Als H., Adamson L., Wise S., and Brazelton T. B. (1978) The infant's response to entrapment between contradictory messages in face-to face interaction. Journal of the American Academy of Child Psychiatry, 17: 1-13 Turner, F. (1991). Beauty: The Value of Values. Charlottesville: University Press of Virginia. Turner, M. (1996). The Literary Mind: The Origins of Thought and Language. New York/Oxford: Oxford University Press. Tzourio-Mazoyer N, De Schonen S, Crivello F, Reutter B, et al. (2002) Neural correlates of woman face processing by 2-month-old infants. Neuroimage 15:454-461 Uzgiris, I. (1984). Imitation in infancy: Its interpersonal aspects. In M. Perlmutter (ed.), The Minnesota Symposia on Child psychology: Vol. 17. Parent-Child Interactions and Parent-Child Relations in Child development, (pp. 1-32). Hillsdale, NJ: Erlbaum. Varela, F. J., Thompson, E. and Rosch, E. (1991) The Embodied Mind. Cambridge, MA: MIT Press. Vygotsky, L.S. (1978). Mind in Society: The Development of Higher Psychological Processes. Edited by M. Cole, V. Steiner, S. Scribner and E. Souberman. Cambridge, Mass: Harvard University Press. Wood, D. and Bruner, J. (1976) The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17, 89-100.
3. How Infants Learn How to Mean
69
COLWYN TREVARTHEN is Professor Emeritus of the University of Edinburgh. He has developed interest in a wide range of subjects in earlier research, including the relationship between brain mechanisms of perception and those for action, and the origins of intersubjectivity ~ awareness between persons. His current research focuses on the foundations of 'communicative musicality' in the companionship between infants and adults. Evidenced in the rhythms and melodies of reciprocal communication, his work demonstrates the intrinsic nature of mutual musical expression and response, and their effects on emotional and cognitive growth. Through his unique perspective which honors the pervasive power of embodied moving and its relationship to physical and mental processes, Trevarthen has shown how shared participation in vocal and action games contributes to cultural understanding, meaning making, and a sense of belonging.
This page intentionally left blank
4
72
"How can good ideas survive in a market society and a post-disciplinary order? We want to learn and not to be taught... We want to meet without being touched... Still we should develop our personality through learning together... We have individual systems, knowledge buffets and general truth factories. Education is high on the political agenda, yet the first item to be transferred to the transit zone whenever possible. So let's envision education and learning embedded in a humanist and ecological society. And move it into a zone of urgency. In the labyrinth of reality. Routing the paths of knowledge and information. Jumping from the learning express to the red zone of fun and switching from the shopping malls of knowledge to the pains of focus and concentration." Marleen Wynants
A Learning Zone of One's Own M. Tokoro andL. Steels (Eds.) 1OS Press, 2004
'3
4
Pretend Play as Learning: Case Studies from the Home Marleen Wynants
Pretend play by children can give us important clues to how the human mind works. Since most of the pretend play occurs at home, parents can be privileged observers and interviewers even if their main role lies in providing the necessary context to enable imaginative games. As one of these privileged observers, for about eleven years now, the continuing pretend play of my daughters reveals that it is a crucial medium in the process of learning, language, memory and psychosocial behaviour. There's emotional grounding involved and it seems intrinsically motivated but that doesn't explain yet where it comes from. Thus evoking the following questions: Why do children engage in pretend play? What is the role of pretend play in social learning and personality development? In what ways can it be an optimal learning context and a wonderful learning tool? In the following pages we will try to answer these questions by zooming in on diverse examples, techniques, core elements and functions of pretend play.
And once she had really frightened her old nurse by shouting suddenly in her ear, "Nurse! Do let's pretend that I'm a hungry hyena, and you 're a bone!" - The Annotated Alice, Lewis Carroll (1998) Was Alice conscious of the effect her exclamation would have on her nurse? Probably!.. Lewis Carroll's favourite character is the queen of play and pretending. Not one single reader has doubts about the imagined world that Alice explores and most readers will jump happily into the rabbit hole with her or peek through the looking glass to zoom in on the odd behaviours of the characters of the explored world. Just like millions of young and older readers embarked happily on the Hogwarth Express on platform 9 3/4 in Harry Potter's London. Just like we all love to set our minds free and follow the stream of thoughts expressed by a character in an opera, a theatre play, a novel or a poem. And even if the
74
4. Pretend Play as Learning
characters or their actions are not real, the emotions always are. With the same ease, we launch ourselves into interior narratives or become completely absorbed by or embedded in something we do. For children this happens all the time and without effort whereas adults require some exercise in focus and concentration before they can step into the flow.
Ezra (9 months) filling the empty bottle with sand and drinking it. Roses, Catalunya 1992. What makes children engage in pretend play? What are the deep aspirations that motivate people to make believe, to start pretending? Young children don't pretend that they will do something. Instead, they believe something and they act accordingly.
1. Baby Let's Play House Pretend play is different from a state of mind like belief. Beliefs can lock us into a way of perceiving that filters reality and hence they can condition our actual experience of life, of reality. Pretend play should rather be seen as the acting out of temporary beliefs. Precisely because of this, pretend play is an often-used tactic in therapy since it offers a context without shame or guilt. The pretending mind is a Utopia for outraging acts and everyday rebellions. But long before we consciously engage in make believe to act out our emotions and beliefs, we use pretend play as a way to make sense of our body, our mind and the world we live in. Pretending is more than just a mere distracting characterisation of the mind. When a baby is pushing some colourful Duplo bricks around while saying "bbbrrrr..." the baby is not merely « playing car » but exploring the world and learning to represent it. The same happens when children run around in the supermarket, pausing to wait on an invisible friend or galloping happily towards their favourite displays. Children engaging in pretend playing do not merely set up a simulation of the real world because as we will see in the coming examples, the dynamic properties of their play are very different from the ones manifested in the world they represent. The nature and purpose of pretend play to train skills and to acquire and organise knowledge has long been studied by theorists, such as Darwin (1877), Piaget (1962), Vygotsky (1967), and Bruner (1976). Still it is only recently that pretend play has become a study object in cognitive science by researchers like Alan M. Leslie (1987), R.D. Kavanaugh (1998) and Lillard (2001).
4. Pretend Play as Learning
75
Infants develop their first representations of reality in a such way that the characteristics of objects and people mix with their own emotional projections on them. When they engage in pretend play, they create a place for the introspective and the expressive self. In this nonthreatening setting they nourish the motivation and involvement necessary for learning about ways of dealing with reality, with their peers, with their parents, with teachers, with emotions, with cognition. As I observe my children playing, the thing that always strikes me, is the concentration and the joy that accompanies their playing. The Chicago psychologist Mihaly Csikszentmihalyi (1990) studied the fulfilling experience of 'flow' that comes to children playing games, to scientific researchers engaged in experiments, to people in conversation, when walking the mountains or making love. These experiences all entail the same sense of happiness and joy through the loss of self-consciousness.
Ezra (4) and Dylan (5) playing chic in Paris, 1996.
The third issue addressed in this article is the role of pretend play as an optimal learning context and learning tool. When observing the various kinds of pretend play that my children engage in at home, it struck me that a major part of their play is about experimenting, about creating representations and organisation, about unlearning and breaking rules. Through pretend play my children nurse their curiosity and try things out. In that way, pretend play is a pretty pure scientific inquiry. But what do they learn ? Through the safe context of pretend play, children work on their motoric, perceptual, emotional, social and conceptual organisation. They also engage in problem solving, developing memory, language and communication skills like attention, expression and negotiation. They explore old and new experiences through mingling facts with fiction, through looking at reality from different perspectives. Last but not least, it looks like an optimal experience since there are a lot of physical, biological and evolutionary games involved at all levels. The inevitable question that rises while observing pretend play in the home environment is: What is the role of the parents? Especially at a time when more and more people start getting convinced that we will need creative minds to solve the problems of the world, developing the children's imagination might be the first step into that direction. As a mother or a father, one can be an observer, a planner and a model. Nurturing diversity is about the most important thing we can do and accepting the child's invitations to play and to leam
76
4. Pretend Play as Learning
with them is part of that. A home can be a private school, a library, a communications centre, a restaurant, a world kitchen, a laundry, a window on nature, with hotel facilities, a hospital, infirmary, meditation and relaxation centre, transport service, sports activities, music conservatory, repair shop, creative workshop... Parents can provide their children with different windows on the world, and teaching them everything you know yourself, is a good start. Next to the home base, the interaction with other children, learning through peers, interacting with real and imagined friends will prove to be of major importance in the development of the child's personality.
2. Pre-Pretend Play Very early on in life, babies 'notion of cause and effect is own action based. They touch a mobile, and something happens. They engage in the exchange of facial expressions with their mother, father or caretakers and they quickly starts making different sounds and eliciting different reactions from their environment. What follows, as early at one years of age, is a period characterised by the sheer pleasure of pointing, naming and categorising things. Later, when pretend play appears, the child learns to shift the meanings of things and to "do as if. A chair can become a horse or a train and happy or sad emotions will be expressed according to the imaginative game the children want to play. The ability to pretend depends on the capacity to represent absent objects and situations, a capacity said to emerge during the second year of life. But already long before that, other playful activities prepare the way: joint story-telling rituals, singing rhymes, taking turns in babbling, mothers and fathers acting silly and pretending, and one of the first collaborative games: Playing rough & tumble. Through playing rough & tumble, children learn to anticipate and develop motor skills. They also explore their emotions and those of the people they are most intimate with. They learn about communication skills like giving / sharing attention, turn-taking and guessing the reactions and intentions of others. Yet other functions of rough & tumble play are the construction of games, the development of narratives and the imaginative creativity involved. Last but not least, in playing rough & tumble, infants express their attachment to their beloved ones just like lovers do when they engage in the very particular rough & tumble play called making love. Most infants, like most of the baby animals, like to play rough & tumble. My children took up the role of an animal, like a dog or cat or lion to explore their immediate environment horizontally and vertically, and try out the affective reactions of their parents and family members they are most intimate with. They change roles according to the flux of the expressed emotions and the animal behaviours they associate with it. From pre-school until the age of 10 my daughters could be seen crawling under tables, stroking their heads against each other, crawling up next to me or their favourite uncle and purring their heads off until I asked them to become real again and snap out of their game. Most of the times the rough & tumble game is just a playful way for getting attention, physical comfort, or love At other times it was a rather provocative game or just another way of retreating into a world that was a lot more exciting than witnessing adults talk after dinner. Sometimes, while either Ezra or Dylan were again barking their beautiful heads off somewhere near me, I had to think of Bertha, the eldest daughter of farming neighbours of my grandparents. The story goes that before they went out in the fields, they would tie Bertha to the table next to the
4. Pretend Play as Learning
77
family dog. After a few weeks, the child barked as loud and convincingly as the dog and not the postman, nor my own grandmother could make the distinction between the girl and the dog's barking. Was the girl pretending? I doubt it very much. Bertha did start talking around the age of 6 when started going to school. Peek-a-Boo Peek-a-boo is a less formal, early form of hide-and-seek. It involves a whole gamma of emotions amongst which surprise, fear, mock-fear and joy, that are shared by the players and the bystanders. A comparable game is "playing sandwich", where parents roll their kids into a blanket and pretend they are making their favourite sandwich. Sometimes the parents act in silence, sometimes they talk aloud and explain what they are doing. Of course, the real fun part comes when the mother or father actually pretends to start eating the sandwich. With the predictable squeaks, laughter and comforting hugs afterwards. Again in this game the children and parents can explore their mutual attachment and a whole range of imaginative ideas and creative expressions to manifest this.
Dylan (I) ignoring my grandfather hiding himself. When he pops out, she will act surprised. Geel, 1992
In playing these games, parents never have to explicitly teach their children how to do it, but by doing it themselves and involving their children, children pick it up spontaneously. Imitating the behaviour of other people is one of the child's most powerful tools for learning the whole range of expressions, body language and verbal language. And humour. Yes! The cognitive phenomenon of understanding and producing humour comes in quite early. The earliest forms of humour emerge, when a certain behaviour doesn't fit the context, and when the children realise that this is done on purpose. One of the first things that made my baby daughters really laugh - and that was before they were one year old was when I put their sootier backwards in my mouth or one of their hats on my head. At these moments they couldn't refrain themselves from bursting out laughing. This was definitely the funniest thing they had ever seen.
78
4. Pretend Play as Learning
Babies and infants watch and hear you do something and later, some minutes, some hours, some days, some months later, when you are not around, they do the same. And not only the same, but also much more.
3. Imaginary roles for Animated things One of the strongest characteristics of young children, is maybe their spontaneous capacity to animate things. Some of the objects that surround us, are there to make us mobile, to protect us, to communicate, to prepare food and to eat it. But there are even much more objects that help us relax, sleep, dream, love and feel loved. Some of them are souvenirs or witnesses of particular moments in the past and just by looking at them or touching them, or smelling them, we set our memories going and create moods and inner visions. That's how we get attached to objects in an emotional way. In order to avoid getting upset and emotionally off-balanced by objects that get thrown out, stolen or lost, Sarat Maharaj, art historian and co-director of Documenta XI, told me: "To me things never get lost or stolen, they are simply not there anymore". It's a nice way of trying to cope with it, but if this will work for a child that has lost its favourite blanket or cherished stuffed animal, I don't know. The difference with young children injecting objects with emotional characteristics is that these objects originally are completely free of emotional memories, because the memory is still in full development. It's really a pure form of animating the inanimate and only later on, when they have been around quite a while, the objects get emotional connotations associated with memory.
TUMBLE AND DRY One day I picked up Dylan's birthday pet and by the look-and-feel of it, I decided it needed a bath. So I took the little stuffed tiger called Boo to the kitchen and was on the verge of putting him into the washing machine when Dylan (3) came near. Dylan: What are you doing with Boo? Me: He needs a bath! (Whereupon I put him in the machine) Dylan (almost crying): Don't! Please! He 'II die! Take him out! Me: Don't worry, he 'II come out fine and....clean! Dylan (still frightened): I don't care! I want Boo back! You 're gonna drown him! Me: Do you want to lick him clean? Dylan: (laughs) Um... No, but we could give him a bath. There was no escape. I took him out of the machine, we filled up a washing basin and gave Boo a nice bath with real shampoo, carefully avoiding to put soap in his eyes...
Because there are so many emotions projected into these objects, they have become far more than what they seem to represent at first sight. Stuffed animals are for a lot of children
4. Pretend Play as Learning
79
the main characters in their assimilation of the world, although most of the time they are transitional. The first detachment babies experience is from their mother. So they attach themselves to one or to some objects - real and imaginary -, to finally detach themselves from these objects too when they come more socially involved in close friendships. The animation of objects by children was closely studied at the psychosocial background of attachment figures and behaviours like "agoraphobia" (Antony Storr, 1987). Studies of the early emotional development of children show that when attachment figures are unreliable or absent, instead of developing increasing confidence, the child will start regarding the world as a frightening and unpredictable place into which it is not safe to venture alone without a supporting arm. Of course, there will always be people like Linus from Charles Schulz's Peanuts whose life seems to depend on a worn out blanket. To Linus his blanket is his proteases, his dance and play partner in rough-and-tumble plays, his protector, his trustful audience and a handy tool when drooling and crying is involved. No wonder it's hard to detach from an object like that because finding all these characteristics in one real person, will always turn out to be a disappointing experience no matter how hard we may try to fool ourselves into it. Children are not alone in this symbolic game. Adults too "fall in love" with objects that evoke strong or positive emotions or embody emotional memories. The fact that Coco Chanel (Tisseron 1999) started sewing copper buttons on a lot of her designs, was due to a former relationship she had with an officer. The skirts of Jean-Paul Gaulthier got their shape not from the design table but were transferred from his memory of the woman who took care of him during his childhood to his working table. What could have turned out into a personal fetishism got transferred into creativity, whereas on the other hand, clothes by Chanel and Gaulthier can become fetishist objects for their buyers. Some people talk about particular objects in their house as if they were family. For others, trees, a park, a house or a monument can have the same affective value as the one they attribute to people. A striking example of this was right after the attack on the Twin Towers in New York City in September 2001. People in Manhattan were reacting emotionally and in shock. "They are gone!" a friend of mine working in Hell's Kitchen exclaimed on the phone. "Marleen! What are we going to do without them?" Thus we keep injecting objects with affective and symbolic characteristics throughout our lives. It all starts when children of about two or three years old set out to explore the fact that other people have independent minds. The "Other" comes in, a crucial step in the development of the "Self. Animating objects is a way of tuning in [or attuning to] with others, with life, with nature. Children engaging in pretend play will transpose habits and objects to other places, times and psychological states and the lived experience is disconnected from its usual context. Moreover they can pretend about the identity of an object, of oneself, of another person, of a place, an event, a situation. (Flavell, Miller and Miller (1993) Objects playing a part in children's pretend plays, do change in function and in characteristics. "Let's pretend this is our house, okay?" The house referred to may be a line drawn in the sand, a wooden bench in the garden or a specific part of the carpet in the living room. A few minutes later, the game may change. One of the children may suggest something like: "The house now becomes a shop and the garage will be the bedroom of the shopkeeper, okay?" The other children are free to agree with this or not. If so, the game goes on, if not, parallel plays will probably develop. Most of the dolls and stuffed animals but also stones, bags and books gathered in our house
80
4. Pretend Play as Learning
throughout the sometimes got didn't like but shooting at the
years, sooner or later had a soul—some for a few minutes, others for years. I quite rude to some of the stuffed animals that I, for one or another reason, there was this particular case that changed my attitude and made me stop imagined world of my daughters with arrows poisoned with realism.
DON'T BE CRUEL There was this awful stuffed cat called Tinny, with little blue misshapen eyes and a realistic kinky imitative fur and a ragged tail that I thought was the most ugly thing I had ever seen. Ezra (7) saved it from drowning in a pool of other animals at her cousin's attic, the place where a lot of stuffed animals go when they die. Countless times I/irmly picked her up, threw her in the furthest corner or close to the nearest dustbin. Dylan and Ezra didn 't understand why I didn 't like her and came up with a successful strategy. One day I was in the bathroom ranging clothes and overheard a conversation that I was meant to hear: Ezra: Look, Tinny is sad. Dylan: Well Tinny what's the matter? Did some of the animals say ugly things to you again? Ezra (with a high pitched voice): No, I'm used to that. But I'm really afraid your mum is going to throw me away. Dylan: Why? Ezra (with a high pitched voice): Dylan: It's because she doesn 't know you.
Because
she
doesn't
like
me.
Ezra (with a high pitched voice): I know, but she doesn't want to get to know me. Every time I get close to her, she throws me with the Duplo 's or in the bag with stuff for the basement. Dylan: But you take care of the little animals! And remember the day that you made Ezra happy again after she had fallen down and hurt her knee? Ezra (with a high pitched voice): Well, maybe she doesn't know these things. Somebody should tell her. She sees my ragged fur and she wants to get rid of me. Dylan (serious): There is nothing we can do about that Ezra (serious): No. But we would all be very sad if you were gone Tinny. Ezra (with a high pitched voice): Thank you, but please protect me because I'm really afraid. Dylan: Don't worry Tinny, we will have a serious talk with her. Ezra: Can we play now?
\ lowered my head and silently crept down the stairs. Tinny is still around, and I still pick her up with two fingers from the ground. t now I put her down more carefully and I don't throw her around anymore because she's getting really old, her fur has become even
4. Pretend Play as Learning
81
looser and she looks more ragged than before, but she earned her respect and I learned my lesson. As parents we can provide the necessary context to make pretend play possible, and to enhance it. Whereas I did my share in the building of camps and tents and providing props, I rarely took part in the games that were played in the camps and tents. My children could do very well without me although sometimes they did invite me to come in and have something to eat - realising very quickly that imaginative food wasn't as attractive to me than real cookies were. Or to have me around for a while, just being there, and when they got what they wanted, they would tell me explicitly that I could leave the premises, so that they could go on with their "real play". As children grow older, their pretend play becomes like staged performances, in which the children act like directors, scene builders, providers of delicate and carefully created props and always multiple roles. BOA TSMAN 'S BIRTHDA Y Scene: Birthday party thrown for Ezra's stuffed bear, called Boatsman. Participants: Ezra, Dylan, their father- Jan, a huge brown bear called Boatsman, some twenty other stuffed animals and me. During the years Boatsman, being huge and flexible, has served as Ezra's pillow, her best pall in sad times, as an authority, comparable to the owl in stories like Winnie the Pooh or Bambi. And Boatsman also has a silly streak and likes being dressed as a girl, at least that's what he pretends... We gathered in Dylan's room where all the animals were sitting in a circle, Boatsman next to Ezra with a paper crown on his head. Jan: How old is Boatsman getting? Ezra: Ask him. He's old enough to tell you. Jan: How old are you Boatsman? Boatsman (through Ezra): 56years old! Me: I hadn 't thought you were already that old! Boatsman (through Ezra): You should know! It was you who brought me here, wasn 't it? Me: Right, I'm sorry Boatsman. Boatsman (through Ezra): Tell me the story again, will you? Me: Well, I went to the Ikea woods when Ezra was a few weeks old and there I saw you and asked you to come with me. There were a lot of brothers and sisters of you out there. Boatsman (through Ezra): I know, my family is really big. We keep in touch by letters. Well, where are the presents? Dylan (speaking for all the stuffed animals and doing different Here! Here! Mine first! Jan: I forgot to bring a present.
voices alternately):
82
4. Pretend Play as Learning
Piglet (a little pig talking through Dylan): That's OK! You can help unwrap mine! Jan: Thank you. Ezra had wrapped up tens of little toys into presents and we had to sit through the whole unwrapping scene with all the stuffed animals giving comments on what the others offered. Some of the more impatient guests were really offended that other guests would offer things like a broken little windmill or a long time neglected and unfinished drawing. At some point Dylan got angry because Ezra had wrapped up one of her favourite Playmobil characters as a present, thinking that way that it would get transferred to Ezra's room, which Ezra denied of course. For some of the animals, the major characters let's say, Ezra had even provided various handmade birthday cards and drawings. To everybody's surprise I came up with an oven warm home made birthday cake, which raised the fun factor and turned everything into a party. We sang, Ezra blew Boatsman's candles, we all had a piece of cake and we still talk about that party as we talk about some of our own birthday parties. The alternation between the real and the imaginative proved to be no problem although the adults involved did have to make an extra effort not to ruin the party with too much realistic small talk...
4. Building Physical and Mental Representations
Ezra (9) arranging the set for a bunch of miniature animals. Provence, 2002.
LET'S SKIP THE PREPARING OF FOOD, IT'S NO FUN Scene: Ezra (9) & Dylan (10) playing with Playmobil miniature dolls, tools and furniture. Ezra: My animals are home. They are eating. Dylan (through a little man called Herman): What did you prepare for them?
4. Pretend Play as Learning
83
Ezra (through a little nameless woman): Oh, there was some stuff in their stable. Shall we go to the disco? Dylan (through Herman): Not yet, don't you want to eat something first yourself? Ezra (through the woman): Great! (Moves the doll into Dylan' s dollhouse) I like your house. Dylan (through Herman): Thank you. I'll make some pasta with tomato sauce! Ezra (through the woman who is sitting at a little table): Mmmm, this tastes so good! Dylan: You haven't got anything yet! I have to cook the pasta and make the sauce. Ezra: No, let's skip preparing the food, it's boring. Let's eat and get to the disco, that is more fun. Dylan: But I want to prepare food. Ezra (through the woman): Good! I'll be back soon. I'll go and check my horses. Maybe I'll eat a pizza and go to the disco on my own. Dylan: That's not fair!
Fair to whom? To the little man wanting to prepare food? Or to herself wanting to play on in her little kitchen? I asked them about it.
What is it that you like in playing with these miniature dolls and environments? Dylan (10): "You can represent real life very easy through them. I love miniature representations of existing things. Like my doll's house. The kitchen in there is the kitchen I would love to have when I grow up. It's this old wooden like kitchen. It really existed and they reproduced it in small size ". Ezra (9): "Playmobil is really worked out in detail. That's so great about it. The dolls get names yes. Like Herman and Petunia. I called my horses Hermione and Ron. Predictable huh? But after all the names I had to come up with after all these years, I was a bit lazy, I didn 'tfeel like thinking about it for too long. " Dylan: "Ezra never plays according to the real world; she always has these crazy characters who cannot act normally. I play with them as if they 're part of the real world. Ezra will never cook with these dolls, she thinks it's a waste of time. Also the fact that these dolls eat, she thinks it's a waste of time really. I love doing these things. I spend minutes in making mashed potatoes, while she sends her dolls into these unbelievable adventures. I have this old ladylike doll and I give her an old fashioned mixer and then she makes this mashed potato stuff. " Ezra: The problem is it takes the time it would take to make real mashed potatoes. That's boring. Who decides what is going to be done next? Ezra: Most of the time the dolls talk to each other and develop the story. But of course we are always talking about the scenario out loud too. "Later on they will go into the woods... "
84
4. Pretend Play as Learning
Pink and pretty predictable?
Barbie and props. Provence, 2002
Barbie is another queen of pretend although her Wonderland is pink and pretty predictable. Children know that from the beginning but: Hey, girls do want to have fun and like to play around in a world that is easy, smiling and forever! In the beginning, the girls play with Barbie dolls as if they were hand puppets, getting them in a good grasp around the knees and using them to engage in dialogues about the pink props that come with them. But pretty soon they will start directing scenes out of the pink and although it's hard to do, most girls try to get Barbie out of the mindless trap she was born in. The picture above shows a Barbie kitchen made out of props and assembled things during our summer stay in Provence. Dylan and Ezra spend hours looking for small pebbles and stones, in discussing the functions they were going to have and the colours they would paint them in. I had provided them with three natural coloured ochre powders, one bright blue and one parrot green thus adding exotic flavours to the pink Mattel equipment. Asked for solid arguments for wanting to play with Barbies, here's what Dylan and Ezra came up with: Dylan: With Barbie you need more props because you use less imagination. Ezra: Barbie's diapers don't need to be changed. They come in handier than other dolls cause you don't have to feed or care for them all the time. Through them, you can talk normally. No babytalking, which can be fun, but after a while you want a normal conversation and with Barbies you can. Me: Do you give them different names? Dylan: Yes, they all have a different name. The names do change too. My favourite Barbie is named Kathleen. Ezra: Sometimes they keep the names that they have been given in the factory, like Esmeralda or Pocahontas. Sometimes we do forget these names and we come up
4. Pretend Play as Learning
85
with new ones. Me: What kind of scenes do you play with them? Kitchen? Living room? Beauty parlour? Dylan: Bed scenes. (Spills of laughter). I love the bed scenes! Also bar- and disco scenes. But I love the bed scenes although my male doll has always this kind of plastic drawn boxer shorts on, which is a bummer... The nice thing about them is that you can imagine a kind of life for them. Partly your own life or what we would like to do. Like with the Playmobil stuff. Our games alternate between fantasy and things that can happen in real life. But with the Barbies, it's mostly entertainment. It's always fun! Ezra: Yeah! Although they also go to the university. They didn't study something in particular, they just hung around and did some tests. One of Dylan's Barbies was super dumb! Dylan: That's not fair! She hadn 't practised for her tests, that's all! Ezra: It's so easy to be Barbie. You never have to proof or do anything. You just are. And you get away with it. Dylan: Playing with them will not last long anymore. Until we're in secondary school? Maybe not even that far. We move to the real world more and more, into music and stuff like that. Instead of "Have you seen the last outfit of my Barbie?" it will become: "Didyou hear the new CD ofManu Chau? "
During their role playing with dolls - - Dylan as Ezra both suggest possible evolutions and changes in the scenario. And if one does not agree with a certain move, the other most of the times does not adapt herself, but continues playing in parallel for a while until they get together again. Needless to say that the world of Harry Potter has influenced the ways they try to come up with eccentric skills, imagined languages and lots of surprising events in their pretend playing. The latest skill my daughters developed, is that they started documenting and photographing some of the set ups and scenes themselves. And building an archive and different stories.
Hermione and Harry posing for Ezra taking their picture. Brussels 2002.
86
4. Pretend Play as Learning
The soul of violins and bikes Even at the age of 11, my eldest daughter Dylan sometimes patted her bike whenever her one-year younger sister Ezra puts it away with too much clinging and banging. So instead of saying to Ezra "Hush, be careful" she talks to her bike as if she has to calm it down in order to prevent that it will attack Ezra, saying: "Don't worry, easy now, don't bite her, she's okay but a bit rough sometimes". It's a very diplomatic way of dealing with a problem, an indirect way of saying to her sister "Be careful with my bike, will you!" but without provoking a direct fight. Although it's a restricted socio-dramatic pretend play, the social and communications skills involved are pretty sophisticated. Another example is the way Dylan treats her violin. She considered it a literal playmate in the sense that she addressed it directly by asking "Ready?" just before she actually started playing. She also used to pat it each time before putting it back into its case. At a certain moment, the patting stopped and a few weeks later, I knew why when she asked for an electric guitar for her eleventh birthday. There was immediately some new patting involved, but I predict that from now on, the patting will decrease, as the decibels will increase... In the examples and observations above, we distinguish some common characteristics of pretend play. A first characteristic is the uses of the symbolic function. Pretend play is symbolic or representational in that it detaches behavioural routines and objects from their habitual, real-life situational and motivational contexts and using them in a displaced, makebelieve stage. Piaget has long established that young children animate things that move, like clouds or waves or water. Also humans relate differently to objects that they animate in their imagination - the result of personification - than to objects that they treat as merely reactive - objectification. So, personification could be an important stepping-stone for human cognitive and emotional growth. It starts with children endowing things with life, they blur the boundaries between animate and inanimate things. But also adults interact with imaginary characters through the fictional characters in books, operas, theatre plays, movies ...(Ackermann, 2000) Media that always need a listener to become alive. We project our convictions and emotions onto them and hence render life to the perceived letters, molecules and pixels. We insert words, characters, music, pictures and even people we meet and observe with our own personality. In that way, children, as well as adults, get a grip on or try to frame reality. A second characteristic is the socio-dramatic quality of pretend play Children are masters in directing and improvising and in setting up games free from externally applied rules. A third characteristic is that children are very actively engaged, there's no daydreaming or idling involved. And last but not least: Watching your children engage in pretend play, makes one thing clear: there's much attention and concentration involved [not true]. The mind is at work. And what is more, and what makes all the difference: the mind is at play! We should keep this in mind when we create learning environments for children. De Luca (1991) cites the scientist and storyteller Gianni Rodari saying: "by using stories and those fantastic methods that produce them, we help children to enter reality through the window instead of through the door. It's more fun. Therefore, it is more useful."
4. Pretend Play as Learning
87
5. Imaginary roles for animate things
Dylan (4) galloping happily on a beach -or maybe she was writing a songline? Mallorca, 1995.
Just like children - and adults - project emotional characteristics on the behaviour of the animals they interact with, children often engage in animal-like behaviours that they already injected with emotional interpretations. It is not uncommon to see children trot or gallop in a supermarket, as if they were horses. While they pretend and make believe, children get completely absorbed in their fantasy world,. What is important to note, here, is that for young children, there's usually no confusion about where reality stops and imagination begins. Dylan's favourite animal when she was younger, was a horse. She loved watching pictures of horses, touching and talking to real horses, sitting on her father's back and kicking her invisible spurs into his sides and whenever she could, she would start trotting or galloping. Until galloping became her way of running and she hesitated for two years before taking part in a sports event at school, being afraid she would start galloping during the 600 meters run. An error that would make her really ashamed of a specific behaviour she once liked so much. But Erzra and I convinced her and she took part in the competition. Without winning, but without galloping ! Let me zoom into another common form of pretend playing in which children engage spontaneously, namely personal role-playing. Not all types of pretend play facilitate the understanding of other people's mental states, but role-playing does. Hence it is a crucial element in the psychosocial development of children. Through it children become able to experiment with their own emotions and with those of others, without risk.
88
4. Pretend Play as Learning
I'M ONLY PLAYING Dylan: Lore, have you finished your sums? Ezra (with a slightly higher voice): Not yet, teach. Can I do them later? Dylan: No Lore, you will have to finish them first and then we are going to paint. Ezra (with a slightly higher voice): Yes! I love painting. Dylan: Yes, but your sums first. Ezra (still with a high voice): Sums are stupid. (Starts tapping with her pencil on the desk) Dylan (Getting annoyed): Stop that Lore or I'll have to punish you like I did with Tom. Ezra (in her own voice): Can we please start painting? Dylan: Not yet, until I say so. Ezra (repeats mockingly): Not yet, until I say so... Dylan (still a bit playful): Stop that! When you've finished your sums, you can copy them twice for me. Ezra (in her own voice): We haven't had these kinds of sums yet in school, I can't make them. Dylan (trying to keep up a nice voice): I will learn you. Ezra: I don't want to learn, I want to play. Dylan (in her own voice and angry now): Stop acting so silly! Ezra: I'm only playing! Dylan: ... Well, then I don't like your way of playing anymore. Sometimes it looks like it is great fun to play the angry teacher and the disobedient pupil, sometimes reality takes over the imaginative world and the pupil won't want to play the pupil anymore. Although Dylan would never give up her teaching role, she always affirmed that Ezra (Lore) is a great pupil, sometimes a bit stubborn, but a nice and eager learner altogether. Playing school at home requires a lot of patience from both the teacher and the pupil. And some intense pretend playing, since they are not alone with the two of them... The rest of the pupils are invisible although they all have desks and chairs, documents with their names on and their own range of sums. While Ezra keeps on filling in everything, from different viewpoints and characters, Dylan corrects all the assignments and invents new ones. Sometimes they take a break and work on subjects together or parallel. Sometimes it seemed that they were rehearsing things they had learned in school, but most of the times, the socio-dramatic aspect of role playing was much more important and their game looked like a voluntary lesson in discipline and even more, in breaking the rules. Most of the creativity went into they ways they thought up new assignments, new courses and the way they kept on personalising everything for every character - present or imagined. They always played female teachers, although some of the imagined pupils were boys and those were not necessarily troublemakers or noisy pupils. But both Dylan and Ezra confirmed to me that they didn't like playing boys. They did like playing with boys, but did not like
4. Pretend Play as Learning
89
pretending to be boys because, according to Ezra "They make too much noise and always want to show off, there's nothing funny or attractive about that". Dylan adds: "Boys are so different from girls that it is hard to imitate them without making them ridiculous, so we don't do it". There is one exception: in chat rooms neither of my daughters will hesitate to take up the role of a boy. It's part of the role playing in online games where names, identity and realistic personal data can take the form of anything you want. By doing so, again, they create a safe context for themselves and explore the world in which they move around, a virtual one, this time. Role-playing is fun, it's even more fun to fool others into your game and to get away with it as a perfect chameleon.
Dylan (11) and Ezra (9) chatting on \s_^H\'Jlu^}l<^bitLll!•
Brussels, 2002.
Exploring and Performing gender issues in pretend play is hard to observe because most of the time gender-oriented plays happen behind closed doors or out of sight. And since I have no young boys in my natural laboratory at home, I cannot elaborate on this. The times that boys were visiting our house, there was shared pretend playing going on but from the age of 5 on, the nature of the games girls and boys played, differed and I saw more parallel plays developing than shared ones. The same props and environment were used, but the plays were different. I remarked that in mixed companies, children would rather start playing formal games, with set up rules, than engaging in games where the rules would have to be invented. When it happened, either the boys or the girls would loose interest in each others' games and rather do their own thing. There would hardly be any discussion or dispute about this, and sooner or later they would get together again for something else. As came up in several studies collected by Deborah Cameron (1998), boys and girls seem to learn different things with words in conversation and in pretend play. At first view, the differences are striking Girls and boys talk differently while pretend playing since they
90
4. Pretend Play as Learning
engage in different plays. It looks as if women and men carry over to their adulthood conversational patterns that they learned from interacting with their same-sex peers during childhood. And yes, the differences between these patterns create the well-known conflicts and misunderstandings in mix conversations. Aki Uchida (1998) states that gender problems are primarily caused by cross-cultural miscommunications. Hence the categories of females and males should not be seen as pre-linguistic variables biologically assigned to individuals at birth, Cheris Kramarae (1990) rightly states, because it prevents the conceptualisation of gender as a social construct. We are doing gender in the way we are dressed and addressed and hence it would be interesting to see how children deal with playing the "other" gender in pretend play. In the examples I discuss here , girls always performed female roles, except in playing animal. They never performed male roles when I was around. When I asked them about it while preparing on this article, they said they did pretend to be boys sometimes, in the classroom situation, but they would rather not elaborate further on it since it was a boring issue.
6. Invisible Playmates One last aspect of pretend play I wish to discuss briefly, is the emergence of invisible playmates or imaginary companions. Imagining friends is just another way to develop social contacts without harming anyone. The invisible animals and persons can be called upon when needed and dismissed when no longer needed. Although at a one point there were a lot of invisible animals around - birds, tigers, little pet dogs, even bees and a cat called Mayonaise - my children rarely talked to invisible boy- or girlfriends. Probably because they are very close in age, there's a strong common grounding and as a consequence they can engage one another quickly into games they both liked. The mechanisms involved in interacting with invisible playmates, is the notion of addressing what we think to our fictional audiences or ourselves. And although the playmates do not exist and the conversational partner is not there, the involved mechanisms of the mind are the same as in communication with real physical beings. Scene: I went to pick the girls up after they had stayed for some days with their father. Dylan, Ezra and I cross the hallway and are out on the street. Ezra (pointing with her finger to the doorstep and mumbling in a bossy way) Me: Whom are you talking to? Ezra: To Tyrano. Me: Tyrano who? Dylan: God! Sometimes you 're so blind! Me: Did I step on an animal that I can't see again? Ezra: It's Tyrano Saurus. He stays here in front of the apartment building. Me: Why? Ezra: To protect Jan. Me: Do you think Jan needs protection? Dylan: Only when we stay with you. There are three of us, and Jan is on his own.
4. Pretend Play as Learning
91
Me: Why a dino? Ezra: She's female! And a darling. But she's merciless with strangers. And Jan knows that she's here. It's okay for him as long as she doesn't scare people in the street. Normally she's visible but when people come near, she makes herself invisible. Invisible animals or imagined animals sooner or later will be an unavoidable part of a household with kids. Some will not last long. Others will disappear for a while and then come back. These invisible playmates can take any shape or form. Your child's friend could be a character from a book, a television show or a movie. My experience is that the age, until which children will play with their invisible animals, will largely depend on the permission and the patience parents have with these uninvited guests. Some are sleeping companions, while others are travelling companions. What is the function of this kind of imagined interaction? On the cognitive level, we can state that while communicating with imaginative friends or with an absent beloved one, children and adults explore communicative, social and psychological skills by going through arguments, motivations or memories. On the emotional level, I could observe that Dylan and Ezra were often looking for extra fun and excitement, and sometimes for additional reassurance and comfort in the imagining of non-existing friends. Children use their imaginary playmates to share secrets with, to communicate concerns or worries, to place responsibilities, beliefs and desires. The mental mechanisms involved might the same as those at work in teenagers idolising movie stars and imagining conversations with them. But again, rather than just another kid's silly game, this kind of imagined interaction has to be seen as highly sophisticated learning process. Maybe they are rehearsing language and communication skills through it maybe it is just one step away from the true interior monologue. Maybe it's just another way of dealing with reality? Or another means of solving a problem or just reflecting upon something, or to make meaning out of life? Probably it's all of these things together.
7. It's not the Tea but the Ceremony Quite often in playing and in learning, it's not the tea but the ceremony that matters. At a certain age, the spontaneous learning passion seems to get obstructed. It is not only a challenge for institutional education to restore this, but it is the responsibility of parents to provide children with an environment in which they can safely explore the world—and in which errors are allowed. Pretend play at home provides this context because through pretend play, children can fail or play trial and error as much as they want. They play the hero, the saviour, but they also eagerly play the culprit, the angry and the mean. In moving from imagination to the real world and back, children explore and digest their "real" life experiences, build the representations necessary to re-engage them in more "mediated" and reflective ways. This in turn fosters a child's cognitive and personal development Many psychologists and therapists see pretend play as acting out. Some philosophers prefer to regard pretend play as the individual in need of a dialogue or trying to deal with solitude. Pretend play hardly ever is limited to simple images, perception or feelings but comprises imagined conversations or arguments, reruns of events with new endings, selfjustifications, complicated plans for the future or difficult decisions that have to be made. In
92
4. Pretend Play as Learning
that sense I would like to state that pretend play plays a fundamental role in the development of the personality and «Self» and in processes like learning, memory and language. In other terms, it's a crucial part of how we humans become who they are. And don't forget: we keep pretending and taking up different roles throughout our whole life, it's not just a child's thing. Sometimes, when playing with a puppet theatre, children prefer a public of real people, at other times stuffed animals will do but most of the time they don't need a public. Why not? Because children who engage in pretend playing are not staging any pre-mediated play but they are staging their own life. "I'm just playing" is a phrase often used to defend a specific behaviour, and not only with children: It becomes "I was only joking" later on in life. That same non-threatening setting is an often-used tactic in second language learning and in therapy. But while some of the pretend plays are true landscapes of desire, others are another step in constructing reality. Whether we are talking about a drawing in which the pencil hardly ever leaves the paper, or about a child building a camp with accidental props, what seems to count is the process and of course the narrative that accompanies these external representations.
First drawings of Ezra (2): This is a party with grandma and grandpa and everybody is dancing and having fun.
According to the Belgian biologist, researcher and educator, Jean-Ovide Decroly (Dubreucq 1993), children develop representations to satisfy their need to identify the data of the environment. But also to appropriate and master them by play and opposition and to produce them by drawings and constructions, to transform them in dreams and fantasy and to evoke them mentally. The more the environment stimulates the activities, the more these representations develop the motor, sensorial, perceptive, affective, intellectual and expressive capacities of the child. Decroly showed through his research and experiments in learning environments that there is a common ground in all the modes of expression, the core
4. Pretend Play as Learning
93
of it being a very intimate experience that proves to be a major motivation in their learning process.
Acknowledgements This article exists thanks to the creativity and playfulness that my daughters Dylan and Ezra manifest spontaneously every day. With special thanks to Edith Ackermann for reading and improving this text and to Karin Beyens for some additional pictures.
References Ackermann, E. (2000). Relating to things that think. In "Play of ideas and ideas of play. 13 Magazine. The European Network for Intelligent Information Interfaces. P. 2-4 Bruner J.S. (1976). From comunication to language - A psychological perspective. Cognition, 3, 255-287. Cameron Deborah (ed) (1998). The Feminist Critique of Language: A Reader. 2nd Edition. London & New York: Routledge. Carroll Lewis (1998) Through the Looking-Glass, in the Annotated Alice, p. 180, New York : Wings Books. Csikszentmihalyi, Mihaly (1990). Flow: The Psychology of Optimal Experience. New York, Harper & Row. Darwin Charles (1877) A Biographical Sketch of an Infant, in Mind, 2, 285-294. De Luca, C. (1991) Gianni Rodari. La Gula sapienza della fantasia. Cantazaro, Abramo. Dubreucq, Francine (1993) Jean-Ovide Decroly - Perspectives: revue trimestrielle d'education comparee vol. XXIII, n°l-2, 1993, p.251-276. Paris, UNESCO:Bureau international d'education. Flavell, J.H., Miller, P.M. and Miller, S.A. (1993). Cognitive development (3rd ed.). p.82, Englewood Cliffs, NJ: Prentice Hall. Kramarae, Cheris. (1990) "Changing the Complexion of Gender in Language Research." In Howard Giles and W. Peter Robinson, (eds.). Handbook of the Social Psychology of Language. New York: Wiley & Sons. Kavanaugh, R. D., and Engel, S. (1998). The development of pretense and narrative in early childhood. In O. N. Saracho and B. Spodek (Eds), Multiple perspectives on play in early childhood education (pp. 80-99). Albany, N.Y.: SUNY Press. Leslie Alan M. (1987). Pretense and Representation: The origins of « Theory of Mind ». Psychological Review, 94, 412-426. Lillard, A.S. (2001). Pretend play as Twin Earth. Developmental Review, 21, 495-531. Piaget Jean (1962) Play, dreams and imitation in childhood. London : Routledge & Kegan Paul. Storr, Antony (1987), Why Psychoanalysis Is Not a Science. Colin Blakemore and Susan Greenfield (eds) In: MindWaves, p.80. Blackwell. Tisseron Serge (1999) , Comment L'esprit vient aux objects, p.41-42, Aubier.
94
4. Pretend Play as Learning
Uchida, Aki (1998). When 'Difference' is 'Dominance' : A Critique of the 'Anti-Power-Based' Cultural Approach to Sex Differences. In Deborah Cameron (Ed) The Feminist Critique of Language : A Reader, 2nd edition, p 280-292. London&NewYork : Routledge. Vygotsky L.S. (1967) Play and its role in the mental development Psychology, 5, 6-18.
of the child. Soviet
MARLEEN WYNANTS graduated in linguistics and literature and holds a masters in audiovisual communication sciences (KULeuven). She started working as content producer for the official Belgian Radio and Television in its pre-commercial stage and writing as a free lance journalist on art and music. From 1985 to 1988 she was the editor of the post-punt magazine Fabiola, leaving the scene the year that Hillel Slovak, Chet Baker, Divine, Sylvester and Roy Orbison died. The birth of her two daughters and the breakthrough of the Internet reoriented her focus and writing towards learning processes, the emergence of creativity and the relevance of gender. Some titles: The first Computers were Women, The Deadlock of the Body (on Alan Turing), Dea Ex Machina, The Child and the Machine, Solving the Problems of the World with Love and Mathematics. She conducts workshops on creativity and technology in primary schools and writes books for children. At present she writes for the transdisciplinary magazine JANUS and is the operational director of CROSSTALKS, the university and industry network of the VUB - Vrije Universiteit Brussel.
Part II. Optimal Experience and Emotion
This page intentionally left blank
5
98
"Many individual differences have their roots in biology, there's no way around it. Some people tend to develop a specific competence in some area, others develop other competencies in other areas. So we can never find a single common experience through which everybody will get interested in mathematics. Although most students will develop some interest, it will not necessarily develop into a pervasive flow experience. I had a wonderful mathematics teacher and I thank her, because what I learned was through her engagement although I never ended up in liking mathematics. Can flow be taught? Perhaps, but in any case, we can help people find out activities or areas of interests where they can use all their resources and engagement." Antonella Delle Fave
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
99
5.
A Feeling of Well-Being In Learning and Teaching Antonella Delle Fave
In the last two decades, researchers from a broad range of disciplines have become interested in the study of well-being and in the formalisation of models of healthy behaviour centred on human strengths and resources. This approach has potentially direct consequences on intervention, in that it emphasises empowerment and prevention instead of problem treatment. Training programs addressed to practitioners in the domains of health, education, psychology, and social policies increasingly take this perspective into account. As a consequence, researchers are facing the challenge of defining well-being, through the identification of its foundational components, both at the individual and at the social levels. The issue has been traditionally investigated in term of objective indicators, such as income, health and housing conditions. However, there is evidence that economic indicators do not provide an adequate evaluation of the developmental resources and goal attainment of a person, or of the success of a nation. As several studies show, it is also important to identify and measure subjective indicators of well-being, referring to individuals' judgements about their own state, satisfaction with life, social relationships, work and health, future goals and personal achievements (Diener, 2000; Veenhoven, 2002). Two approaches have been developed in the research on well-being: hedonism and eudaimonism. The former emphasises the search for pleasure, the latter the pursuit of self-actualisation, considered as the fulfilment of one's true nature (Ryan and Deci, 2001). Within the eudaimonic perspective, this paper aims at describing two specific constructs: psychological selection (Csikszentmihalyi and Massimini, 1985) and optimal experience, or flow (Csikszentmihalyi, 1975), that can be used to design and evaluate training programs in the domain of learning and education. Emphasis will be put on the importance for psychologists and educators to gather information on persons' subjective perception of daily activities and opportunities for action and development. The subjective perspective is essential to design intervention programmes centred on individual resources, and not on social expectations, and to offer people appropriate solutions reflecting their own experience and perceived needs.
100
5. A Feeling of Weil-Being in Learning and Teaching
\. Psychological selection Individuals do not live in a vacuum: they inherit a genotype from their parents, and they build their culture type by acquiring cultural information from their environment (Richerson and Boyd, 1978). Presently, most researchers interested in the study of human behaviour emphasise the joint influence of biology and culture. Socio-biology and evolutionary psychology focus on the biological roots of individual behaviour (Wilson, 1975; Barkow, Cosmides, and Tooby, 1992). Bio-cultural theories stress the interplay of biology and culture, as two different inheritance systems (Boyd and Richerson, 1985). In particular, the features of culture as a dynamic system have been investigated in the last two decades. Culture emerged as a peculiar human product, thanks to species-specific biological traits that promoted the transformation of the natural environment and the ability of humans to observe and describe themselves and the world. Cultural information units are stored in the human CNS (intra-somatic culture), as well as in material and symbolic artefacts (extrasomatic culture, Cloak, 1975). Thanks to culture, and more specifically to artefacts, humans could settle in various ecological niches, and survival became more and more dependent on cultural learning (Tomasello, Kruger, and Ratner, 1993). A selection process can be recognised in culture, which determines the differential survival and reproduction of cultural information. The selective transmission of cultural information units across generations shapes human communities and their changes in time (Durham, 1991). However, each cultural information originally was an answer to a specific problem or question, created and tested by one person or a small group of individuals. This has enormous implications for the potential role of individuals in the building of society. Being both reproducer and transmitter of biological and cultural information units, each human being actively influences the survival and replication of both biological and cultural pools (Delle Fave and Massimini, 2002). This influence is again based on a selective process, the psychological selection of biocultural information (Csikszentmihalyi and Massimini, 1985). The process is shaped by two specific human features: the subjective-objective awareness and the limited amount of attention resources (Csikszentmihalyi, 1978). Individuals cannot pay attention to all occurring environmental stimuli. Thus, they select a subset of information - daily activities, situations, and social contexts - to be involved in. As several studies have pointed out, the main factor directing psychological selection is the quality of experience.
2. Optimal experience Several studies have shown that psychological selection is guided by the quality of experience subjectively perceived in the interaction with the environment. Most part of daily life consists of apparently irrelevant episodes and experiences, but psychological selection operates on them, and its long-term outcomes are built on this micro-level information (Massimini and Delle Fave, 2000). Individuals differentially reproduce a subset of the available information units - activities, interests, relationships, values, behavioural norms - through their selective allocation of attention. More specifically, they preferentially invest their attention in environmental opportunities associated with positive and rewarding states of consciousness.
5. A Feeling of Weil-Being in Learning and Teaching
101
Particularly relevant for this approach has been the identification of optimal experience, or flow (Csikszentmihalyi, 1975; Csikszentmihalyi and Csikszentmihalyi, 1988). Optimal experience is characterised by the perception of high environmental challenges, adequate personal skills, high levels of concentration, enjoyment and engagement, loss of selfconsciousness, control of the situation, focused attention, positive feedback, clear ideas about the aims of the activity, and intrinsic motivation (Deci and Ryan, 1985). Optimal experience is a complex and highly structured state of consciousness, and it is preferentially associated with highly structured tasks and with activities supporting autonomy and creativity (Csikszentmihalyi, Rathunde and Whalen, 1993) As clearly shown in several studies, optimal experience is not centred on fun or positive mood: people do not emphasise the emotional component, when they describe this experience. Rather, they emphasise their involvement with relevant challenges requiring active engagement, and the satisfaction derived from improving personal skills. This has implications in the long-term perspective. The values of engagement in the various activities can be indices of complexity and predictors of future goals and activity cultivation (Greene and Miller, 1996). In particular, even compulsory activities, such as work and studying, can be perceived as sources of flow thanks to the engagement and concentration they require, and can therefore influence vocational choices and life goals (Delle Fave and Bassi, 2000). Similarly, the most positive experiences in leisure are reported in activities such as sports, games, arts and hobbies, that merge the fun with concentration, involvement and goal setting (Verma and Larson, 2003). Stebbins (1997) defined them as "serious leisure", in contrast with "casual" or "non-serious leisure". These activities provide enjoyment and intrinsic motivation, at the same time promoting intentional effort toward specific skill acquisition. Their crucial aspect is structure, that is a clear set of rules and procedures which foster agency, engagement, and autonomous action toward meeting challenges and pursuing goals. Cross-cultural studies showed the recurrence of optimal experience regardless of gender, age, culture of the participants and its association with the most varied activities, provided that they are valid opportunities for action, engagement and high skill investment (Massimini and Delle Fave, 2000).
3. Fostering personal development Optimal experience promotes individual development in that, to replicate it, a person will look for increasingly complex challenges in the associated activities and will improve his or her skills accordingly. This process, defined as cultivation, fosters growth of complexity in the individual behaviour as a whole. In the long run, psychological selection results in the life theme (Csikszentmihalyi and Beattie, 1979), defined as a set of activities, social relations, and life goals uniquely cultivated and pursued by each individual. In this perspective, optimal experience represents the basic unit of psychological selection. Like biology and culture, psychological selection operates as an evolution mechanism: In time, it promotes individual differentiation and growing complexity (i.e., internal order and
102
5. A Feeling of Weil-Being in Learning and Teaching
integration). Given the interdependence of individuals and culture, the growth of complexity involves constructive information exchange with the environment. This process supports development, that is the harmonisation of individual life theme with environmental opportunities for action. The extreme flexibility that characterises humans at the biological level has also been detected at the psychological level, in studies concerning persons who had to revise their life theme and to find new opportunities for optimal experiences after a trauma or disease causing disability (Delle Fave and Massimini, 2003). The key role of optimal experience in fostering personal growth and social integration were also detected in the field of mental health and social maladjustment. Cross-cultural applications of these concepts in psychotherapy and rehabilitation have produced encouraging results (Delle Fave and Massimini, 1992,2000). Several studies have shown that the perception of intrinsic motivation and autonomy plays a key role in fostering individual well-being (Ryan and Deci, 2000). This topic has been widely investigated, and the findings highlighted the cross-cultural consistency of the association between autonomy and well-being, suggesting the ability to generalise of this construct (Chirkov and Ryan, 2001). Following intrinsic motivation, a key component of optimal experience, individuals can creatively devote themselves to the cultivation of specific activities, discovering innovative goals and thus spreading new information. With the support of optimal experience, the human ability to create and to select opportunities for action that enhance psychic complexity can be considered the basis of cultural change.
4. Assessing optimal experience The role of optimal experience is twofold: It supports the preferential replication of the associated activities within one's own life span, at the same time promoting the survival and spreading of such activities, or any other kind of information, within the culture. Several research procedures have been developed to investigate the daily fluctuations of subjective experience and the occurrence of flow. Among them, the most widely used are an experience sampling procedure and retrospective questionnaires. Experience Sampling Method (ESM) was developed by Csikszentmihalyi, Larson and Prescott (1977). It investigates contextual and experiential aspects of daily life through online repeated self-reports that participants fill out during the real unfolding of daily events and situations. ESM therefore allows researchers to investigate subjective experience overcoming the problems related to the retrospective collection of information, such as distortions and rationalisations (Larson and Delespaul, 1992). During a standard ESM session, participants carry for one week an electronic device sending random signals 6-8 times a day from 8 a.m. to 10 p.m.. They are asked to fill out a form at each signal reception. Questionnaires comprise open-ended as well as scaled questions. The former investigate the external context at signal receipt (activity, location and companionship), the content of thought, as well as the desired activities, places and social interactions, if any. The quality of experience perceived at signal receipt is assessed through Likert-type scales measuring the level of affective, cognitive, motivational variables. Additional information is gathered about participants' evaluations of the level of personal satisfaction, short- and
5. A Feeling of Well-Being in Learning and Teaching
103
long-term importance of the activity, opportunities for action perceived in the situation (challenges) and personal abilities (skills) in facing such opportunities. Since self-reports are repeatedly filled out, numeric values of each variable are transformed into z-scores before analysis, starting from the individual statistics for each participants. Thanks to the features of the ESM data, a model of analysis has been developed (Experience Fluctuation Model, EFM), that allows researchers to study the relation between the perception of challenges and skills, and the quality of experience (Massimini and Carli, 1988). The model, used in a broad range of studies, is built on the Cartesian plane and is divided into eight sectors called channels, in which channels 1, 3, 5 and 7 are centred upon the two main axes, starting from 90° and proceeding clockwise; the others are positioned on the bisectors of the four right angles (Figure 1). Each channel corresponds to a particular ratio between the standardised values of challenges on the y-axis and those of skills on the x-axis. EFM allowed us to identify a relation between the values of challenges and skills and the quality of experience. Through the application of the model, a recurrent association emerged between specific experiential patterns and channels. In particular, when challenges and skills are perceived above mean (Channel 2), optimal experience is reported. When challenges are above average and skills below it (Channel 8), participants describe a state of Anxiety. The perception of challenges below and skills above average (Channel 4) corresponds to a state of Relaxation. Finally, when the values of challenges and skills are perceived below average (Channel 6) a state of Apathy is reported. Subsequent ESM data analysis can focus on time-budget (daily distribution of activities, locations, social contexts) and on the fluctuations of experience according to activities and social interactions. Data can also be used to draw inter-and intra-group comparisons, as well as cross-cultural
Channel 1
Channel 2 OPTIMAL EXPERIENCE
Channel?
Channel 3
Channel-! RELAXATION
Channel 5
SKILLS
investigations. Figure I. The Experience Fluctuation Model (SM = Subjective mean). Flow Questionnaire (Csikszentmihalyi, 1975) is designed to investigate the occurrence of optimal experience and its psychological features. In its most widely used version (Delle
104
5. A Feeling of Well-Being in Learning and Teaching
Fave and Massimini, 1991), participants are first asked to read the following three quotations that describe optimal experience: "My mind isn't wandering. I am totally involved in what I am doing and I am not thinking of anything else. My body feels good... the world seems to be cut off from me... I am less aware of myself and my problems". "My concentration is like breathing... I never think of it.. I am quite oblivious to my surroundings after I really get doing in this activity ...When I start, I really do shut out the world. Once I stop 1 can let it back again". "I am so involved in what I am doing... I don't see myself as separate from what I am doing". Participants are then invited to report whether they have ever had similar experiences in their life, and —if yes— to list the activities or situations associated with it (also defined optimal activities). No limitations are given as concerns the number of activities to be quoted. Subsequently, participants are asked to select from their list the activity that is associated with the most intense and pervasive optimal experiences, and to describe the experience reported during this activity through 0-8 point scales investigating cognitive, affective and motivational variables: involvement, clear-cut feedback from the activity, intrinsic motivation, enjoyment, perception of clear goals, challenges, skills, focus of attention, distraction, boredom, anxiety. All the answers to the open-ended questions are coded and included in broader functional categories, derived from previous studies. To evaluate the quality of experience during the selected optimal activities, mean scores are calculated for each variable. The individual and environmental conditions which contribute to the onset and maintenance of optimal experience are also investigated (Massimini, Csikszentmihalyi and Delle Fave, 1988). The average quality of experience in the main daily activities, such as studying, work, family interactions, being alone are investigated as well. Data allow researchers to specifically identify the psychological and phenomenological components of optimal experience, to evaluate the daily opportunities for action participants perceive in their environment, the quality of experience associated with daily routine activities, and the quality of daily life from the subjective perspective. We usually administer Flow Questionnare together with two other instruments. The first one, Life Theme Questionnare, provides information on participants' positive and negative life influences, present challenges, future life goals, and on the specific role played by school and family as the main socialisation factors in their lives. The other one, Order/Disorder Questionnaire, investigates situations and activities associated with the perception of internal and/or environmental order and disorder, and the daily strategies used to avoid excessive disorder or to create a situation of dynamic and flexible order. The three questionnaires together provide information on participants' quality of life, perceived determinants of individual history, goals and expectations, skill cultivation through selective allocation of psychic resources on optimal activities, and personal growth trajectories.
5. A Feeling of Well-Being in Learning and Teaching
105
5. Optimal experience and learning The association of optimal experience with culturally meaningful activities, such as learning, is an important premise for successfully coping with the challenges of the productive world. In general, studying is the major daily task for children, adolescents and college students. In cross-cultural investigations, studying covers a remarkable percentage of time in the daily life of teenagers (Brown, Larson and Saraswathi, 2002). Italian adolescents describe studying as the most engaging daily activity; however, intrinsic motivation and mood are lower than in most other situations (Delle Fave and Bassi, 2000). When studying at home, students are often alone and wish to be doing something else: they report high engagement and average control of the situation, but values of mood and intrinsic motivation significantly below the average. However, researchers have shown that solitude can be a very fruitful opportunity for creative behaviour, personal growth and identity building (Larson, 1997). When attending classes, students live in an institutionalised setting, where they primarily have to deal with classmates and teachers, following adult rules of behaviour and interaction. Usually, the school setting is a highly regulated context in which little room for free interaction is available. Our findings with a sample of Italian high school students showed that school interactions were associated with high engagement, on the one side, and with low control and intrinsic motivation, on the other side (Delle Fave, Bassi and Massimini, 2002). This can be related to the pressure toward achievement and the external evaluation both by teachers and by peers. These findings highlight the need for improving students' intrinsic motivation, in order to make school and learning experience more rewarding, and not only engaging. Several studies have shown that teenagers' rejection of their schools' values is related to the undermining of intrinsic motivation. Children who have more fully internalised the regulation for positive school-related behaviours are those who feel securely connected to and cared for by their teachers (Ryan, Stiller, and Lynch, 1994). School should also aim at teaching youth - in particular, adolescents - how to find meaningful challenges for development and identity formation in everyday life. Findings show that even compulsory learning activities can offer opportunities for optimal experience, provided that the school context is adequately supportive of autonomy and creative expression. In a ESM study with Italian adolescents, learning activities accounted for the 36.6% of the situations associated with optimal experiences (Delle Fave and Bassi, 2000). Thanks to the engagement, concentration and skill development provided by learning activities, study can be preferentially selected and cultivated as a component of personal Life Theme, thus fostering personal growth and development (Csikszentmihalyi, 1982). Since learning activities can structurally become sources of enjoyment and intrinsic reward, adolescents should be equipped to perceive this potential, and to identify opportunities for engagement and personal growth in their work. As clearly stated in the previous pages, flow does not mean fun. It is a complex and structured state of consciousness, made of cognitive, motivational and emotional components, and each of these components plays a specific role in shaping the experience. Flow helps individuals actively build their own life goals and identity. A school system based on passive learning and on the transmission of ready-made
106
5. A Feeling of Well-Being in Learning and Teaching
easy notions can lead students to apathy, lack of involvement in school subjects, and search for alternative, sometimes deviant sources of optimal experience. On the contrary, students should be encouraged to use their abilities and to find engaging and rewarding opportunities for action, during both work in class and homework (Ranson and Martin, 1996).
6. Optimal experience and teaching Teachers represent behavioural models that can be crucial for the students' subsequent career choices (Flink, Boggiano, and Barrett, 1990). The quality of experience they report at work influences the educational process, and — in a broader perspective — the effectiveness of cultural transmission and students' desire for learning (Lepper and Cordova, 1992; Richer and Vallerand, 1995). Csikszentmihalyi (1982) pointed out that optimal experience at work is one of the prerequisites to be an effective teacher. Moreover, students get more easily involved in school subject matters and improve their performances when teachers communicate them the potential for autonomy and self-determination in learning The affective climate perceived in the classroom fosters students' involvement and active participation in classes, promoting autonomy and relatedness while discouraging rough competitiveness (Ryan and Powelson, 1991). The association of teaching with optimal experiences was confirmed in our own findings, derived from the administration of Flow Questionnaire and Life Theme Questionnaire to 80 primary and secondary school teachers (Delle Fave and Massimini, in press). Both teaching in classroom, and individual work (preparing lessons, writing notes, evaluating students' tests) were quoted among the opportunities for optimal experience. Teachers also reported to perceive knowledge acquisition and exchange as part of their life theme, a means for pursuing personal development and growth in complexity at the psychological level. Besides teaching, they reported to preferentially devote their attention to books, artistic hobbies, and study. When asked about the reasons for becoming a teacher, they indicated the prominence of intrinsic motivation in choosing an activity which allowed them to cultivate intellectual sources of enjoyment, and to transmit interest in knowledge to their young students. On the contrary, the most negative experiences were associated with children's disengagement and lack of interest: this finding is consistent with the results reported by other researchers (Burke, Greenglass, and Schwarzer, 1996). The quality of children's learning widely depends on the quality of teachers' work, but teaching effectiveness, in its turn, is related to the global quality of experience educators associate with their job. This should be taken into account in educational programs, as well as in studies concerning occupational psychology and job satisfaction (Evans, 1997). People enjoying their work can represent models for their children, friends, or students. People who find intrinsic rewards and opportunities for self-expression in a specific domain —be it art, science, sport, manual work— can bring these more or less overt opportunities to the attention of others. Providing teachers with opportunities to improve their skills, their subjective perception of work and their role in society as well can have an enormous impact in shaping young generations, and in the development of culture (Alderman, 1999). In this perspective, the quality of attention educators devote to their work will eventually influence cultural
5. A Feeling of Well-Being in Learning and Teaching
107
selection. This is especially true in that - as discussed above - teaching does not only consists in the transmission of cultural information. It also provides youth with behavioural models, values and social skills. Teachers play a key role in promoting commitment to lifelong learning, co-operation in group work and engagement in society. . The active involvement of youth in the learning process, their perception of learning activities as opportunities for optimal experiences, and their awareness of the cultural impact of their individual behaviour will make them more responsible citizens.
References Alderman, M.K.. (1999). Motivation for achievement: possibilities for teaching and learning. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Barkow, J.H., Cosmides, L., & Tooby, J, (1992). The Adapted Mind: Evolutionary Psychology and the Generation of Culture. New York: Oxford University Press. Boyd, R., & Richerson, P.J. (1985). Culture and the Evolutionary Process. Chicago: Chicago University Press. Brown, B., Larson, R.W., & Saraswathi, T.S. (Eds.) (2002). The worlds' youth: adolescence in eight regions of the globe. Cambridge: Cambridge University Press. Burke, R., Greenglass, E.R., & Schwarzer, R. (1996). Predicting teacher burnout over time: effects of work stress, social support, and self-doubts on burnout and its consequences. Anxiety, Stress and Coping: an International Journal, 9, 261-275. Chirkov,V.I., & Ryan, R.M. (2001). Control versus autonomy support in Russia and the US: effects on well-being and academic motivation'. Journal of Cross-Cultural Psychology, 32, 618-635. Cloak, F.T. (1975). Is a cultural ethology possible? Human Ecology, 3, 161-182. Csikszentmihalyi, M. (1975). Beyond Boredom and Anxiety. San Francisco: Jossey Bass. Csikszentmihalyi, M. (1978). Attention and the holistic approach to behavior. In Pope K.S. and Singer J.L. (Eds.), The Stream of Consciousness, 335-358, New York: Plenum. Csikszentmihalyi, M. (1982). Intrinsic motivation and effective teaching: a Flow analysis. In D. Bess (Ed.), Motivating professors to teach effectively (pp. 15-26). San Francisco: Jossey Bass. Csikszentmihalyi, M., & Beattie, O. (1979). Life themes: A theoretical and empirical exploration of their origins and effects. Journal of Humanistic Psychology, 19, 45-63. Csikszentmihalyi, M., & Csikszentmihalyi, I. (Eds.) (1988), Optimal Experience. Psychological studies of Flow in Consciousness. New York: Cambridge University Press. Csikszentmihalyi, Larson, & Prescott (1977). The ecology of adolescent activity and experience. Journal of Youth and Adolescence, 6, 281-294. Csikszentmihalyi, M., & Massimini, F. (1985). On the psychological selection of bio-cultural information. New Ideas in Psychology, 3, 115-138.
108
5. A Feeling of Weil-Being in Learning and Teaching
Csikszentmihalyi, M., Rathunde, K. & Whalen, S. (1993). Talented Teenagers. New York: Cambridge University Press. Deci, E.L., & Ryan, R.M. (1985). Intrinsic motivation and self-determination in human behaviour. New York: Plenum Press. Delle Fave, A., & Bassi, M. (2000). The quality of experience in adolescents' daily life: developmental perspectives. Genetic, Social and General Psychology Monographs, 126, 347-367. Delle Fave, A., & Bassi, M. (2003). Italian adolescents and leisure: The role of engagement and optimal experience. In S. Verma & R. Larson (Eds.), Examining Adolescent Leisure Time Across Cultures: Developmental Opportunities and Risks. New Directions in Child and Adolescent Development Series. San Francisco: Jossey Bass. Delle Fave, A., Bassi, M., & Massimini, F. (2002). Quality of experience and daily social context of Italian adolescents. In A.L. Comunian, U.P Gielen (Eds.). It's all about relationships, 159-172. Lengerich: Pabst Science Publishers. Delle Fave A., & Massimini F. (1991). Modernization and the quality of daily experience in a Southern Italy village. In N. Bleichrodt, & P.J.D. Drenth (Eds.), Contemporary Issues in Cross-Cultural Psychology (pp. 110-119). Amsterdam: Swets & Zeitlinger B.V. Delle Fave, A., & Massimini, F. (1992). Experience Sampling Method and the Measuring of Clinical Change: a Case of Anxiety Syndrome. In M.W. deVries (Ed.) The Experience of Psychopathology, 280-289. Cambridge: Cambridge University Press. Delle Fave A., & Massimini F. (2000). Living At Home Or In Institution: Adolescents' Optimal Experience And Life Theme. Paideia. Cadernos de Psicologia e Educacao, 19, 55-66. Delle Fave, A. & Massimini, F. (2002). Cultural change and human behaviour: evolution or development? In A. Delle Fave and Min B. Pun (Eds.), In pursuit of a sustainable modernisation: culture and policies in Nepal, 13-32,. Milano: Arcipelago Edizioni. Delle Fave, & A., Massimini, F. (2003) Making disability into a resource. The Psychologist, 16, 9-10. Delle Fave, A., Massimini, F. (in press). Optimal experience in work and leisure among teachers and physicians: individual and bio-cultural implications. Leisure Studies. Diener, E. (2000). Subjective well-being. American Psychologist 55, 34-43. Durham, W.H. (1991). Coevolution. Genes, culture and human diversity. Stanford, CA: Stanford University Press. Evans, L. (1997). Understandig teacher morale and job satisfaction. Teaching and Teacher Education, 13, 831-845. Flink, C., Boggiano, A.K., & Barrett M. (1990). Controlling teaching strategies: undermining children's self-determination and performance. Journal of Personality and Social Psychology, 59, 916-924. Flinn, M.V. (1997). Culture and the evolution of social learning. Evolution and Human Behavior, 18, 23-67.
5. A Feeling of Well-Being in Learning and Teaching
109
Greene, B.A. & Miller, R.B. (1996). Influences on achievement: Goals, perceived ability, and cognitive engagement. Contemporary Educational Psychology, 21, 181-192. Larson, R. (1997). The emergence of solitude as a constructive domain of experience in early adolescence. Child Development, 68, 80-93. Larson, R.W. (2000). Toward a psychology of positive youth development. American Psychologist, 55, 170-183. Larson, R., & Delespaul, P.A.E.G. (1992). Analysing experience sampling data: A guidebook for the perplexed. In M. deVries (Ed.), The experience of psychopathology - Investigating mental disorders in their natural settings (pp.58-78). Cambridge: Cambridge University Press. Lepper, M.R., & Cordova, D.I. (1992). A desire to be taught: instructional consequences of intrinsic motivation. Motivation and Emotion, 16, 187-208. Massimini F., Csikszentmihalyi M., & Delle Fave A. (1988). Flow and Biocultural Evolution. In M. Csikszentmihalyi, & I. Csikszentmihalyi (Eds.), Optimal experience. Psychological Studies of Flow in consciousness, 60-81. New York: Cambridge University Press. Massimini, F., & Carli, M. (1988). The systematic assessment of flow in daily experience. In M. Csikszentmihalyi, & I. Csikszentmihalyi, (eds.) Optimal Experience. Psychological studies of Flow in Consciousness, 266-287. New York: Cambridge Univ. Press. Massimini, F., & Delle Fave, A. (2000). Individual development in a bio-cultural perspective. American Psychologist, 55, 24-33. Massimini, F., Inghilleri, P., & Delle Fave, A. (Eds.) (1996). La selezione psicologica umana. Milano: Cooperativa Libraria IULM. Ranson, S., Martin, J. (1996). Towards a theory of learning. British Journal of Educational Studies, 44, 9-26. Richer, S.F., & Vallerand, R.J. (1995). Supervisors' interactional styles and subordinates' intrinsic and extrinsic motivation. The Journal of Social Psychology, 135, 707-722. Richerson, P.J., & Boyd, R. (1978). A dual inheritance model of human evolutionary process: basic postulates and a simple model I. Journal of Social and Biological Structures, 1, 127-154. Ryan, R.M., & Deci, E.L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist. 55, 68-78. Ryan, R.M., & Deci, E.L. (2001). On happiness and human potentials: a review of research on hedonic and eudaimonic well-being. Annual Review of Psychology, 52, 141-166. Ryan, R.M., & Powelson, C.L. (1991). Autonomy and relatedness as fundamental to motivation and education. Motivation and Emotion, 60,49-66. Ryan, R, Stiller, J., & Lynch, J. (1994). Representations of relationships to teachers, paretns, and friends as predictors of academic motivation and self-esteem. Journal of early adolescence, 14, 226-249. Stebbins, R. A. (1997). Serious leisure and well-being, hi J.T. Haworth (Ed.) Work, leisure, and well-being (pp.117-130). London: Routledge.
110
5. A Feeling of Well-Being in Learning and Teaching
Tomasello, M., Kruger, A.C., & Ratner, H.H. (1993). Cultural learning. Behavioral and Brain Sciences, 16, 495-552. Veenhoven, R. (2002). Why social policy needs subjective indicators? Social Indicators Research, 58, 33-45. Verma, S. & Larson, R.W. (Eds.), Examining Adolescent Leisure Time Across Cultures: Developmental Opportunities and Risks. New Directions in Child and Adolescent Development Series. San Francisco: Jossey Bass. Wilson, E.O. (1975). Sociobiology.: the new synthesis. Harvard Mass.: Bellknap.
ANTONELLA DELLE FAVE, M.D., specialised in Psychology, is professor of General Psychology at the Medical School of the University of Milano, Italy. She conducted several cross-cultural research studies in Europe, Asia, Africa and America, gathering data on the quality of daily experience and on its long-term developmental impact by means of singleadministration inventories and experience sampling procedures. Her studies aim at investigating the active and creative interaction of individuals with their natural and cultural context, through their daily psychological selection of the environmental information. This process has been investigated in over 4500 participants belonging to different cultures and subcultures, and widely varying in age, health conditions and lifestyle. She is presently involved in international co-operation projects on disability, social maladjustment and quality of life. Delle Fave has been specifically studying the relation between optimal experience at work and at school and bio-cultural selection.
6
112
"What fascinates me is this initial moment when your attention is triggered, when you discover something that totally absorbs your interest. I would call this an initiatic state and according to me this is the key state. It is a necessary condition to start the educational process. But how to reach this state? How to use it? Once you have triggered attention, you have won the case. Maybe we can develop technologies to help in this, but that is not my primary focus. My own challenge is not only to create a flow machine, but to create initiatic experiments. I am interested to find out how we can design technology that helps create these initiatic experiments."
Francois Packet
A Learning Zone of One's Own M. Tokoro andL. Steels (Eds.) IOS Press, 2004
113
6.
On the Design of a Musical Flow Machine Francois Packet
This paper addresses the issue of designing interactive systems that create flow experiences in users. I first describe an interactive musical system called the Continuator, which is able to learn the musical style of users in an agnostic, continuous fashion. I then describe experiments conducted with professional musicians and with 3 to 5-year old children and the Continuator. I show that these interactions are - almost - typical of the Flow phenomenon, as introduced by Csikszentmihalyi. I then focus on the abstraction of the design principles behind the Continuator and propose the notion of Reflective Interactive System as a class of applications which trigger Flow experiences. Based on the analysis of the various psychological experiments conducted so far, 1 identify the issue of flexibility in interaction protocols as a crucial step to enhance the efficiency of Reflective Systems.
1. When are interactions interesting? Virtually all things done by a computer program are interactive today. From web sites to word processors, from video games to entertainment robots, users are constantly engaged in various forms of dialogs with computer programs. In most cases, these dialogs are designed to help users solve precise and well defined tasks. For instance, information retrieval systems are designed to allow users to find quickly information they look for. Many web sites are - at least in theory - often designed to minimize the number of clicks needed to find specific information. Video games propose interactive devices designed so that users can quickly issue commands to move, shoot, or perform various actions in real time, without having to think about their input devices. In all these cases, the design principles put forward consist in satisfying precise criteria: utility, optimisation, conciseness or transparency. Of course there are some exceptions to the utilitarian view of interactivity: artistic installations in the domain of digital art (as
114
6. The Design of a Musical Flow Machine
found in shows such as Ars Electronica) consist often in providing the user with some sort of hopefully novel aesthetic or sensory experience, which does not necessarily correspond to a task or game to play with precise rules to follow. However, this lack of aim in purely artistic experiments makes these installations often hard to understand. Actually, the very idea of an aesthetic experience as such, decoupled from any relation to content is itself debatable. Whatever the subject matter, all interactions at work in man-machine interfaces do not create equally enjoyable experiences for users. This very issue of what makes an interactive system appealing, enjoyable or attracting to users is systematically eluded. It is also the subject matter of this chapter: what makes an interaction interesting ? This bold question is addressed under the form of a particular experiment in music interaction, called the Continuator project, whose goal is precisely to engage users in exciting, appealing interactions. We first introduce the project and the basic mechanisms of the system, and then describe some experiments involving children. Then we show the relations that these experiments may have with the theory of flow, as introduced by Csikszentmihalyi (1990). In the last section we generalize from the experiments with the Continuator and propose the notion of Reflective Interactive System as a class of applications having the same properties than the Continuator. We propose that Interactive Reflective Systems are a possible way to build flow machines. We conclude on proposing extensions and other examples of this class of system and issues remaining to be solved.
2. The Continuator The Continuator project stems from major frustrations of the author regarding music interactive systems. The issue of interacting with a computer to make music has long been addressed by many researchers, with many different goals in mind. Today there seems to be two categories of music making systems. On the first hand, purely interactive systems in which users may trigger various kinds of musical effects. These systems have been experimented for a long time by the pioneers of computer music. For instance, Jean-Claude Risset composed pieces with a Yamaha Disklavier (a piano forte with a Midi input and output) in which the computer played various types of accompaniment based on his input (Risset & Van Duyne, 1996). More recently, the Korg company issued a synthesizer of a new kind, called the Karma, which proposes thousands of such musical effects (Kay, 2000). This notion of musical effect is particularly well adapted to describe the type of interaction at play: when a user presses a key, a chord, or some sort of predefined musical sequence, the system reacts by producing, in turn, a sequence based on the user input. The versatility of the machine makes it possible many effects ranging from arpeggiators to automatic accompaniments in many styles. On the other hand, a lot of research has been devoted to understanding the notion of musical style, and building systems that produce music in particular styles. On of the first attempt in building an automatic composer is probably Mozart with his Musikalisches Wurfelspiel in 1787 (Chuang, 1995). This composition took the form of a dice game. Each
6. The Design of a Musical Flow Machine
115
throw would determine a number. This number, associated with the number of the current measure would determine the new measure, out of a collections of pre-programmed measures composed by Mozart himself. From a table of 176 minuet measures, the device could generate, thanks to two 6-sided dices, 1.3 1029 different minuets. The device invented by Mozart was in fact a rudimentary implementation of so-called Markov models, invented much later, but using the same principle: transition tables containing probabilities about transitions between two events. Applying Markov models to analyze and generate music has become, since Mozart, a tradition in itself. Today, many systems have been proposed to analyze automatically corpuses of music material, under their score or Midi form, and to generate new material based on these analysis. The most spectacular of these attempts was probably done by Cope (Cope, 1996) with his EMI system. David Cope's system is able to generate convincing musical pieces in the style of virtually all classical and modern composers, ranging from Jean-Sebastian Bach to Scott Joplin. However, to achieve a reasonable level of accuracy and faithfulness to the style being mimicked, Cope has to enter by hand many information to the system, in particular concerning high-level musical constructs such as forms, beginning, endings, etc. which are not naturally captured by Markov models. However, whatever their level of accuracy music generation system are not interactive: the processes of analysis and generations are separated, with possibly many hand made parameter tweaking inside. Unfortunately, purely interactive systems are never intelligent, and conversely, intelligent music generators are never interactive. The Continuator is an attempt to combine both worlds: interactivity and intelligent generation of music material in a single environment. The Continuator System The Continuator system consists of one MIDI input (typically from a synthesizer) and one MIDI output (typically returning to the same synthesizer). Its operation in the standard mode involves no interface other than the MIDI instrument itself. The user plays musical sequences of any kind, either monophonic, polyphonic, in any playing style. When the phrase is terminated, the Continuator generates a musical phrase in response. This musical phrase has the characteristic of being stylistically similar to the phrases played by the user so far. Technically, it is a continuation of the last input phrase, hence the name of the system. Although it is difficult to define the notion of musical style precisely, we have adopted, like our predecessors in music generation research, a notion of style consisting of the statistical distribution of notes, chords and musical elements in general as well as their ordering. The Continuator, like post of music style replication systems, is based on a Markov model of musical phrases, and the model of the style created by the system retains melodic patterns, harmonic progressions, dynamics and rhythmic patterns of the corpus used for learning. An important consequence of this approach is that the phrases generated by the Continuator are similar but different from the phrases played by the user. The Continuator may therefore be seen as an engine for producing variations of arbitrary musical material. Some Musical Examples To illustrate the working of the Continuator, simple musical examples are given below (see Figure 1 and Figure 2). These examples are noted exactly as they are played, i.e. without rhythmic quantization. They show that the Continuator adapts quickly to arbitrary styles
116
6. The Design of a Musical Flow Machine
and is able to generate musical material that "sounds like" the user input on a relatively small scale. Issues related to capturing higher-level structure are not discussed here as they are not relevant for our purpose (refer to Pachet 2002a for more details). More sophisticated examples of music created by the Continuator can be found on the web site of the author. The most important aspect of the Continuator is the fact that the musical material generated always conforms stylistically to the input. Also, the Continuator keeps on learning from whatever input is given. As a consequence, the behavior of the system improves over time: if the user produces phrases which are stylistically consistent, but unique, the Continuator will learn more faithfully and will produce musical phrases that are increasingly accurate, with respect to the musical style of the user.
Figure 1. A simple melody (top staff) is continued by the Continuator in the same style (bottom staff).
Figure 2. A simple chord sequence (top staff) is continued by the Continuator in the same style (bottom staff)
Implementation and Design There has been considerable research done in the fields of artificial intelligence and information theory regarding the technical issue of learning a musical style automatically in an agnostic manner. Shannon introduced the concept of information based on the probability of occurrence of events in communications (messages) in his seminal 1948 paper (Shannon, 1948). This notion was soon after used to model musical styles, for instance by (Brooks et al., 1957). These early experiments showed that it was possible to create pieces of music that would sound like given styles by simply computing and exploiting probabilities of note transitions. More precisely, given a corpus of musical material (typically musical scores or MIDI files), the basic idea was to analyze this corpus to compute transition probabilities between successive notes. New music can then be
6. The Design of a Musical Flow Machine
117
produced by generating notes using these inferred probability distributions. One of the most spectacular applications of Markov chains for the generation of music is probably (Cope, 1996), although his musical results are not entirely produced automatically. A good survey of state-of-the-art Markov-based techniques for music can be found in (Trivino-Rodriguez et al. 2001), including in particular variable-length Markov models, which capture stylistic information more finely. The Continuator system is yet another species in the world of musical Markov systems, although with novel features. In our context, we wanted to learn and imitate musical styles in a faithful and efficient manner and make the resulting mechanism useable as an actual musical instrument. This raised a number of technical issues, whose solutions were progressively integrated in the Continuator. The architecture of the Continuator consists of two modules: an analysis module and a generator module. The analysis module takes as input MIDI sequences played in real time. The system contains three main parts: 1) A phrase end detector, which is able to detect that a musical phrase had "ended". This detection is based on an adaptive temporal threshold mechanism. The threshold is inferred from the analysis of inter onsets intervals in the input sequence. As a result, if the input sequence is slow (or, rather, contains few notes per seconds) then the threshold is increased, otherwise it is decreased. This simple mechanism ensures that the continuation produced will be seamless, temporally. 2) A pattern analyzer. Once detected as completed, these input sequences are sent to a pattern analyzer, which builds up a Markovian model of the sequence. The complete algorithm is described in (Pachet, 2002b), and consists of a left to right parsing of the sequence to build a tree of all possible continuations for all possible prefixes of the sequence. To speed up learning, the system also learns all transpositions of the sequence. 3) A global property analyzer. Various global properties of the input sequence are also analyzed, such as: the density (i.e. number of notes per second), the tempo and the meter (location of strong / weak beats), the overall dynamics (i.e. loud or soft), etc. These properties are used to produce a continuation which is musically seamless with the input. The generator is responsible for producing the continuation of the input sequence. The actual production of the musical material exploits the Markovian graph created by the analysis module (Pachet, 2002b). It essentially consists of producing the continuation on a note-by-note basis. Each note is generated using the Markovian probabilities inferred during the analysis stage. Technically it uses a variable-order Markov generation that optimizes the relevance of each single note continuation by looking for the longest possible subsequence in the graph. Special care has been taken to perform meaningful segmentations of the input phrases for the learning phase. Indeed, real-world input phrases are never composed of perfectly successive notes or chords. In order to "cut" input phrases into chunks, which are then fed to the learning system, a segmentation process is able to detect note or chord transitions and possibly cut across unfinished notes. The module also stores the possible "residual" discrepancy, and restores it at generation phase, so that the material retains the rhythmical "naturalness" of the original style. This continuation sequence is, however, crude, in the sense that it does not necessarily have the global musical properties of the input sequence. Therefore, a mapping mechanism is
118
6. The Design of a Musical Flow Machine
applied to transform the brute continuation into a musical phrase that will be played just in time to produce seamlessness. Currently, the properties which are analyzed and mapped are tempo, metrical position, and dynamics. More details can be found in (Pachet, 2002a). Playing Modes The basic playing mode of the Continuator is a particular kind of turn-taking between the user and the system determined by three principles: 1) Automatic detection of phrase endings. The Continuator detects phrase endings by using a (dynamic) temporal threshold (typically about 400 milliseconds). When a time lapse exceeds this threshold, the Continuator takes the lead, and produces a musical phrase. 2) The duration of the phrase generated by the Continuator is parameterized, but in most cases the duration is set to be the same as the duration of the last input phrase. 3) Priority given to user. If the user decides to play a phrase while the Continuator is still playing, then the system will stop and return to listening mode (and eventually apply again principle 1). These principles, in the current implementation, are hard-coded in the system. Moreover, they are set without explicitly telling the users. Experiments with the system has shown that these rules are usually easily learned by the user in an implicit way - the behavior of the system is usually obvious, even for children.
3. Experiments with the Continuator Experiments with Professional musicians Apart from the technical issues related to system design and the evaluation of the "stylistic similarity" of the music produced by the system, it soon appeared that there was an important dimension of the project that was more difficult to describe in a standard scientific paper, related to the subjective impressions of users playing with the system, or watching it in action. Indeed, one impressive result of the experiments with professional musicians is that the very use of the system provokes intense subjective impressions and reactions. Rarely but occasionally, these reactions are negative. For instance a composer from Ircam reacted aggressively against the system as soon as he understood the machine would create (compose) music, and would not even try it. His credo was that machines should only be used to do things humans could not do. Since the Continuator only generated music from human inputs, he therefore considered the system uninteresting and did not need further inspection. Another one considered a priori that the system was not interesting because it was not able to capture stylistic information when the user plays only one note: indeed, the Continuator analyses transitions between notes or chords, and a one-note sequence does not
6. The Design of a Musical Flow Machine
119
contain any transition. The remark was true, but the composer refused to forgive this limitation. Apart from these two negative and ideological reactions, all the other reactions were very positive. The Continuator captures the attention of the audience beyond the traditional "demo effect" of many computer music presentations. In particular, a systematic Aha effect was noticed (Pachet 2002b) for professional users as well as beginners. Aha effects have been introduced in experimental psychology to characterize sudden moments of realization, understanding or inspiration. They are also being studied in neuroscience, where evidence of neuronal activity corresponding to Aha effects has been shown (Mogi, 2003). Experimentation with the system invariably induces users to reflect on their own musical personality (Pachet, 2002c). Bernard Lubat, a Jazz pianist and drummer, at the forefront of progressive Jazz in Europe played with the system many times. He evoked with great precision (in particular during performances at Uzeste festival and during a concert at Ircam in October 2002, see Figure 3) how the system would speed up his own evolution in improvisation, allowing him virtually to "play ahead of his current thinking". Gyorgy Kurtag Jr., a composer and improviser, described the Continuator as a kind of "amplifying mirror", and his regular use of the system throughout the year 2001 changed his way of improvising and composing music (this collaboration resulted in a composition performed at the Vienna Festwochen 2002 music festival, "The Hollow of the Deep Sea Wave »).
Figure 3. Bernard Lubat during his performance. At some point, he "launches" the Continuator with a characteristic gesture. Later, he makes gestures in a "pretend play" mode while listening to the music produced by the Continuator.
120
6. The Design of a Musical Flow Machine
Experiments with Children When the first author started playing with his daughter and the Continuator (she was 3, see Figure 4), her positive reaction was quite surprising and significant. She started to become interested in playing the keyboard - she would laugh at the system's "answers" and was able to focus her attention longer on musical playing.
Figure 4. A 3 year-old child playing with the Continuator at home.
The idea to push the experiment further with additional children was therefore quite natural. In most schools, music is still taught using outdated methods, in which children are confronted with formalisms before they have experienced the enjoyment of playing or listening to music. With a system like the Continuator, basic playing capabilities might be learned more easily, and earlier than with standard music education practice (piano lessons usually start at the age of 6 at the earliest n most conservatories). Most importantly, the Continuator - or in a general a class of systems able to learn and react - could develop a genuine desire for music in children, and consequently prepare them for traditional classical training in a more productive way. This need to bring more fun and interactivity in the classroom has long been advocated by various psychological studies (Webster, 2002; Delalande, 1984). More precisely, our vision of education as a pleasing experience falls within the boundaries of studies concerning flow (Csikszentmihalyi, 1990), as discussed in the last section of this paper. Preliminary experiments with children took place at a French kindergarten (Ecole Bossuet Notre-Dame in Paris). Later, more systematic experiments, involving in particular crossed experimentations, were conducted under the direction of Anna-Rita Addessi (University of Bologna) to investigate the issues brought to light in the first set of experiments further. A detailed analysis of these sessions from the perspective of experimental psychology is under way (Addessi & Pachet, 2004). Only the key observations are mentioned here. The goal of these experiments was to test basic hypotheses about the effect of the Continuator on the playing abilities of 3 to 4 year-old children. More precisely, the following hypotheses were made: 1 - Increase in attention span: Children can play longer with the Continuator than just a keyboard alone.
6. The Design of a Musical Flow Machine
121
2 -Aha and surprise: The Continuator produces noticeable Aha effects with children, as is the case with professional musicians (note that children rarely find things surprising, a flying pianoforte would probably not surprise them any more...). 3 - Autonomy: The Continuator motivates children to play music. For instance, a child, knowing that the Continuator is there, may express the intention of playing alone more often than with standard musical instruments. 4 - Exploration and playing modes: The Continuator pushes children to explore new playing modes. The results of the experiments with a normal piano suggested that children would usually stick to single playing modes, including playing with only one finger, playing clusters with two hands, playing ascending or descending diatonic scales (i.e. white keys), etc. 5 - The Continuator can develop various kinds of attachment behaviors in children, similar to what has been observed with the Tamagotchi or Aibo (Kaplan, 2001). These experiments consisted of several sessions. Each child (3 to 5-year-old) was invited to play with a keyboard (a Korg Karma with piano sounds and no additional effect connected to a pair of amplified loudspeakers). The Continuator was set to a mode that played phrases in the same style as the child, and of approximately the same number of notes as the input phrase. The threshold for triggering the continuation was set to about 500 rns. The child was left alone with experimenters in a familiar classroom. The protocol consisted in two phases: The child would first be told to play with the keyboard as he/she wanted, with no particular instruction. When the child stopped playing or express significant boredom, he/she would be told that the system would now try to play with him/her. At that point, the Continuator would be turned on. The session stopped when the child stopped playing. Several sessions involved also two or more children. Children were brought in one at a time by their teacher. The sessions were recorded with a video camera. A certain number of interesting points were observed:
-
Aha effects were indeed produced and noticeable. Reactions of the children ranged from enhanced attention to surprise when the Continuator started to play initially. The Continuator did seem to augment the attention span of most of the children. On two occasions, I had the session had to be terminated because the duration exceeded 40 minutes. There were two exceptions: one child was very tired and played very little, with or without the Continuator. Another child seemed to enjoy playing the keyboard enormously with or without the Continuator. The children who engaged in long interactions (one of them played 30 minutes and had to be stopped) also appeared to develop the ability to listen with great attention. The interaction mode of the Continuator (stopping when the child plays, and playing when the child stops) induced "turn-taking" behavior from the children, without explicit directions. The emergence of turn taking in such a context is not a trivial phenomenon, as shown for instance by the works on complex dynamic systems by Ikegami (lizuka and Ikegami, 2002). Some children exhibited a wide variety of playing modes. Apart from the classical playing modes (e.g. playing isolated notes, chords, arpeggios) the 1 children invented new playing modes such as playing with the sleeves, kissing the keyboard, playing from behind, playing with the palm, etc. Some of them also theatrically accompanied the launching of the Continuator with particular gestures like raising hands at the end of a musical phrase (see Section 4.5.3).
122
6. The Design of a Musical Flow Machine
The Continuator as a Flow machine The Continuator appeared in fine as a machine that promotes musical enjoyment in various forms. The system's ability to maintain children's attention for long periods of time — remarkable for this age group — and in general its ability to attract and hold the attention of users of all ages can be interpreted through the theory of flow introduced by psychologist Mihaly Csikszentmihalyi (Csikszentmihalyi, 1990). Csikszentmihalyi's notion of flow describes the so-called optimal experience as a situation in which people obtain an ideal balance between skills and challenges. Two emotional states of mind are particularly stressed in this theory: anxiety, obtained when skills are clearly below the level needed for the challenge, and boredom, when the challenges are too easy for the skill level. In the middle lies flow. Other states can also be described in terms of balance between skills and challenges (see ). We can think of the Continuator as a flow machine in the sense that it produces by definition a response corresponding to the skill level of the user. This approach also allows for the progressive scaffolding of complexity in the interaction, which is not the case for most pedagogical tools designed with a fixed pedagogical goal in mind.
Figure 5. Csikszentmihalyi's Flow diagram describes various emotional states such as boredom or anxiety according to the relation between skills and challenges for a given activity (From Delle Fave, this volume)
More precisely, Csikszentmihalyi describes the state of flow as consisting of several fundamental traits where the balance between challenges and skills is probably the most important. The other traits can be discussed in light of the experiments conducted: Focused attention. The experiments show clearly that children are engaged in focused activity both when both playing and listening. The ability to listen and concentrate for several seconds and listen to music is remarkable at this age. As pointed out by Carla
6. The Design of a Musical Flow Machine
123
Rinaldi (Rinaldi, 2003), listening is probably one of the most important abilities of children to discover the world around them. In her view, children are researchers constantly making up theories about the world and evaluating them. With the Continuator, we observed this phenomenon rather systematically, i.e. children engaging in deep, concentrated listening of the effect of their playing on the system. Figure 7 shows several examples of this phenomenon. Another typical situation encountered in sessions involving two children was the phenomenon of joint attention. More precisely, one of the children would force the other to stop playing to listen in order to the system. This situation, which we call "aspetta" (the Italian word for "wait"), is illustrated in Figure 8. Of course this behavior was not observed with professional musicians, who so far experimented with the system alone. Ease of concentration. This is particularly clear given the fact that no instruction is given to the children whatsoever. They play with the system in a self-motivated way, without any external constraints. Clear-cut feedback. The Continuator produces clear feedback (in fact this is the only thing it does). The interaction in some sense is reduced to the analysis of the feedback produced by the machine. Control of the situation. Children are in control of the situation most of the time. They understand quickly that they can interrupt the system whenever they want. The limitations in control are due to the difficulties that may arise when interpreting some of the system's outputs (see example in the next section). Intrinsic motivation. The most striking result of the experiments is related to the intrinsic motivation of the children, who were not told anything about the rules of the system. Excitement. Excitement is clearly shown most of the time in particular in the early phases of the sessions. We separate here excitement from surprise in the sense that the surprise effect is most often short in duration, whereas the excitement phase lasted much longer, sometimes for 20 minutes or more. Excitement was observed in most of the cases. Interestingly, the children were excited mostly by what the system was playing, rather than by what they were doing. Figure 6 shows some expressions of this excitement. Change in the perception of time and speed. A systematic study concerning interaction times is under way, but it is clear already that at least for some of the children time did pass very quickly: some sessions had to be terminated by the experimenters when the time limit was reached. There is, however, one flow characteristic that does not apply directly to the Continuator experiments: Clear goals. No goal was given explicitly to the children except to play until they were bored. Indeed, improvisation is generally not goal-oriented. Similarly, web sailing is usually non goal-oriented but many claim that it can nevertheless create flow (King, 2003). More importantly, It can be argued that children did create spontaneously goals during their interaction. For instance, several sessions involved children trying to push the system to replicate a particular frantic musical style that they had played some minutes ago.
124
6. The Design of a Musical Flow Machine
Figure 6. Various expressions of musical excitement. Excitement is mostly provoked by listening to the system, rather than by actually producing music.
Figure 7. Various gestures showing listening and concentration.
6. The Design of a Musical Flow Machine
125
Figure 8. « Aspetta »: when one child forces the other to stop in order to listen to the machine.
Phases to Reach Flow In order to precisely define the role and importance of the Continuator in developing musical abilities, the phenomena described above deserve careful analysis on an individual basis (e.g. attention span, listening and concentrating, etc.). However, It is also worth studying the "life cycle" of these phenomena, or in other words, the progression of these f flow traits over time. We have observed that they always progress in the same order (see Figure 9).
Figure 9. A tentative sketch of the "life cycle" of the interaction mode with the Continuator.
One of the most interesting aspect of this time line is the role of Surprise/Aha: this occurs only once, at the beginning of sessions, or shortly after, and it may be argued that these Aha constitute a sort of initiatic experience, acting as a phase transition: once this step is performed, the users change completely their attitude, both towards the system and towards the environment (the experimenters). They become at once involved and self interested.
Observations Concerning Interaction Modes During the analytical stage (i.e. after the initial periods of surprise and excitement), many other behaviors were also observed with clear indicators. For instance, several children developed spontaneously innovative playing modes. Besides rediscovering standard playing modes such as playing individual notes, arpeggios or chords, they would produce sometimes remarkable arpeggios and clusters, but also new modes such as playing with elbows, turned around with hands in back, kissing the keys with their lips, etc.).
126
6. The Design of a Musical Flow Machine
An interesting phase in the interaction occurs when children, after having mastered the basics of the system, somehow abstract the concept of a "musical phrase". This is indicated by a typical gesture demonstrating a pretend "launching" of their own musical phrase (as if it were a golf ball). This gesture determines the end of their musical phrase and also creates the expectation of the system's response. It is remarkably similar to what professionals do themselves (see Figure 3, Bernard Lubat performing at Ircam and Figure 10). This gesture can also be interpreted as a desire to pretend-play, as described e.g. by (Wynants, 2004).
Figure lO.Children raising their hands and virtually « launching » the Continuator, after finishing a musical phrase. Another phenomenon worth mentioning - and worth studying more in depth - is the children's ability to engage in turn-taking behaviors in a spontaneous way. As mentioned before, the "rule of the game" (in this case a particular kind of master/slave turn-taking) was not explained to the children. However, they learned it, somehow implicitly, extremely quickly, after a few interactions. This ability to learn the rule of turn making is nontrivial. In particular, a child who knows how it works must also have an understanding of beginnings and endings of musical phrases: of his, as well as the system's. A few cases showed that this skill in ascribing an intentional ending to phrases generated by the system develops very early in the sessions. Figure 11 shows a typical scenario in which a child is confronted with an interesting situation regarding phrase endings. In this case, a particular continuation produced by the system started with the ending of the child's phrase followed by additional notes. This situation, although rare, can occur because the Continuator has no particular notion of beginnings or endings. Figure 12 shows the situation graphically. The phrase played by the child is schematically represented by two consecutive chunks of a few seconds. The last chunk is repeated by the Continuator, and followed by another one.
6. The Design of a Musical Flow Machine
127
Figure 11. A child clearly showing its interpretation of the Continuator's phrase endings, (a) shows him just finishing a phrase and waiting for the answer. The answer generated by the Continuator turns out to contain the last part of his input phrase followed by some additional notes. At the end of the first part of the continuation (i.e. the repetition of the child's phrase ending) the child gets ready to play again, assuming that the phrase played by the Continuator is finished. It is not (c), and the child shows his misunderstanding with a facial expression. Eventually (d), the Continuator ends his phrase, and the child resumes playing.
Figure 12. A short scenario demonstrating the complexity of the interpretation of phrase endings. The phrase structure is represented schematically using rectangle of various shapes. The important aspect of the scenario is that the last chunk of the input phrase is the first one of the continuation, thereby creating a false sense of ending for the child.
128
6. The Design of a Musical Flow Machine
4. Designing Interactive Systems The experiments conducted with the Continuator are interesting because they teach us something about the nature of human behavior, and in particular about what human beings find exciting. They also give us precious hints for the design of artificial systems that help to generate flow experiences. I now highlight only two key ideas which appear crucial for the success of the Continuator: reflection and emergent interaction protocols. Reflection in Interactive Systems A lot of research has been conducted into the properties that make interactive software usable. In the field of music, Sidney Pels (2002) analysed several interactive systems designed for creating novel aesthetic experiences and found that an ideal interactive system requires intimacy. This can be caused by the sensation that the system is somehow an extension of the user's body, thus ensuring the efficiency of the user's actions. The Continuator has this property because it uses traditional interfaces such as the keyboard. But it has something no existing instrument has: a learning component which introduces a totally new dimension of interaction. Because the system is able to learn and imitate the user's musical personality and style, the Continuator acts as a dynamic mirror, and I claim that most of the interesting behaviours observed in the experiments stem from this feature. This reflective capability brings a number of intriguing characteristics:
-
-
Similarity or Mirroring effect. What the system produces "looks" like what the user herself is able to produce. Agnosticity. The system's ability to reproduce the user's personality and style is learned automatically and agnostically i.e. without human intervention. No preprogrammed musical information is given to the system whatsoever. Incrementally. Interactive systems are not designed for short demos only. Because the user is constantly interpreting the output of the system, and altering his playing in response, it is important to consider the longer term behaviour of the system. Incremental learning ensures continuous evolution. Each interaction with the system contributes to changing its future behavior. Open Systems. Incremental learning is a way to endow the system with an organic feel, typical of open, natural systems (as opposed to pre-programmed, closed-world systems). A preliminary version of the Continuator used an algorithm based on Lempel-Ziv for learning. The Lempel-Ziv algorithm is not complete: several learning steps are needed to ensure that all the information contained in a new sequence is actually learned. As a consequence, the system is less faithful (and learns more slowly), but this very slowness increases the organic feeling of teaching something, which is less the case in the current version (which learns at once all the information contained in each new sequence). Seamless. The system produces material that is virtually indistinguishable from the user's input. This characteristic does not apply in the case of "classic" hyperinstruments, where the sonic effects are entirely produced through the system, and therefore no output is directly produced by humans.
6. The Design of a Musical Flow Machine
129
One important consequence of reflection is that the "user's center of attention" is not directed towards the end-product (the music), but towards the subject herself as it engages in the interaction. Engaging in an interaction thus becomes a means of discovering oneself, or at least exploring one's ability in the domain at hand (in our case, musical improvisation). This natural, deep interest in exploring oneself - particularly during the early childhood years - may explain why the Continuator is so appealing and we can design probably other successful interactive systems using the same principle. In some sense, these systems are an extension of the "second self (Turkic, 1984). Not only does the machine seem to "think," but it thinks like the user. This leads to a reversal of roles: the student becomes the teacher who teaches the machine about herself. Emergent Interaction Protocols The experiments so far have also shown various limitations of the current design of the Continuator. One limitation suggests that interaction protocols need to be flexible and emergent. Interaction protocols specify the rules of the game. They determine how and when the system decides to play. Like in conversations, these rules can vary, and a simple question/answering sequence is by far not the only possible protocol: lectures, Smalltalk (in the common sense meaning), exams, group conversations, baby talk, talking to oneself, etc. are all conversational situations in which the modes of interaction differ greatly. The issue of interaction protocols is closely related to the idea of music as conversation, put forward for instance by Bill Walker in his ImprovisationBuilder system (Walker, 1997; Walker & Belet, 1999). The ImprovisationBuilder is able to take turns with the player, and to detect, in the case of collaborative music playing, whose turn it is, using a simple analysis of the various musicians' inputs. However, this system as well as all the music interaction systems so far, rely on fixed interaction protocols. The question is whether it is possible to find mechanisms so that the machine adapts its protocols automatically and finds the best protocol suited to the context. Currently, several interaction protocols were designed and experimentally tested with the Continuator. Here are some of them, by increasing order of complexity (and represented graphically in Figure 13). Turn-taking. This mode is represented graphically as a perfect succession of turns, with no gap. The Continuator detects phrase endings, then learns and produces a continuation. It stops as soon as the user starts to play a new phrase. Turn taking with delay. The same as above, except that the Continuator stops only when the user finishes a phrases. This produces an interesting overlapping effect in which the user and the Continuator can play at the same time. Single note accompaniment. The Continuator produces an appropriate chordal accompaniment each time a note is played, and with the same duration. It stops the chord when the key is released. Phrase-based accompaniment. The same as above except that the chord is produced only at the beginning of a phrase. Collaborative. In this mode, the Continuator plays an infinite stream of music (based on material previously learned). The user can play simultaneously, and what he/she plays is taken into account by the Continuator, e.g. harmonically. The user's actions are here like high-level control, instead of a question to be answered.
130
6. The Design of a Musical Flow Machine
Figure 13. Various interaction protocols with the Continuator.
These various modes are in turn highly parameterised: The phrase length of the continuation, the dynamics, the tempo, etc. all depend on the interacxtion mode. In practice, it is easy to see that an infinite number of concrete interaction protocols could be defined, all tailored to a particular situation. The main problem with these protocols is that they are fixed, therefore sometimes not appropriate. Several observations (with children and professionals) showed that this rigidity is a severe limitation: For instance, a child enters the room and remains seated looking at the keyboard. Nothing happens because the Continuator waits for a phrase before it starts playing itself. In this case, an initiative by the system so as to trigger an interaction would be more relevant than just waiting. During an interaction, the parameters of a chosen protocol may be unadapted. For instance, we have observed that short answers in early stages of the interaction are preferable, but longer ones are more adapted as the user gets involved in the playing. Conversely sometimes a continuation which is too long can confuse the user (as was the case with one of the children who did not know how to stop the system). During a session, different modes may be needed. For instance, while the turntaking mode appears to be very natural at the beginning, several users (children and professionals) expressed the desire to play at the same time as the system. This requires indeed an explicit change of interaction mode (e.g. with a button), which breaks the fluidity of the interaction, and hence breaks the flow experience. Parameters other than the simple musical inputs should also be taken into account to influence the interaction protocol. The child's face in Figure 11 is a clear indication that the interaction was not as expected. The need for the system to control its own interaction protocol is therefore crucial to enhance its capacity to adapt. More generally, a fully flexible system should be able, like humans, to infer on-the-fly the protocol induced by a particular situation, as well as invent itself new protocols. A good music teacher would for instance switch from turn taking to, say, a teaching mode, and back to an accompaniment mode, etc. To achieve this functionality we need to make the system somehow self-reflective, i.e. able to manipulate explicitly the interaction protocol at a meta-level: select a protocol, create
6. The Design of a Musical Flow Machine
131
one, or infer one from the context. Such an endeavour is usually referred to as reflection in the world of language design (Demers and Malenfant, 1995). Building systems with such an architecture is not so simple. Successful attempts at designing self-reflective systems have usually been restricted to so-called structural reflection, i.e. the ability of a program to reason about its data structures. Conversely, systems able to switch between protocols (e.g. for mobile services) are able to do so only within a fixed list of predefined protocols. To extend flexibility in protocols requires the representation of the intentionality in interactions. The speech act theory proposed by Searle (1969) allows to represent utterances as so-called illocutionary acts, such as stating, requesting, commanding, and so forth. Every speech act consists of an illocutionary Force F applied to a proposition P. (Covington, 1998). This theory has given birth to a number of languages such as KQML (Knowledge Query and Manipulation Language), which are used for specifying various interaction protocols between agents, notably for electronic commerce (Finin et al, 1994). However, speech act theory and KQML are primarily aimed at facilitating knowledge sharing and transfer, in contexts where the set of possible illocutionary forces is known a priori. A language such as KQML may be used to specify each interaction between the machine and the user. But KQML does not solve (or even address) the motivational issue nor the problem of interpreting the "messages", or inferring optimal protocols.
5. Conclusion I have introduced the bold question of what makes an interactive system appealing, interesting and attention grabbing, taking examples from experiments performed with the Continuator system, a musical interactive system. I have shown that the Continuator may be seen as a particular example of a flow machine, i.e. a machine which generates flow-like experiences in users. The reflection principle, which causes a mirroring of user's behavior after a learning process, appears to be of critical importance. Current limitations concern primarily the lack of flexibility in interaction protocols. The Continuator proposes many different interaction modes, but is not able to detect automatically the best mode, nor to create new ones on-the-fly to adapt to unexpected situations. This is clearly an important research line for the future, which can build further on a long history of research in computational reflection. Reflection for building Reflective Systems: although it seems a play on words, I believe this is not entirely coincidental.
Acknowledgement The research in this paper rests on the work and enthusiasm of many people, not in the least the children that participated in the first joyful experiences with the Continuator, and the professional musicians, in particular Bernard Lubat, Alan Silva, and Gyorgy Kurtag Jr., who experimented with the Continuator in live performances. I also thank Anna-Rita Addessi for collaborating on the Bologna experiments and for discussions on the significance of some of the underlying principles presented here. For the present paper, I thank Atau Tanaka and Mary Fardoob for comments on earlier drafts. I also thank Marleen Wynants for helping to shape the present paper, particularly the part on designing interactive systems.
132
6. The Design of a Musical Flow Machine
References Addessi, A.-R. & Pachet, F. (2004) "Musical style replication in 3/5 years old children: experiments and analysis", submitted to British Journal of Music Education. Brooks, Hopkins, Neumann & Wright. "An experiment in musical composition." IRE Transactions on Electronic Computers, 6(1), 1957. Chuang, John. (1995). Mozart's http://sunsite.univie.ac.at/Mozart/dice/
Musikalisches
Wiirfelspiel,
web
site:
Cope, David. (1996). Experiments in Musical Intelligence. Madison, WI: A-R Editions, 1996. Covirigton, M.A. (1998) "Speech acts, electronic commerce, and KQML," Decision Support Systems volume 22. Csikszentmihalyi, Mihaly. (1990) Flow. The Psychology of Optimal Experience. Harper and Row, New York. Cope, David. (1996). Experiments in Musical Intelligence Madison, WI: A-R Editions. Delalande F. (1984), La musique est un jeu d'enfant, Buchet/Chastel, Paris. Demers, Francis-Nicolas and Malenfant, Jacques (1995). Reflection in logic, functional and object-oriented programming: a short comparative study. In IJCAI '95 Workshop on Reflection and Metalevel Architectures and their Applications in AI, pages 29-38. Pels, S., (2000). Intimacy and Embodiment: Implications for Art and Technology, Proceedings of the ACM Conference on Multimedia, Marina del Sol, CA, pp. 13-16, Oct. Finin, Tim Fritzson, Richard McKay, Don McEntire, Robin (1994). KQML as an Agent Communication Language Proceedings of the 3rd International Conference on Information and Knowledge Management (CIKM'94). lizuka, Hiroyuki and Ikegami, Takashi. (2002) Simulating Turn-Taking Behaviours with Coupled Dynamical Recognizers, The 8th International Conference on the Simulation and Synthesis of Living Systems, University of New South Wales, Sydney, NSW, Australia 9th-13th December. Ingalls, Daniel (1981). Design Principles Behind Smalltalk, BYTE August 1981 Special Issue on Smalltalk. Kaplan, Frederic. (2001). Artificial Attachment: Will a robot ever pass Ainsworth's Strange Situation Test?, in Hashimoto, S., editor, Proceedings of Humanoids 2001: IEEE-RAS International Conference on Humanoid Robots, pages 125-132. Kay, Stephen. (2000). The Korg Karma music work station, Korg Inc. King, Andrew (2003). Speed Up Your Site: Web Site Optimization. Chapter 2: Flow in Web Design.
6. The Design of a Musical Flow Machine
133
Mogi, Ken. (2003). Onceness, in The Future of Learning, Tokoro, M. and Steels, L. Eds, IOS Press. Pachet, Francois. (2002a) The Continuator: Musical Interaction with Style. In ICMA, editor, Proceedings of ICMC, pages 211-218, September 2002. ICMA. best paper award. Extended version in Journal of New Music Research, 31(4), 2003. Pachet, Francois (2002b) Interacting with a musical learning system: the Continuator. In C. Anagnostopoulou, M. Ferrand, A. Smaill, editor, Music and Artificial Intelligence, Lecture Notes in Artificial Intelligence (vol. 2445), pages 119-132, Springer Verlag. September 2002. Pachet, Francois. (2002c) Playing with virtual musicians: the Continuator in practice, IEEE Multimedia, April-June 2002. Rinaldi, Carla (2003) The Joys of Preschool Learning, in the Future of Learning: Issues and Prospects, Tokoro, M. and Steels, L. Eds, IOS Press Risset, Jean-Claude and Van Duyne, Scott. (1996). Real-time Performance Interaction with a Computer-Controlled Acoustic Piano, CMJ 20(1), pp. 62-75. Searle, J. R. (1969), Speech Acts. An Essay in the Philosophy of Language. Cambridge. Shannon, C. E. (1948). "A mathematical theory of communication", Bell System TechnicalJournal, vol. 27, pp. 379-423 and 623-656, July and October, 1948. Steels, L. (2004) The Autotelic Principle. In: Hafner, V., F. lida, R. Pfeifer and L. Steels (eds.) (2004) Embodied AI. Proceedings of the Dagstuhl Seminar. Lecture Notes in Computer Science. Springer Verlag, Berlin. Trivino-Rodriguez, J. L. Morales-Bueno, R. (2001) Using Multiattribute Prediction Suffix Graphs to Predict and Generate Music, Computer Music Journal 25 (3) pp. 62-79. Turkic, S. (1984). The second self: Computers and the human spirit. New York: Simon and Schuster. Walker, W. (1997) A computer participant in musical improvisation. In CHI 97 Electronic Publications. Walker, W. Belet, B. (1999) Applying ImprovisationBuilder to interactive composition with Midi piano. In ICMC 99, Hong-Kong. Webster, Peter R. (2002). Computer-Based Technology and Music Teaching and Learning, in R.Cowell-C. Richardson (EDS). The New Handbook of Research on Music Teaching and Learning, Oxford University Press, pp. 416-439. Wynants, M. (2004). Pretend Play as Learning: Case Studies from the Home. This volume.
134
6. The Design of a Musical Flow Machine
FRANCOIS PACHET is a civil engineer, with a PhD in artificial intelligence from University of Paris 6. He has been assistant professor at Paris University for 5 years, and now leads a research team at Sony Computer Science Laboratories in Paris. His main research interest is the notion of "Chrono-attentiori": the ability of a temporal process to attract and sustai attention in time. He investigates what makes temporal forms of entertainment (music, games, movies) interesting, appealing, attention grabbing and sustaining. Francois Pachet has led several projects dealing with particular Facets of chrono-attention from a system design perspective (how to design systems which are chrono-attentive): MusicSpace (how to make interactive music listening attractive through spatialization), Continuator (how to interact and generate interesting musical dialogues and interactions), the Popular Music Browser (how to provoke serendipity and A-ha! effects in music catalogues browsing). His recent interests focus on chrono-attention from a psychological perspective (how does our attention system works), and educational perspective (how to exploit chrono-attention in teaching).
7
136
"I am worried about too much technocentrism or seiento-centrism. It is not because we can do certain things with science, that we have a complete insight in the phenomena. We should not be blind for the enormous complexity of reality. For example, in architecture and urbanism, there are people who carry out a rational scientific analysis and then impose it on a city. Later it turns out that their ideas do not work, that the city does not evolve like they thought it would. Sometimes it is a real disaster. The same is true for education. There are things that policymakers and educational planners with all their science do not grasp, whereas a lot of teachers and parents have many intuitions and wisdom from their practice. Even though we reflect here upon learning and education in a scientific way and even though we propose scientific ideas and technologybased tools, I want to stress, particularly towards politicians and policy planners, that there are limits to these scientific insights and that the basic intuitions of parents and educators and their experience should always take primacy." Luc Steels
A Learning Zone of One's Own M. TokoroandL. Steels (Eds.) IOS Press, 2004
13 7
7.
The Architecture of Flow Luc Steels
1. Introduction Nowadays many people argue that learning must be fun, but what does that mean and how can it be achieved? The fun involved in learning can surely not be like the kind of shortterm 'kick' one gets from going down a roller coaster. Although this gives instant pleasure (at least to some people), true learning requires a more sustained, long-term motivational force that stimulates the learner to keep seeking new challenges and gets her through periods where the going is tough. There are people who have a strong inner motivation to engage in challenging activities and consequently they are the ones that tend to excel. These people spend long hours in a highly focused state and seem to enjoy tremendously what they are doing. They are highly effective learners and seek out learning opportunities themselves. Typically these learners are very creative and make major breakthrough-contributions to their fields. But such focused activity is also naturally observed in children. When they are truly engaged, they strongly resist having to interrupt their play, and they generally don't like to be put in a rigid framework that constrains their activities. We need to understand better the nature of such behaviour and how it can come about, so that we can build learning environments that stimulate self-motivated learning and teach students what the joy of learning is like. This can be achieved by a theoretical investigation in the concepts underlying self-motivated learning and development, and by looking at case studies where high states of motivation are clearly reached by students and teachers alike. Such case studies can be found in some pre-school environments, like in Reggio Emilia (Rinaldi,2003), in successful art academies and design institutes, like the Antwerp Fashion Academy or the Ivrea Design Institute, in innovative business schools, and even in some enlightened science universities where students are immersed in laboratory practice and work intensively in group towards the realisation of particular projects. Unfortunately, most schools and curricula in contemporary societies, particularly at the level of higher education, do not exhibit the necessary characteristics to generate sustained self-motivated learning. This paper first surveys discussions on motivation in psychology and its impact on educational practice. Then I report briefly on attempts to make cognitive science models of
138
7. The Architecture of Flow
'autotelic agents' which have an inner drive towards self-development. The paper concludes with some reflections on what the implications for education could be. 2. Theories of Motivation Reinforcement Learning Most models in psychology and neuroscience today are still rooted in the behaviourist framework of reward and punishment, originally coming from the work of Skinner and his associates (Skinner, 1953). Consequently the pedagogical strategies that we find in contemporary schools are still to a large extent based on these ideas. There are three major assumptions: First, behaviourism argues that the environment (or a trainer) can trigger adaptation and learning by providing rewards, for example food or other means that give direct pleasure, or punishments, for example through the inducement of corporal pain. Rewards reinforce specific behaviours because they inform the individual that they are beneficial, in other words that a viable state can be reached and maintained. Punishments signal that the behaviours that were enacted need to be abandoned or new knowledge and skills need to be acquired. Second, classical behaviourism proposes that learners start with a repertoire of reflex behaviours and an innate value system. New behaviours are shaped by reward and punishment produced by the environment. When a trainer or educator hands out the reward or punishment, she can push development in specific directions and the trainer's value system may become progressively internalised by the trainee. This opens up the possibility of cultural transmission of value systems. Third, classical behaviourism proposes that this reinforcement framework is an adequate theory of action selection and motivation, in the sense that the main purpose of an individual is to seek reward and avoid punishment, and all the rest (acquisition of new behaviours and internalisation of a value system) follows. Behaviourism has had a tremendous impact both on education and on behavioural and brain sciences. In the educational domain, it has lead to an emphasis on disciplinary relations between teachers and students where fear for punishments and exams are the primary instruments for motivating students. Technologically, it has lead to so called 'programmed instruction' systems, as practiced for example in 'language laboratories'. In the behavioural and brain sciences, it has given rise to a tradition where results from experiments on rats are routinely transferred to humans and where representation making, meaning, personality development, etc. are taboo subjects. Criticism of the reinforcement paradigm In the past decades, there has been a growing feeling that the behaviourist approach is not a very 'humane1 way to organise education, even though it can sometimes give very good results for the acquisition of specific skills. Moreover several educational thinkers, such as Illich(1973) or Postman(1972), have pointed out from the early seventies onwards that there are severe disadvantages using this approach in the long term, particularly for educating children to become responsible, autonomously acting members of the
7. The Architecture of Flow
139
community. They argued that this approach to education stifles creativity and personality development. They proposed an alternative approach: to give pupils more freedom, abandon systematic grading, and frame the teacher in the role of supporting the learning process rather than being the source of reward and punishment. The pupil is encouraged to explore domains of knowledge by herself, instead of being supplied with pre-established answers. Specialised schools, like the Freinet schools (Sivell,2004), the Reggio experiment in Italy (Rinaldi,2003), or the Sri Atmananda Memorial schools in India are examples where this approach has worked - albeit usually only for pre-school or the first grades in primary school. But this alternative learning model has not scaled very well to the whole school system and is now regarded as inapplicable on a large scale. Where it has been tried beyond primary school, academic achievement of students are plummeting and teachers are confronted with pupils that lack the motivation to learn, resulting in boredom and subsequent violent or deviant behaviour. It seems that the Utopian models proposed in the nineteen seventies only work when the student body is very homogenous, the community, including the parents, fully co-operate, and the non-school environment (particularly the leisure activities) do not overpower children as they do in our contemporary society. In a recent essay that was distributed to all teachers in France, the French minister for education Luc Ferry (2003) blamed the current educational crisis in France to a breakdown of the traditional model, with the teacher as authority and transmitter of knowledge. He attributed this breakdown to the propagation of ideas related to 'free1 alternative schooling and suggested a return to the traditional teacher-centred model. French teachers welcomed his proposals with public burnings of the book and month-long strikes! The latter was actually a symptom of a deeply rooted crisis in the teaching profession, partly caused by an abandonment of the traditional role of the teacher. In other words, the teachers are as much lost as the students.
3. Flow theory So how can we go beyond the traditional reinforcement learning framework without falling into the other extreme of a completely unstructured free environment? The theory of selfmotivated learning and behaviour may give us a potential clue. This theory was originally developed by the humanistic psychologist Csikszenmihalyi, based on studying the activities of painters, rock climbers, surgeons, and other people who showed to be deeply involved in some very complex activity, often for the sake of doing it, i.e. without direct reward in the form of financial or status compensation (Csikszentmihalyi,1978). He called these activities autotelic. "Autotelic" signifies that the motivational driving force ("telos") comes from the individual herself ("auto") instead of from an external source, administered by rewards and punishments. Autotelic activities induce a strong form of enjoyment, which Csikszentmihalyi has characterised as "flow". The word "flow" is a common sense word and so there is a risk to interpret it too broadly. Csikszenmihalyi intends a restricted usage, being a state which often occurs as a side effect of autotelic activities:
140
7. The Architecture of Flow
People concentrate their attention on a limited stimulus field, forget personal problems, lose their sense of time and of themselves, feel competent and in control, and have a sense of harmony and union with their surroundings. (...) a person enjoys what he or she is doing and ceases to worry about whether the activity will be productive and whether it will be rewarded, o.c. p. 182. Because the activity is enjoyable, the person who experiences this enjoyment seeks it again, i.e. it becomes self-motivated. Moreover due to the high concentration and the strong selfmotivation, learning takes place very fast. The learner is eager to find the necessary sources and tools herself and spends time on the acquisition of skills, even if they are not exciting in themselves, as long as they contribute to the autotelic activity. Given this definition of flow, it is quite obvious that many people will have experienced some form of flow in their life, and that children in particular enter into flow experiences quite often, particularly during play. Flow is sometimes associated with the ultimate high experience of the rock climber that has finally managed to climb Mount Everest, but that is an exceptional situation. Flow - as defined here - is much more common and can just as well happen in every-day experiences like playing with children or engaging in a long-term love relationship. It is also important to distinguish flow from directly pleasurable activities like going down a roller coaster. A key difference is that the activity must in itself be challenging - otherwise there is no feeling of satisfaction after difficulties have been surmounted. Moreover there must be a steady progression in the nature and particularly the level of the challenge. This is the reason why child rearing can be so enjoyable and fascinating. A child keeps developing all the time - which is what makes the interaction fun - and that creates continuously new challenges for the parent to figure out what she is thinking, what she might want to do or not do, and so on. The rock climber can also scale up the level of difficulty with which rocks are being climbed or the kinds of rocks that are tackled. Similarly, the musician can first play easy pieces and then steadily move up. She can first play with other amateur musicians and then play with better and better musicians. The performance can be first for a few friends, but then for a larger and larger unknown audience. An obvious key question is: What makes activities autotelic? This is where Csikszenmihalyi's most important contribution comes in. He argues that it lies in a good balance between high challenge, generated through the activity and perceived as meaningful to the individual, and high skill required to cope with this challenge: Common to all these forms of autotelic involvement is a matching of personal skills against a range of physical or symbolic opportunities for action that represent meaningful challenges to the individual, o.c. p. 181 When the challenge is too high for the available skill or the opportunity for action so bewildering that no clear course can be seen, anxiety sets in, particularly when there is no hope to develop appropriate skills by learning. The person gets paralysed and eventually may develop symptoms of withdrawal and depression. When the challenge is too low for the available skill, boredom sets in and the long-term reaction may be equally negative. So the optimal regime is somewhere between the two, when there is a match of challenge and skill (figure 1).
7. The Architecture of Flow
141
Figure 1. The flow region is defined by a balance between challenge and skill. When challenge is too low for the skill level reached, there is a feeling of boredom. When skill is too low with respect to the level of challenge, there is a feeling of anxiety. The individual needs the ability to decrease challenge when this become too high, or alternatively be given the time to build up more skill to cope with the challenge. But it is equally important that the individual is in control of the challenge level, so as to increase challenge when the skill has become higher, otherwise activities become boring and selfmotivation is lost. Flow is not a steady state but occurs in the process of reaching a balanced state with high challenge and skill. This balanced state never lasts. When an activity is carried out often, skill normally increases so that the activity becomes boring and new challenges need to be found. So the individual seeking the flow experience is always 'on the move1. Implications of the Flow concept The insights of Csikszenmihalyi and those who have elaborated these insights further (see for example papers referenced in Csikszentmihalyi and Selega(2001)) are potentially of enormous significance. They hint to an alternative to classical behaviourism and hence to a very different approach to psychology and neuroscience compared to the one suggested by the behaviourist paradigm. They also complement the information processing approach to psychology, which has tended to ignore issues related to emotion, motivation and affect. The implications for education are very deep. The current school system, in as far as it is built on the concept of reward and punishment, does not create the conditions for flow, in fact it often totally ignores them. The meaningfulness of an activity is not considered as primary (often it is not even meaningful to the teacher). Instead activities are imposed in a top-down fashion and based on statewide uniform programs, such as the so-called 'college unique' program in France. The learners have no individual control over the challenge levels in an activity. This results in giving up any effort to learn as soon as they see no way to cope, or to get bored and become totally disinterested in the school as a whole when the challenge is not high enough. A typical example of a traditional teaching situation where the conditions for flow are destroyed, has been described by John Matthews, who is an authority in visual art education for children (Matthews, 1993). In a provocative essay entitled "Art Education as a form of child abuse", he argues that the natural desire and enjoyment that children experience in
142
7. The Architecture of Flow
developing their visual skills is totally ignored by the educational system. Specifically in the case of Singapore it even becomes completely wiped out so that children lose all interest in the matter: Generally, from the primary level onward, rigid, highly prescriptive approaches to drawing are employed. Typically, the teacher instructs, in a series of strict stages, toward a preenvisaged, fixed end-point, usually of the most stereotypical, banal kind. The children are required to obediently follow every step, and if variation is allowed at all, it is of the most trivial kind. Identical art works are the result. If a child has varied a little from the fixed procedure, he or she is often criticised, if not punished. Many recorded observations testify that, out of a class of forty children, the one or two art works which possess any vitality are condemned and their makers accused of an inability to concentrate or follow instructions. Other recordings made in private art classes for the very young show adults sometimes altering their children's drawing or even re-drawing the image wholesale. The drawing process is often totally dominated by the adult, who gives strict instructions about the sequence of strokes, from which any variation is strictly criticised. Teachers sometimes even physically manipulate the child's drawing arm itself. By the time children get to the end of primary school, they have lost all interest in drawing or painting and are visually illiterate. The culprit is obviously that goals are imposed in a top-down fashion through reward and punishment instead of being generated in a selfmotivated way. The 'free school1 paradigm proposed by Illich and others goes a step in the right direction because it accepts that the learning environment must be adapted to the individual student and that the learner must have control to some extent about what should be learned and when. But at the same time, this paradigm equally ignores the necessity for the teacher and the school environment to create scaffolded opportunities for action that generate the challenges for self-motivated learning. So two things are needed. A learner must be able to feel some control of the challenge level, but at the same time, the environment is crucial in generating new opportunities and providing structure to the learning experience. For example, the rock climber must be able to choose a more difficult or more challenging rock but there must of course be rocks in the first place. The rocks are the learning environment for the climber, a continuing source of increasingly more challenging opportunities for enhancing the skill. In the same way, the school, the teachers, and the parents must create the learning environment, which must be full of opportunities, challenges, but also the resources to progressively cope and grow in the acquisition of a particular skill. According to this viewpoint, the most important thing that children have to learn in school is not so much knowledge or skill about a particular subject matter (for example, copying the drawing of the teacher in exactly the same way), but rather the experience that there are such things as autotelic activities, which are not only highly rewarding on a long-term time scale at a deep level, but also a foundation for self-actualisation. The flow concept is also relevant for the teachers: Teaching can be a highly rewarding activity, when there is a smooth dynamical interaction between teacher and students and students progressively become better in a certain domain of expertise under the guidance of the teacher. But this requires that the teacher must have a way to control the situation
7. The Architecture of Flow
143
enough to generate the flow experience. However, a teacher won't get a grip on the teaching process and motivation will break down if a class contains too many students of different levels and backgrounds, if the students are not at all interested to engage themselves in the learning process, or if the content changes too fast and is imposed topdown by administrations. The school must be equally rewarding for the teacher as for the student. If the teachers are not enjoying what they are doing, then they are not going to be able to sustain appropriate learning environments either. It has been argued that interference from political and bureaucratic institutions that impinge on the freedom of teachers to set the course of action is one of the reasons why teachers become anxious or bored with their jobs. The studentteacher dynamic has been broken in contemporary schools from both sides (Nevejan,2003). Impact The first obvious way in which the ideas associated with flow theory can impact education is to propagate them widely in the teacher community and among politicians or administrators that set or enact educational policy. This work has been taken up by Csikszentmihalyi in a series of popular publications, such as (Csikszentmihalyi,1990). The second approach is to engage in a scientific study of flow. This challenge has also been taken up by several researchers, including Antonella Delle Fave(2004). Most of this research at the moment is observational in nature; it tries to find out how people characterise self-motivated activities, when they have them, which groups have them more than others, how they are triggered, etc. This work is of course extremely useful to get an empirical grip on the phenomenon, but there is also another approach possible, which is to develop operational models of the cognitive mechanisms that underlie autotelic behaviour. Traditional symbolic cognitive models developed in AI (Newell, 1990), as well as behaviourist and neo-behaviourist models, such as connectionism (Churchland and Sejnowski,1995) have well worked out formalised theories of the architecture of a 'cognitive agent'. This has made it possible to define and study systems based on reinforcement learning (Sutton and Barto,1998) and even to use these models in artificial devices like robots that are acquiring crucial features of the world and how to behave in it (Steels and Brooks, 1995), (Pfeifer, 2000). Recent work (Steels, 2004) has shown that it is in fact possible to develop mathematical and computational models for autotelic behaviour and instantiate them in self-developing robots. This research can tell us something about the kinds of internal mechanisms that are required to induce autotelic behaviour and how they work. It can then be used to stimulate the development of these mechanisms in human children or to incorporate them in learning environments. Yet another approach (explored by Fra^ois Pachet (2004)) is to build devices that elicit flow in users. One example he discusses is the 'Continuator', a music system that acquires the statistical properties of musical input introduced through a keyboard and then mirrors this with its own output which consists of variations on the learned patterns. Experiments with human subjects show that this simple principle (when implemented well) generates enormous excitement in children and adults alike. Because the system continuously learns, the interaction remains always novel and users try to elicit more and more complex behavior. Building this kind of 'flow machine' (in the sense of machines that help to generate flow) has a dual purpose: They can tell us something about why humans become
144
7. The Architecture of 'Flow
very excited in interacting with such devices, and also what features a system needs to have in order to elicit flow. In my own work, I have focused on the second approach, i.e. trying to make operational models of agents that simulate aspects of autotelic behaviour. The next section reports briefly on the kinds of models that I have been able to construct so far.
4. Autotelic Agents The models of agent architecture that cognitive scientists have attempted to construct involve a large number of components: for sensory perception, motor control, categorisation, memory, situation recognition, action planning, action selection, motivation, communication, etc. Each of these components takes some input (for example a bitmap produced by a camera in the case of a vision system) and maps this onto some output (for example an analysis of the scene). Although earlier versions of agent architectures did not take learning into account, it is now common to make every component in the agent's architecture adaptive: new low level visual feature detectors arise, new actions are learned, new plans are made and stored for later, new perceptual categories are acquired, new conventions for communication invented or adopted, etc. Today we have already a vast range of quite detailed operational models for many of the components and learning strategies for these components (Pfeifer and Scheier, 2000). Some are argued to be realistic models of the brain. The learning strategies are either unsupervised, meaning that they attempt to find recurrent structures in large data sets using statistical methods, or supervised, meaning that they rely on a re-enforcement signal coming from the environment or from a teacher. Supervised learning mechanisms operationalise the behaviourist re-enforcement framework discussed earlier in the paper. In all proposals so far, there is however no overall motivational mechanism that is helping to decide what to do, while pushing the agent towards greater levels of achievement. Many agents disussed in the literature are purely reactive, responding to environmental stimuli which might be a good model for an ant but not for a human person. If any kind of development is modeled, it is completely steered by human designers with carefully scaffolded input data and reward functions. Learning of a particular task is usually studied in complete isolation, for example, the researcher focuses on the acquisition of new visual features and supplies input-output pairs focusing exclusively on this task. To make cognitive agents that are autotelic, I argue that four ingredients need to be added to the agent architectures proposed so far in the literature:
7. The Architecture of Flow
145
Figure 2. The quality of experience can be defined as a function of challenge and skill levels. They are compared with the standard subjective mean (SM) skill and challenge levels experienced by the agent in the past (see Delle Fave, this volume)
1. The agent needs a way to sense the quality of experience, in the sense of Csikszentmihalyi (1990) (see figure 2). Because quality of experience is a function of skill and challenge, it follows that the agent needs internal sensors that are capable to track skill levels, and categorisation processes that estimate the degree of challenge associated with a particular situation in the environment. Because quality of experience is based on a comparison to a subjective mean (SM) of skill and challenge levels, it follows that challenge and skill levels need to be tracked over longer periods of time. Clearly developing sensors and categorisations of skill and challenge is not a static thing, but must be learned and continuously adapted based on feedback of performance. 2. The agent needs a way to increase or decrease the challenge level of a task. This can be done by seeking out (or avoiding) environments or tasks that have more (or too much) complexity, or by regulating internally with what degree of complexity environmental stimuli should be treated and with what precision actions executed. For example, a vision system might perform a very detailed scene analysis, recognising all objects in the scene and their interrelationships (in other words set a very high challenge level), or alternatively it may perform a simple rough scene analysis with only vague contours of objects and no further recognition (implying a very low challenge level). 3. The skill levels of the agent change due to learning. So if a situation is encountered frequently, skills are exercised and skill levels automatically increase. Alternatively, if a skill is not exercised, it deteriorates, causing the skill level to decrease. (Compare this to experience in learning a foreign language. Knowledge of the language rapidly increases when one speaks it a lot but it decreases just as quickly when one no longer has the opportunity to speak.) Exercising a skill requires psychic resources (memory,
146
7, The Architecture of Flow
processing time), and this may be under partial control of the agent, so that skill levels can effectively be decreased causing the agent to pay less attention. 4. In most action selection models so far, the criteria for choosing an action are uniquely based on the utility of the action towards achieving specific goals. Action is here taken in a broad sense and also includes perception, i.e. attending to a particular object in the environment. In the case of autotelic agents, optimising the quality of experience becomes an additional criterion for action selection. In other words, an action that has less utility for the agent, but has greater value in optimising the quality of experience, might still be preferred. This principle needs to operate on different time-scales: Long term benefit is larger if a steep slope of skill increase has to be achieved to cope with a particular challenge because that will enable future experiences with higher skill and higher challenge. 5. Our experiments have shown that autotelic agent need occasionally to 'shake up' the challenge levels they have set themselves, to avoid getting into locally optimal states which miss out on opportunities to reach globally higher skill and challenge levels. In other words, the agent cannot always rely on the environment to pose new opportunities and challenges, but must seek them out, thus generating an ever increasing selfdevelopmental spiral of higher challenge followed by matching skill. 'Flow' is usually characterised as a state in which there is a high concentration and sharp attention to a specific task and aspects of the environment relevant for that task. In the agent architecture discussed here, attention is operationalised by a focusing of mental resources (time, memory) and an increase of challenge levels for the relevant components or tasks (with subsequent decrease of irrelevant components). Flow automatically develops when the agent self-regulates challenge and skill in order to maximise the chance that optimal experience might occur. So flow is an emergent side effect of the model and not a causal factor in itself. Once a model of a single autotelic agent exists, it is possible to experiment with a population of autotelic agents which each have their own developing skill and challenge levels. Often one agent must lower his challenge levels, and hence the complexity of his behavior, because the agent with which he is interacting is not yet able to deal with this complexity. On the other hand, an agent that is interacting with another agent that has higher challenge levels, needs to increase his skills to cope with the interaction. So agents coordinate their challenge levels and can push each other up towards reaching higher challenges, and hence opportunities for reaching higher skills.
5. Implications for education This is not the right place to go into further technical detail of these operational models, nor discuss simulation results (see Steels, 2004). Instead we turn to the question of what we can learn from these models with respect to bettering human learning environments. I believe the implications are quite straightforward and essential for the practical implementation of learner-centered learning: 1. It appears essential that students develop an awareness of their own skill levels. In some sense exams and tests already do this. But often tests are seen as hurdles that need to be overcome, and once they are overcome, everybody in the same class is seen as being on
7. The Architecture of Flow
147
the same level. This is clearly not valid. Every student has their own skill levels which are dynamically changing, and tests should be presented to the student as a tool to monitor their skill levels. 2. Students should develop competence in estimating the challenge level in a particular task situation. Usually the teachers scaffold challenge, for example by first giving simple exercises and then gradually providing more complex ones. This is in fact one of the major tasks of the teacher, and Vygotsky's notion of the zone of proximal development captures this idea very well. But in real life, one must be able to decide yourself when a challenge is appropriate for your skill, and this implies that one must develop appropriate categories for judging challenge levels. 3. Students should be stimulated to develop an increased awareness that skill can be increased through learning and that skill needs to be maintained by continuously exercising it. The motivation for learning has to come from the realisation that increases of skill are going to bring a whole range of new challenges within grasp. They should be able to see that a particular problem or task is too difficult right away but that work on a supporting skill can help to achieve the right level. 4. Finally the learner needs to be given control of the challenge level, perhaps not all the time because then there is a risk, particularly with less ambitious students, to remain stuck with skill levels that are below the ultimate capacity of the student. But, as reported by Csikszentmihalyi in many of his case studies, self-control of the challenge level is one of the absolute prerequisites for reaching a flow experience. Self-motivation dies out if somebody is felt to have no control of his destiny. It is obvious that many curricula and learning environments in one way or another incorporate some of the principles expressed here, thanks to the intuitions of good teachers. But just as many school environments violate the principles suggested here. The more we learn about autotelic experiences and how they come about, the better we can tailor learning environments to exploit the natural tendencies of human beings to seek optimal experience and balance challenge and skill through self-motivated learning and self-control. In many cases, it is a matter of giving learners much greater control in setting challenge levels, deciding what resources to bring to bear, seeking out environments in which supporting skills might be acquired, or interacting with peers that have slightly higher skills.
6. Conclusions Flow is part of a long-term engagement, in which the individual is continuously trying to go beyond existing skills to reach new heights. Only then the deep experience of satisfaction can occur that sustains new learning. When a school curriculum can be organised in such a way that students feel progressively challenged - without experiencing the anxiety of overreaching their skill - very powerful natural learning processes get triggered and students become self-motivated. This paper briefly reported on efforts to make operational models of 'cognitive agents' that incorporate some of the intuitive ideas associated with flow. These models include components for sensing skill and challenge levels and determining the quality of experience, mechanisms to increase or decrease challenge and skill levels and thus focus
148
7. The Architecture of Flow
physic energy, and action selection mechanisms that take optimal experience into account in addition to mere utility with respect to short-term goals. Much remains to be discovered about the nature of autotelic activities and the design and experimentation with cognitive architectures that integrate notions of flow is a very new area, bjut I hope that the ideas expressed in this paper make it clear that progress is being made and that the potential impact on the future of learning is high.
Acknowledgements I thank all the participants from the various workshops on the 'Future of Learning1 for many stimulating thoughts and comments, and Mario Tokoro for continuing encouragement to pursue these topics in the best circumstances imaginable at the Sony Computer Science Laboratory in Paris. Discussions with Frederic Kaplan and Fra^ois Pachet have helped to fine-tune the agent architectures of flow which form the background to this paper. I also thank Marleen Wynants for her critical comments on the notion of flow in general and on the present paper in particular.
References Churchland, P. and Sejnowski, T. (1995) The Computational Brain. The MIT Press, Cambridge Ma. Csikszentmihalyi, M. (1978) Beyond Boredom and Anxiety: Experiencing Flow in Work and Play. Cambridge University Press, Cambridge. Csikszentmihalyi, M. (1990) Flow. The Psychology of Optimal Experience. Harper and Row, New York. Csikszentmihalyi, M. and I. Selega (Editors) (2001) Optimal Experience: Psychological Studies of Flow in Consciousness. Cambridge University Press, Cambridge. Delle Fave, A. (2004) A Feeling of Well-Being for Learning and Teaching. This volume. Elman, J. (1993). Learning and development in neural networks: The importance of starting small. Cognition, v. 48. p. 71-89 Ferry, L. (2003) Lettre a tous ceux qui aiment 1'ecole. F. Jacob, Paris. Hedegaard, M. (2003) In: Tokoro, M. and L. Steels (eds.) The Future of Learning. Issues and Prospects. IOS Press Amsterdam, 2003. p. 79-92. Illich, I. (1973a) Deschooling Society, Harmondsworth: Penguin. Johnson, M. (1996) The Infant Brain. In: Tokoro, M. and L. Steels (eds.) The Future of Learning. Issues and Prospects. IOS Press, Amsterdam, p. p. 101-116. Matthews, J. (1993) Art Education as a form of child abuse. Lecture for the National Institute of Education Singapore.
7. The Architecture of Flow
149
Nevejan, C. (2003) Integrated Learning Environments. In: Tokoro, M. and L. Steels (eds.) The Future of Learning. Issues and Prospects. IOS Press, Amsterdam, p. 33-49. Newell, A. (1990) Unified Theories of Cognition. Harvard University Press, Cambridge, MA. Papadimitrious, C. and K. Steiglitz (1998) Combinatorial Optimization: Algorithms and Complexity. Dover, New York. Pfeifer, R. and C. Scheier (2000) Understanding Intelligence. The MIT Press, Cambridge Mass. Postman, N. (1972) The End of Education. Redefining the value of school. Vintage Books, New York. Rinaldi, C. (2003) The Joys of Pre-school learning. In: Tokoro, M. and L. Steels (eds.) The Future of Learning. Issues and Prospects. IOS Press, Amsterdam, p. 57-78. Sivell, J. (2004) Tools for Embodied Teaching: Celestin Freinet and the Learner-Centered Classroom. (This volume) Skinner, B.F. (1953). Science and Human Behavior. Macmillan, New York. Steels, L. and R. Brooks (eds.) (1995) The Artificial life Route to Artificial Intelligence. Building Situated Embodied Agents. Lawrence Erlbaum, New Haven. Steels, L. (2003) Social Language Learning. In: Tokoro, M. and L. Steels (eds.) The Future of Learning Issues and Prospects. IOS Press, Amsterdam, p. 133-159. Steels, L. (2004) Autotelic Agents: Robots that have fun. In: Hafner, V., F. lida, R. Pfeifer and L. Steels (eds.) (2004) Embodied AI. Proceedings of the Dagstuhl Seminar. Lecture Notes in Computer Science. Springer Verlag, Berlin. Sutton, R. and A. Barto (1998) Reinforcement learning: an introduction. MIT Press, Cambridge Ma. Thelen, B. and L. Smith (1994) A dynamic systems approach to cognition and development. MIT press, Cambridge Ma.
150
7. The Architecture of Flow
LUC STEELS is director of the Sony Computer Science Laboratory in Paris and professor of Artificial Intelligence at the University of Brussels (VUB). He was educated at the University of Antwerp (Belgium) and MIT (USA). His research has covered a broad range of interests in autonomous cognitive robotics and artificial life. Steels currently researches meaning and the origins and learning of language through computational and robotic models. He has developed the framework of evolutionary language games as a way to study the acquisition of concepts and language grounded and situated in experience. The framework allows a systematic study of transactional lestming and could lead to novel learning environments for language acquisition. Steels published and co-edited a dozen books on various topics in artificial intelligence and regularly publishes in top scientific journals, with recent articles in Behavioral and Brain Sciences, Proceedings of the Royal Society, Robotics and Autonomous systems, Artificial Life, a.o. He is a member of the New York Academy of Sciences and a fellow of ECCAI, the European AI organisation.
Part III. Pedagogy and Technology
This page intentionally left blank
8
154
"My dream about the future of education is a grand dream: I look around and I see that we have the Microsoft Office Suite, we have the Corel Office, but I'm waiting for the time that we're going to have a Classroom Suite, a big conceptual structure with parts that could be subdivided and field tested separately, worked out with psychologists, with technicians, with teacher educators, with teachers, with students. The classroom suite should be a broad welcoming set of tools that would be very flexible, that would invite content from a range of different sources, commercial sources, real life sources, children, teachers, and everybody who cares, and that somehow we would be able to start structuring our virtual classroom world the way businesspeople structure their virtual office world." John Sivell
A Learning Zone of One's Own M. Tokoro andL. Steels (Eds.) IOS Press, 2004
155
8
Tools for Embodied Teaching: Celestin Freinet and the Learner-Centered Classroom John Sivell
Despite its origins more than seventy years ago, Freinet Pedagogy's learner-centred approach to educational technology remains both thought-provoking and realistic. As discussion and examples in this chapter illustrate, present-day teachers, striving to create or select teacher- and learner-friendly technologies, still must resolve theoretical and practical issues that are fundamental to the Freinet Pedagogy tradition!
1. The Modern School Movement Celestin Freinet went to Normal School in Nice just before World War I and died at age seventy in 1966. During his lifetime computers obviously had no place in the schoolroom; so, apparently Freinet's educational world was confronted by nothing comparable to the adventures and challenges presented by the modern-day availability of micro-computers and the Internet. In fact, however, Freinet was very much involved with the effective pedagogical exploitation of generally comparable forms of new instructional technology, which in their day presented similar opportunities and difficulties to him and his colleagues. Hence Freinet's carefully theorised but also highly practical advice on such matters continues to be astonishingly relevant to the current debate on electronic classroom resources. Freinet's viewpoint remains valuable precisely because it grew out of lengthy experience with simpler (and less overawing) devices like the home movie camera, the opaque projector, or the child-sized printing press. It probably enabled him to develop a more clear-sighted and critical outlook than is sometimes manifested in today's discussion of computer-based educational technology. This is not to imply that Freinet was sceptical of the potential offered by instructional technology : far from it. He believed in technological innovation and was committed to assuring the best possible results from its use. But he was keenly aware that, for optimal success, attention must be directed towards the human dimension of the interface between inventive classroom equipment and the
156
8. Celestin Freinet and the Learner-Centered Classroom
pupils and teachers who would be using it. This practically-grounded focus on technology in its human context is what makes Freinet's work such an important reference when thinking about computers and the Internet as educational tools. Freinet's educational initiative was formally known as the Ecole Moderne (Modern School) Movement, sometimes with the expression Techniques Freinet (Freinet Techniques) joined to it. Informally, the movement is most often called Pedagogic Freinet (Freinet Pedagogy), but the official words Ecole Moderne were chosen carefully : whatever is 'modern' can maintain that description only by constantly up-dating itself. For nothing that stays the same, no matter how innovative it may have been at the outset, will be 'modern' a few years hence. Similarly, the expression Techniques Freinet - specifically in contrast to Methode Freinet (Freinet Method), which was deliberately rejected - underlines the evolving and innovative nature of the movement (Elise Freinet, 1969:134-5). The goal was to develop an open-ended set of effective teaching techniques, some devised by Freinet himself and others contributed by other members of the movement, so as to create an abundant range of both practically and theoretically satisfactory options among which teachers could select according to their individual judgement regarding the needs of a specific time and place, and to which they could constantly add. Thus, the very nomenclature of the movement suggests an openness to emergent opportunities, and Freinet's work with educational technology vividly reinforces that impression. Rather like today's computers, Freinet's educational technology was strongly associated with middle-class Western work styles and social values. But Freinet taught mainly in rural Provence, both geographically and culturally distant from the busy streets of Paris or other urban centres. Without his close and sensitive observation of the habits, preferences and skills of the children in his classroom, and also without the necessary freedom to take pedagogical decisions based on those insights, Freinet could never have made effective learner-centred use of even such simple (but initially alien) technology as the home movie camera, the opaque projector, or the child-sized printing press. The same may be said a fortiori of today's computer-based educational technology.
2. Work, Tools and the Expert Craftsperson The inter-related ideas of work, tools and the expert craftsperson are key concepts in Freinet Pedagogy, even to the extent that their special importance may be said to distinguish the Modern School Movement from other approaches to teaching and learning, especially in the United States (Lee & Sivell, 2000:114). Freinet gave careful thought to these concepts, and his understanding of them significantly influenced his outlook on educational technology. Probably the easiest way into Freinet Pedagogy as a whole is through the idea of work. As the child of a tiny mountain village, Freinet was well aware of work in the sense of hard, dirty and boring drudgery. However, as real as such drudgery was, he was unwilling to dignify it with the glorious name of work (Freinet, 1994, vol. 1:145-8, or vol. 2:128-9). Rather, Freinet reserved the word work for creative, satisfying effort to accomplish something of value to oneself or one's community. As tiring and painful as it might at times be, genuine work was always rewarding in the long run, and it was synonymous with choice, expertise and pride. He called it work that "edifies" (Freinet,
8. Celestin Freinet and the Learner-Centered Classroom
157
1994:123). Not surprisingly, all of Freinet Pedagogy could be summarised as an evolving set of techniques designed to promote as much real work (and as little pointless drudgery) as possible in the classroom. This ideal is crucial to the learner-centred use of educational technology. For Freinet, genuine work was divisible into two categories: work-play and play-work. Work-play was commonly an adult activity, associated with the professional or recreational activity of men and women in the real world, making things, carrying out projects, or rendering useful services (Freinet, 1994, vol. 1:161-71). As such, a certain proportion of work-play involved the use of instruments or materials and the acceptance of responsibilities beyond what might be appropriate for children, although of course many work-play opportunities were within the grasp of youngsters (e.g. gardening, simple woodworking or, in the classroom, creating a school magazine, to name just a few). But because some kinds of work-play - for instance, handling heavy or dangerous machinery, managing a farm, being a parent, or building a house - would plainly be out of reach for nonadults, Freinet noted that children often engage in equally valuable play-work, the childfriendly equivalent of such activities, carried out on a far smaller scale and frequently with symbolic elements to represent real-life entities (e.g. a stick for a plough, or crickets for cattle; Freinet, 1994, vol. 1:173-5, 192). Both work-play and play-work, Freinet observed, had a place in the classroom, although with a preference for work-play. In order to engage in work, tools are required. When Freinet speaks of tools in an academic setting, he includes what most people would intend by that term: saws, hammers, shovels, and so on, but with two provisos. First, any such adult-type tools should be scaled down to their child-sized equivalents without loss of true functionality. In other words, a saw for young pupils should be small enough for a child's arm to hold and guide, but it should really cut: classroom tools of this sort must make it possible to engage in genuine work, not just make-believe. This respect for the child's right, whenever possible, to engage in work-play obviously comes back to the principle of learner-centred education. Second, the list of potential tools goes far beyond those associated with manual labour; Freinet stresses that recommending classroom work does not mean that schools should become technical training institutes (Freinet, 1994, vol. 1:143). This point is very important. Concretising and demystifying the process of arranging the school environment, Freinet includes in the category of tools any classroom resource, device or strategy that is designed to promote real work. Just as a sharp and sturdy saw makes it easier for a carpenter to do his work, so will pupils (and teachers) best pursue their own academic work when they have an appropriate classroom library, self-correcting work-sheets, scientific equipment (e.g. microscopes or balances), and classroom routines (e.g. individual projects, study contracts, and opportunities to publish their work). The quality of the resulting work depends on a fruitful interaction between skilful workers - both teachers and pupils - and well-adapted tools. Because the category of academic tools is so broad, including strategies as well as equipment of various sorts, not all tools will be forms of educational technology, but everything we recognise as educational technology will be classified as tools. Through this over-arching association of new tools with traditional ones, teachers are empowered to view technology as something meaningful in human terms and not as a kind of bizarre, hyper-complicated and untouchable magic from another world.
158
8. Celestin Freinet and the Learner-Centered Classroom
A tool is considered as something logical and comprehensible that can be judged, selected, rejected or modified in terms of its capacity to satisfy teachers' and pupils' need to engage in productive and satisfying work. Freinet, taking the persona of Mathieu - a wise old shepherd who exchanges opinions about life and work with a married couple of school teachers, in Les Dits de Mathieu - accentuates it this way : "There are ... mechanical tools that turn over the earth for you ... But as for me, ... I like to sift the soil by hand... That's how it is: one and the same job can be a chore or a liberation. It's not a matter of novelty but of edification and richness ... Don't run after novelty; the most sophisticated mechanism becomes boring if it does not serve the fundamental needs of the individual". (Freinet, 1994, vol. 2: 125-6) Mathieu's argument is not against technical innovation, but rather for "edification and richness" that can satisfy "the fundamental needs of the individual." New technology does not in itself guarantee progress in the desired direction, although it is certainly not necessarily retrograde, either. The point is simply that the skilled expert must judge each option in human terms and then decide which will be most effective, case by case. From the Modern School perspective, more innovative technological tools are grouped alongside more traditional paper resources. A representative outline looks as follows: Tools to arouse interest in an activity, either by just raising the topic or by actually providing initial guidance on how to approach it. An example would be the Freinet classroom's 'what's new?' technique, a version of show-and-tell designed to bring experience outside the school into the classroom and hence create interest in further writing or research activities. Tools to help guide research and concerted effort. Examples would include the work library and sets of self-correcting task cards. Tools for integration and extension, either by practice or by transfer to new applications. An example would be the class newspaper (journal scolaire), wherein children's writings can have an extended life as more public documents. Tools for communication. Examples would be classroom printing or the use of desktop publishing or World Wide Web publishing packages to disseminate a class newspaper. Tools to organise and evaluate learning. Examples would be ... individual and collective work schedules... Tools for manual work of various sorts; these are everyday tools, such as hammers, saws, shovels, craft instruments, and so on. (Lee &Sivell, 2000:41) Thus, educational technology represents far more than a translation of paper materials into a more modern-looking form. Each tool is conceptualised as bringing something particular to the overall educational blend, rather than just mimicking current resources in a higher-priced and more elite format. Any such initiative would focus on the technology at the expense of the child: a clear contradiction of learner-centred education. Let's explore a specific example. The most famous instance of educational technology recommended by the Modern School Movement was the child-sized printing press. At the time of this innovation, the basic question was very simple: why go through the trouble and
8. Celestin Freinet and the Learner-Centered Classroom
159
expense of designing and then paying for a child-sized printing press, not to mention the nuisance of teaching pupils how to use it? Today the same challenge could be raised with regard to the cost and inconvenience of purchasing and installing a computer (which in most Freinet Pedagogy classrooms has replaced the printing press). Instead of just assuming that anything new will naturally be better, Freinet addressed the issue head-on. For him, the issue was the same for any tool. How could the child-sized printing press justify itself in terms of promoting genuine work in the classroom? Freinet's assessment was a learnercentred one, directed towards the fulfilment of deep human needs: "Writing is pointless unless one feels a need to use it to communicate one's ideas further than just by speaking ... We have put that motivation into practical effect through the following technique: free writing, copying or printing, illustration, and creation of a school magazine transmitted to parents and exchanged for magazines from other participating schools ". (Freinet, 1994, vol. 2: 35) This explanation makes clear the pedagogical rationale for all classroom work. Real work work-play or play-work - is fundamentally enjoyable. It is inherently rewarding; it offers a "sense of power" and accomplishment (Freinet, 1994, vol. 1:158). For Freinet, the inclination to derive satisfaction from work is a basic human drive, valid for everyone even though the specific kinds of work that each individual appreciates might well vary from person to person and from culture to culture. This approach to educational technology is not applicable only to the particular resources known to Freinet during his own lifetime. The whole point of the Modern School Movement - as already noted above - was to establish an adaptable, up-datable framework for thinking about education that would welcome emerging trends and resources as long as they promoted the basic values of the movement. Freinet never deviated from his belief that techniques and technology should constantly evolve: unlike the promulgators of highly centralised methods meant to be transmitted intact to generation after generation of teachers, Freinet knew that the vigour of the Modern School depended crucially on on-going development based on input from participating teachers: "The paths that [Les Dits de Mathieu] outlines are certainly not always perfectly smooth... It will be up to you... to improve their quality... so as to make them into reliable highways on which good practitioners may in future set out with confidence". (Freinet, 1994, vol. 2:11)
For those practising Freinet Pedagogy today, the principle remains the same. Successful technological innovations will be judged on their merits, and the key child-centred criterion will always be the capacity to support enjoyable and motivating classroom work.
160
8. Celestin Freinet and the Learner-Centered Classroom
3. The Modern School Movement as a Professional World A particularly noteworthy feature of the Modern School Movement's emphasis on individual teachers' freedom (and duty) to make context-specific and child-centred decisions about pedagogical strategies or technologies is the very high level of professionalism required for participation in the movement. From the very beginning just after World War I, Freinet organised professional gatherings and publications to bring teachers from all over France into contact with each other so that they could collaborate as professionals. From the outset the idea was to bring teachers (and therefore pupils) into the process of developing the materials and techniques that they would use. Teachers' knowledge and ownership were emphasised. Nowadays, in an era when international companies provide paper and electronic resources for a global education market, and when at any given moment the top-rated teaching techniques often seem to be generally similar the world over, the goal of increasing individual teachers' involvement in such matters is particularly important. Freinet's first professional magazine, founded in 1926, was L'Imprimerie a I'Ecole (School Printing), named to identify the educational technology that initially united his group. As time went on, given the increasing range of techniques and resources developed by the Modern School Movement, the journal was renamed L'Educateur Proletarien (The Proletarian Educator) and, finally, just before World War II - in a vain effort to deflect criticism from the Vichy regime in France - there was a third name change, to L 'Educateur (The Educator). L 'Educateur is still going strong, with a distribution mainly in France but also abroad. It currently appears ten times yearly, with each issue running to about forty pages. This professional magazine is managed by a team of practising teachers, and all editorial decisions are likewise made by classroom instructors. It is specifically designed to close the gap between theory and practice; occasionally university professors (including myself) may have brief or even lengthy items published in L 'Educateur, but mainly the magazine transmits work written by teachers for teachers. Distinctly exceptional for its rigorous focus on expert classroom teachers as respectable authorities in their own right, this publication promotes the ideal of educators as independent professionals and decisionmakers. In tandem with this professional magazine, Freinet established a teacher owned and operated co-operative to identify, field test, refine and ultimately produce teacher-made classroom materials of all kinds: one of the first great needs was for reliable and affordable child-sized printing presses, but sets of self-correcting materials, graded readers, subjectmatter resource booklets, and films and cassettes gradually joined the list of available materials as teacher needs and the relevant technologies advanced. These, of course, are the necessary tools of the expert classroom craftsperson. The first incarnation of this cooperative was the Cooperative de 1'Enseignement Laic (Public Education Cooperative; abbreviated as the CEL), founded in 1928; subsequently, as part of a wider restructuring, the Publications de I'Ecole Moderne Fran9aise (Publications of the Modern School Movement; abbreviated as the PEMF) was set up to continue this function (1986). This latter body now concentrates on paper materials along with a certain number of CDs, some online materials, online help with using Freinet Pedagogy tools, and a discussion forum (all at www.pemf.fr). Its paper catalogue runs to about 60 pages and is also available online
8. Celestin Freinet and the Learner-Centered Classroom
161
(www.pemf.fr). As with L 'Educateur, the aim is to narrow the theory-practice gap, in this case by providing an opportunity for classroom teachers to be directly involved in the creation and distribution of materials made to their own specifications. The basic principle is always the same: professionalism will be promoted when there is a close and harmonious inter-relationship among work, tools and expertise. Another interesting initiative was La Gerbe (The Bouquet). This began as a collaborative school magazine in 1927, open to contributions from any participating Freinet Pedagogy classroom; classes would submit 100 neatly printed copies of a particularly appealing article from their own classroom newspaper and those copies would be bound with others from other schools to create 100 copies of a composite edition. This publication provided an especially elegant forum for children's work. However, the plan soon changed. By 1932 La Gerbe was handled by a commercial printing house and in time it was replaced by Enfantines (Children's Stories), which specialised in publishing only longer texts that were better suited to this higher-tech medium than to a simple classroom newspaper. Despite the undoubted appeal of higher production values in a commercially printed booklet, Enfantines fulfilled a function that was different from La Gerbe. Its increased sophistication necessarily had the side-effect of distancing the publication process from the pupils themselves. Ordinary classroom newspapers therefore continued to have an important role to play (and they still do, although now they are often on the WWW or else produced on paper with desktop publishing packages rather than on child-sized presses). In this way, we can see that the tools evolved so as to keep in touch with real pupil needs. The selfaware and reflective intellectual system of the Modem School Movement functioned just as it should, and learner-centred values were thus preserved. Turning to the present day, it is interesting that the PEMF does not produce many computer materials: even a CD about the life and work of Freinet, first released in 1996 for the centenary of his birth, is now out of print, and there is just a limited number of other audio CDs. Reading materials and self-paced task cards are the mainstay of the PEMF, along with professional publications. Nonetheless, computers and the Internet have become absolutely central to the Modern School Movement in other ways. There are, as one would expect, a Modern School Movement website (www.freinet.org) and another for the International Federation of Modern School Movements (Federation Internationale des Mouvements d'Ecole Moderne) (www.freinet.org/fimem). Additionally, many Freinet classrooms have their own sites, with electronic magazines and art galleries that extend the evolving tradition begun by La Gerbe and then continued by individual class magazines. But no doubt the liveliest computer application is the Freinet Pedagogy list-serve, which may be joined (whether or not one is an official member of a regional Modern School Movement group) by signing up at the Modern School Movement website. This list-serve is used almost exclusively by practising teachers, with rather rare contributions from graduate students or professors. The range of topics is quite varied, but the large majority fall into five categories: •
requests for help or advice with a variety of more or less technical problems: for instance, software availability or bugs, comparisons of various equipment options (e.g. different digital cameras), or location of desired books, equipment or software
162
8. Celestin Freinet and the Learner-Centered Classroom
•
requests for advice around administrative issues, such as dealing with unsympathetic inspectors or principals, or communicating with parents • discussions of mainline pedagogical questions: for example, approaches to literacy instruction or to the teaching of arithmetical or scientific concepts • debates and advice regarding the great social issues of the day, which currently tend to focus on problems with racism or school violence • and questions about how to get started with Freinet Pedagogy The Freinet Pedagogy list-serve is an excellent example of a computer-based tool. It relies on relatively simple technology, with correspondingly wide accessibility and low cost, in order to facilitate an important work-related activity: discussion of vital professional concerns. The list-serve effectively balances technology with human values; its inherent structural features serve a useful communicative function while imposing few constraints on participants. And it is open-ended. The list-serve is not an artificial model of the world; rather, it offers a convenient mode of contact with the real world, extending the individual's power to reach others beyond what would be available without the tool.
4. Teachers are professionals Direct teacher involvement with the creation and field testing of pedagogical tools is a hallmark of the Modern School Movement. In contrast to the low esteem in which teachermade materials are often held in other circles - frequently discounted as parochial, slap-dash and generally of poor quality - such materials are viewed by Freinet Pedagogy not only as necessary for effective site-specific teaching, but also as a welcome sign of professional commitment and energy. Moreover, criteria for learner-centred teaching and for professional excellence in this respect coincide. The Modern School Movement stresses the identity of teachers as not only the users but also the creators of theoretical knowledge. This is made plain by Article Six of the Modern School Movement Charter: Grassroots experimental research is absolutely essential to our effort to modernise the school through co-operation. This determination to help each teacher become not only a practitioner but also a researcher explains, as well, the movement's strict insistence that other groups must not be permitted to infringe on teachers' rightful professional domain, as specified by Article Seven: Teachers of the Modern School have sole responsibility for the direction and use of their co-operative efforts. With respect to pedagogical tools, including computer-based educational technology, the effect of this Article is to underline the necessity for resources that empower teachers to select and develop what they judge to be best for the educational progress of their pupils, rather than simply imposing what some outside expert feels would be a generally good idea. This outlook will obviously have important implications for desirable design features in computer-based resources. As a result, before moving on to look at some practical examples of computer applications, it is worthwhile at least briefly to contextualise the Freinet Pedagogy viewpoint within the broader literature on teacher professionalism.
8. Celestin Freinet and the Learner-Centered Classroom
163
Above all, the Modern School Movement's forceful assertion of teachers' rights and responsibilities as combined theoreticians and practitioners is consistent with what has now become a widespread understanding that there are grounds for questioning the older view that "researchers' knowledge is the best foundation upon which to build a professional knowledge base [for teachers] because of its generalisable and trustworthy (scientific) character", since : " (...) the knowledge teachers use is of a very different kind than [what is] usually produced by educational researchers ... Called 'craft' knowledge by some, it is characterised more by its concreteness and contextual richness than by its generalisability and context independence. From this point of view, bridging the gap between traditional research knowledge and teachers' practice is an inherently difficult, perhaps intractable, problem". (Hiebert et al., 2002: 3) Indeed, over the past twenty years or so, a vast educational literature has grown up concerning this aspect of teachers' professional development, and the themes stressed by the Modern School Movement are entirely in tune with much of that work. The Ecole Moderne's approach is not so much to bridge the gap as to eliminate it by encouraging teachers to become small-scale researchers as well as practitioners. Thus, it is appropriate to associate Freinet Pedagogy with the Action Research tradition, in which the practical "action" of teaching is viewed as "a form of research" (Winter, 1987: 4). From this perspective, teachers deserve to be recognised as professionals whose needs, knowledge and aspirations are directly relevant to the development and implementation of effective pedagogical tools.
5. Teacher- and learner-friendly technologies Although, like most teachers, I am not a programmer, some of my materials-development activities have led to the creation of a number of educational tools that rely on computers and the Internet. Rather than advancing them as ideal models for specific use or imitation, the rest of this chapter demonstrates some ways in which teachers can assert ownership of situation-specific educational technology applications and, above all, highlights the features of openness and flexibility that - as suggested by the Freinet Pedagogy approach -characterise teacher- and learner-friendly technologies. Customising PC-Write An interesting example of teacher- and learner-friendly technology is a thorough customisation that I made of the now-discontinued word-processing program, PC-Write, described in full detail with complete coding in Sivell, 1990. PC-Write was designed in an open format, so that almost all features could readily be modified or controlled using very simple, high-level coding techniques. My purpose for the customisation was to devise a version of the word-processor specifically adapted to the needs of university-level English as a Second Language (ESL) learners — and their teachers
164
8. Celestin Freinet and the Learner-Centered Classroom
— in the late 'eighties and early 'nineties. In those days, computer literacy was less widespread than at present; so, one of my goals was to turn off most of the complex (and rarely used) word-processing features of the program, in order to avoid unexpected and daunting surprises resulting from unintended keystrokes. The very flexible design of PCWrite made it easy to limit text-formatting options to such mainline choices as italics, bold or underline, with many of the more arcane selections - if chosen by accident - simply beeping. Remember, this project took place at a time when that type of simplicity was requested and much appreciated by both teachers and students: it allowed learners to focus on the specifically linguistic aspects of creating a text, without technical distractions, which was what everyone wanted. Additionally, all of the Help screens - accessible in PC-Write for total customisation - were re-written with ESL students' linguistics needs in mind. Some became simple grammarreference files but, more interestingly, others were created as interactive rhetorical resources. In this connection, it should be noted that students learning English face the challenge of developing a grasp of the discourse markers that typically structure English expository prose: for example, moreover, in addition, or also for additive relations, or by contrast, but, or on the other hand for oppositions, and so on. Each of the new rhetorical Help screens focussed on a particular discourse structure - addition, opposition, causality, chronology, etc. - and offered a range of lexical options, with explanations as necessary. Also, through a simple system of macros (handily available in PC-Write) students could select a rhetorical option and it would conveniently 'jump' onto the page for them. And the content for these Help screens was created and revised in collaboration with a succession of teachers and learners. This customisation of PC-Write is instructive for two reasons. On the positive side, PCWrite represents a reasonably good example of commercially-available software that (at the time) was sophisticated enough to be pleasing in itself, but also flexible enough to permit creation of a situation-specific pedagogical tool. In other words, the software offered a highly effective balance between expert programming in the basic program itself, along with scope for local teacher expertise in the particular customisation. But on the negative side, the ultimate fate of the initiative was also interesting. After several successful years, the PC-Write site licence was abruptly discontinued; it was replaced by a site licence for WordPerfect, a more technically advanced (or at least better marketed) option that totally lacked the flexibility of PC-Write. Consequently, all of the pedagogical advantages were lost, in favour of a program imposing a vast range of confusing secretarial features with no value whatsoever for language learners. Indeed, before beginning to use this impoverished tool, teachers had to devote very considerable classroom time to teaching students how to avoid its pitfalls. Thus, the change of software made no educational sense, and the new licence was more expensive, too. Moreover, this very substantive pedagogical decision was represented as a merely technical matter, which could best be determined without consultation with any of the teachers using the lab. In fact, there was a distinct impression that the change was
8. Celestin Freinet and the Learner-Centered Classroom
165
dictated so as to oblige teachers - those backward techno-peasantsf- to adjust to software which specialists in another, quite separate field felt qualified to impose on them: a perfect illustration of how to undermine teacher professionalism at the same time as depriving students of the benefits available from taking educational expertise into account. Fortunately, computer literacy among in-coming ESL students rose steadily in the early nineties, so that after a few years the pedagogically distracting complexities of WordPerfect became less problematic, but the problem of non-adaptability remained, with teachers and students never regaining the educational advantages of the software that had been abandoned. Class-Oriented WWW Sites Reflection on issues around teacher control over the appropriateness of educational technologies is also applicable to various types of WWW site that may either be visited or actually created. At the most technologically advanced end of the spectrum, there are numerous sites specifically designed for educational use and accessible world-wide. For instance, in my own field of English as a Second Language (ESL), we find such sites as Dave's ESL Cafe (www.eslcafe.com) or Randall's ESL Cyber Listening Lab (www.esllab.com), as well as collections of ESL materials at sites like ESL Gold (www.eslgold.com), the National Centre for (Adult) ESL Literacy Education (www.cal.org/ncle), Lesson Plans and Resources for ESL, Bilingual and Foreign Language Teachers (www.csun.edu/~hcedu013/eslindex.html), or The Internet TESL Journal (http://a4esl.org). Similar resources exist for other disciplines as well, of course. WWW sites like these offer a highly professional look and feel, which can be important learners accustomed to sophisticated computer games and other software. Furthermore, selecting appropriate resources is reasonably easy when sites propose a wide range of options, which is often the case. When materials presented at such sites suit a particular teacher's and class's needs, they are not merely convenient; they are fun, visually and aurally attractive, and technically sophisticated. With respect to guaranteeing situation-specificity, however, teachers are mainly restricted to one tactic: careful selection of available items when and where applicable. Thus, teacher decision-making is circumscribed by the limits of what distant webmasters have decided to provide. Because these materials are not teacher-originated, they are difficult or impossible for teachers to adapt closely to the needs of their students. They are less than ideal educational tools in Freinet Pedagogy terms. Designed at a distance from the classrooms where they will be accessed, such sites have potential pedagogical weaknesses along with their technological strengths. In order to bring control of WWW materials closer to individual classrooms, two techniques are available. The first is the exploitation of what may be termed realia: resources that are suitable for pedagogical use in a certain context, but that were not originally designed for educational purposes. This concept antedates the Web but is highly * This denigration of teachers is reminiscent of the advice of Skinner (1968; just a year after the full edition of Freinet's Dits de Mathieu!) that "many [teachers]... are unaware that technical help is available, and many are afraid of it when it is pointed out" (p. 259).
166
8. Celestin Freinet and the Learner-Centered Classroom
relevant to present-day pedagogical use of the WWW. When using WWW realia, constraints on free selection are almost entirely lifted. For example, when ESL students do a project involving a search of the Web, or when teachers select WWW pages for use in their classes, these realia are experienced as localised and individualised learning tools on account of the extreme freedom and precision with which they can be chosen. Consequently, mining the Web for educationally useful realia is one way to devise pedagogical tools over which teachers and learners have a very high degree of control. Used in this way, WWW technology enables ESL learners to engage in personally meaningful communication. A second technique, extending the simple use of WWW realia, entails actually setting up a class WWW site. A simple illustration of this widely-practised technique is a site that I created with a class of Thai ESL learners in the spring of 2001 (at the time of writing, still posted by student request at www.brocku.ca/appliedlang/sivell/srin/index.html). Visitors will find, among other items, a number of pictures and texts (all realia) about dolphins - a topic chosen by the students - which individual learners located on the Web and assembled to their own satisfaction as a project involving the practical use of English. Additionally, this site includes materials created by the Thai students themselves: for example, concrete poems and haikus, as well as group photos and bios. All such WWW activities are effective pedagogical tools in Ecole Moderne terms: starting from individual choice and creativity, they reach out to the wider world in ways that genuinely interest students, and they promote work-play with an authentic outcome - a personal collection of resources on the Web that others will enjoy. To quite an extent, the nature of the resulting pedagogical tools will depend on the method of creation that selected. The simplest option is for teachers to access one of the readystructured but content-free templates for educational sites on offer without charge on the WWW (for instance, the TeacherWeb site, at http://teacherweb.com). Alternatively, it is possible to start more or less from scratch by using site-building software like Microsoft's Front Page Express, by choosing a specifically education-oriented environment like WebCT, which we have adopted at Brock University (actual course sites are password protected, but information and examples are available at www.webct.com), or by relying on the HTML-conversion possibilities now available within the major word processors. Each of these alternatives enables teachers to respond to student preferences and needs by building individualised WWW sites designed just for them, although working from an online template will allow for rather less flexibility than starting from scratch. The best teacher-designed sites are extremely appealing: for example, the excellent classroom sites in France, Belgium and Quebec listed at http://freinet.org/creactif. Moreover, learners can use site-building software or word processors with HTML capacity to construct their own sites. But the key element in all cases is harmonious interaction between the technology and its users, both teachers and learners. The highly customisable features of PC-Write offered a very desirable balance between teacher- or learner-friendliness and what was (for the time) high technological sophistication. As a result, that program exemplified the ease of use, the potential for individualisation, and the capacity to promote the satisfying work-play commended by Freinet Pedagogy in effective tools. A similar balance is necessary with respect to today's WWW technology for pedagogical use. Whatever the technology in question - Web
8. Celestin Freinet and the Learner-Centered Classroom
167
browsers, realia or classroom-directed WWW sites, or website-creation packages of various kinds - success depends on flexible resources promoting a partnership between the original computer programmer and the classroom teacher. Although much or all of the initial programming effort may occur outside the educational realm and far from the individual classroom, good pedagogical tools will result when teachers and learners are free to employ the technology in ways that represent their own interests and ambitions. Such technology supports teachers' professional independence by facilitating the development of teachermade tools with both high production values and excellent applicability to individual learning needs.
6. Conclusions and wish-list In one sense the entire history of educational technology, from its simplest beginnings right down to its present-day computer-based complexities, may be viewed as an on-going effort to bridge the chasm between two solitudes: the non-classroom expertise of technical specialists, and the pedagogical insights of excellent teachers. Stepping back a little, about the same dichotomy may be seen in the broader debate around teacher professionalism, where the contributions of researchers and theorists tend to be in contrast or even in conflict with those of classroom practitioners. When poles such as these remain separate, they almost inevitably become competitive, to the real benefit of neither side. In terms of educational theory and practice - including the specific case of educational technology - the necessary bridge between the two solitudes is the concept of praxis: the skilled and highly effective practical action of reflective teachers who are deeply conscious of both the goals and the theoretical foundations of their work (McNiff et al., 1997). Praxis, as Freinet Pedagogy would agree, is the bedrock on which teacher professionalism must be founded. And excellent pedagogical tools are essential not just for the exercise but also for the development of praxis. A fruitful partnership between technical and pedagogical forms of expertise is the defining characteristic of praxis and, equally, it is a key feature of effective educational tools as envisaged by Freinet Pedagogy. Praxis operates where theory and practice meet, where outof-classroom skills and inside-classroom expertise unite. Therefore, my wish is for computer-assisted educational technology that makes it easier for teachers to develop and exercise the praxis of respected professionals by creating high-quality teacher-made materials. Such a model in fact is not particularly unusual. In the world of office management, for example, we are accustomed to using integrated suites of programs that work together so as to facilitate the operations of individual companies. No two companies will be identical, but business software suites allow for this diversity by offering broadly applicable options that skilled professionals will customise in sophisticated ways. The highly complex programming in such packages, which no end-user could hope to equal, has the specific goal of providing as much technical support for individual users as possible, while still imposing little or nothing in terms of the form or content of the work those users may want to produce. Naturally, imperfections do remain in each of these suites but they are constantly evolving and improving; no user would deny that they have enormously increased in effectiveness over the years. At times, of course, we still resort to extremely primitive, one-time-only quick fixes, like jotting a note on a yellow sticky, because that happens to be the best solution at the moment. At other times, most organisations engage
168
8. Celestin Freinet and the Learner-Centered Classroom
specialists to do exceptionally complicated work that cannot reasonably be carried out with the normal suite of tools. But the heart of any modern business operation is an integrated package of office programs that make the place function, because it facilitates the kind of environment in which team members can most effectively learn and use professional skills for the benefit of the enterprise. The same concept applies to the classroom. Many different forms of educational technology have a place and no doubt will always do so. It would be short-sighted to denigrate any particular option because, used wisely, all can contribute in one way or another. But the heart of teaching is praxis, so that the pedagogical tools at the real centre of educational activity will inevitably be those which promote individual teachers' creativity, independence and expertise, which will benefit learners by making the most appropriate materials available to them. Regrettably, unlike the world of business, the world of education does not yet have a wonderful range of what might be termed classroom suites to choose from. In fact, it appears that we have none at all which genuinely play that complete role in the manner that the best office packages do. We do have a number of more focussed software packages, which are constantly developing in very encouraging ways and which address one part of the overall pedagogical environment quite well: a good example would be WebCT, a reasonably flexible environment that conveniently facilitates several aspects of web-based instruction. However, a good office package has more than - say - a word processor; it boasts at least a database, a spreadsheet, and a computer-presentation component, too. A good educational suite would need comparable versatility, and so far there seems to be no entry in the field. So, my wish is for an ambitious set of integrated programs to support the creation of diverse, classroom-specific pedagogical tools. Given the importance of the WWW, it is probable that software for web-based instruction might be the logical starting-point for such a suite. A reasonable extension of that kernel would surely take the form of a thoroughly flexible and customisable exercise-authoring program, far surpassing the either very limited, or else very user-unfriendly options that now exist. And to make the technology more adaptable, it would be important to have seamless compatibility with the word processor, scanner software, and database of a major office package. Such a coherent suite of educational software could do much to facilitate the exercise of professional praxis among teachers, which in turn would promote the development of situated and learner-centred educational environments. Bringing this project to fruition would doubtless require a lengthy commitment from a team of programmers, psychologists and educators, but the potential outcome would be a giant step forward in the crucial work of helping individual teachers access the power of modern educational technology in original and effective ways.
References Freeman, Donald & Karen Johnson (1998) Reconceptualizing the knowledge-base of language teacher education. TESOL Quarterly, XXXII, iii: 397-417.
8. Celestin Freinet and the Learner-Centered Classroom
169
Freinet, Celestin (1994) (Euvres Pedagogiques. 1 vols. Paris: Editions du Seuil. Contains among others L'Education du Travail (original edition, 1964), vol. 1:23-322; L'Ecole Moderne Francaise (original edition, 1944), vol. 2:7-97; and Les Dits de Mathieu (original edition, 1967), vol. 2:99-203. Freinet, Elise. (1969) Naissance d'une Pedagogic Populaire. Paris: Maspero. Hiebert, James, Ronald Gallimore & James Stigler (2002) A knowledge base for the teaching profession: what would it look like and how can we get one? Educational Researcher XXXI, v: 3-15. Lee, William & John Sivell (2000) French Elementary Education and the Ecole Moderne. Bloomington, 1ND: Phi Delta Kappa Educational Foundation. McNiff, Jean, Pamela Lomax & Jack Whitehead (1997) You and Your Action Research Project. London: Hyde. Oja, Sharon. & Lisa Smulyan (1989) Collaborative Action Research: A Developmental Process. London: Palmer Press. Sivell, John (1990) Pedagogical exploitation of macro-commands in PC-Write. In Louise Craven, Dana Paramskas & Roberta Sinyor (Eds.) CALL: Selected Articles and Reports. La Jolla, CA: Athelstan. Skinner, B. F. (1968) The Technology of Teaching. Englewood Cliffs, NJ: Prentice-Hall. Winter, Richard (1987) Action-Research and the Nature of Social Inquiry. Aldershot UK: Avebury.
Suggested Readings Celestin Freinet (1994) Oeuvre Pedagogique. Ed. M. Freinet. 2 Volumes. Paris: Seuil. W. Lee & J. Sivell (2000) French Elementary Education of the Ecole Moderne, Bloomington, Indiana: Phi Delta Kappa Educational Foundation.. John Sivell (Ed.) (1994) Freinet Pedagogy: Theory and Practice. Lewiston, N.Y. : EdwinMellen Press. Jim Cummins & Dennis Sayers (1995) Brave New Schools : Challenging Cultural Literacy Through Global Learning Networks. Toronto: OISE Press.
170
8. Celestin Freinet and the Learner-Centered Classroom
Appendix: Charter of the Modern School Movement (http://freinet.org/pef/charte.htm) • • • • • • • • •
•
Education means fulfilment and empowerment, not accumulation of knowledge, nor unquestioning obedience, nor memorisation. We oppose all indoctrination. We reject the illusion of self-sufficient education, removed from the great historical and political developments of which it is a part. Tomorrow's school will be the school of work. The school should be child-centred. It is the child who, with our assistance, constructs his or her own personality. Grassroots experimental research is absolutely essential for our effort to modernise the school through co-operation. Teachers of the Ecole Moderne have sole responsibility for the direction and use of their co-operative efforts. Our Ecole Moderne Movement is eager to maintain friendly, collaborative relations with all organisations sharing its goals. Our relations with the administration: ... we are prepared to share our experience... but we are determined to maintain our independence... in keeping with the demands of the co-operative action of our movement. By its very nature, Freinet Pedagogy is international.
JOHN SIVELL obtained a B.A., English Language and Literature, (Trinity College, University of Toronto), a Diploma in English Studies, (King's College, University of Cambridge), an M.Ed., in TESL and Educational Technology, (University of Wales Institute of Science and Technology, Cardiff, Wales) and a Ph.D., English Literature, (University of East Anglia, Norwich, England). As a teacher in the TESL- Teaching English as a Second Language - program, he tries to establish an equilibrium between scholarly and theoretical work, on the one hand, and practical applications of that same expertise, on the other. Sivell is a Faculty resource person for the Reading stream within Brock University's Intensive English Language Program, and has written or co-written a number of classroom ESL Reading textbooks published by Full Blast Productions, for instance (1992) From Near and Far, (1993) Jigsaw Activities for Reading and Writing, (1995) Canada from Eh to Zed: People, and a number of others.
9
172
"I'd like to pick up the idea that learning should be fun since this is what came out of the first workshop. We started to unpack this concept of what we mean by fun, how it related to other concepts as joy and play. And I still think that we need to unpack it more in its context and in its grounding. Sometimes learning shouldn't be fun!! It should be hard. There is no question about it. I'd like to see that being explored more and in particular in terms of the kind of technologies that we are developing. I'm concerned that there is too much emphasis being put on supporting virtual learning, virtual communication... We are moving into the computer too much, and I still think we need to learn how we interact with each other and physical artifacts and not forget that that is an important skill." Yvonne Rogers
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
173
9.
Playing and Learning in Digitally-Augmented Physical Worlds Yvonne Rogers and Sara Price
The arrival of pervasive technologies and ubiquitous computing offers many opportunities for designing learning experiences with a difference. Novel forms of interactions can be developed that exploit the 'physical' and the 'digital' worlds in a diversity of ways. In particular, there is much scope for bringing physicality and embodiment into computerbased learning experiences promoting different forms of engagement, reflection and interpretation. In this chapter we present our approach - technology-mediated learning describing how we have developed a range of digitally-augmented physical spaces to support reflection and interpretation in children. Our claim is that more active forms of learning can be facilitated through engaging children in these kinds of spaces than is currently possible when interacting with educational desktop software. To illustrate our approach we describe three studies where we developed and evaluated a range of novel technology-mediated learning experiences.
1. Embodied learning interactions It is well known that interacting with one another and the environment is an integral part of learning and development. In particular the constructivism school of thought, originating from Piaget's and Papert's work, views knowledge as being created by learners in their own minds through interaction with other people and the world around them. Inspired by this philosophy, a variety of computer-based activities have been developed, that support learning through interactivity. Most well known are the various Logo programming environments and the genre of micro-worlds. One of the most popular and successful forms of computer-based interactivity involved programming a toy turtle to physically move around the floor by typing commands at the computer. However, with the advent of graphical user interfaces, many computer-based learning activities began to take place exclusively at the computer display. Interactions with the environment and physical toys,
174
9. Digitally-Augmented Physical Worlds
like the Logo turtle, were superseded by interactions with virtual representations in the computer. The physical Logo turtle became a virtual one, that when interacted with at the computer screen, resulted in shapes, designs, and pictures being drawn on the computer screen. The mixed couplings between digital actions and physical effects - epitomized by programming the physical turtle to move around the floor - were replaced by same couplings between digital actions and digital effects. One of the downsides of this shift into the digital sphere is that it means that children's attention becomes focused primarily at making things happen at the computer screen. While, arguably, just as stimulating and engaging as making a turtle move around a physical environment, this form of displaybased learning is more like a closed system; where the children's digital interactions are essentially divorced from their interactions with the physical world and each other. Some researchers recognised the constraining effect GUI interactions can have on the learning experience and started building computational toys that enabled children to program them to create physical behaviours and actions. Most notable was the 'programmable bricks', 'thinking tags' and 'crickets' developed at MIT that can be programmed using a version of the logo programming language. Children are encouraged to experiment and test their hypotheses about the physical world by designing, programming and engineering the bricks to behave in creative ways (e.g., Resnick, 1996; 1998; 2000). Embedding physical artefacts with digital information has also been found to provide a number of opportunities for reasoning about the world through discovery and participation (e.g., Soloway, Guzdial and Hay 1994; Tapscott 1998; Forrester and Jantzie 2000). Furthermore, the digital world of information has begun to be coupled with familiar and novel arrangements of electronically embedded physical objects (e.g., Ishii, 1998; Underkoffler and Ishii, 1998; Piper and Ishii 2002; Annay, 2001). Everyday actions and artefacts are physically manipulated to make changes in an associated digital world, capitalising on people's familiarity with their way of interacting in the physical world. An example is the Illuminating Light system that was designed to allow students to learn about lasers and holography by manipulating models of optical elements using physical objects that were coupled with various forms of digital augmentation in the form of light patterns (Underkoffler and Ishii, 1998). In a similar vein, we have been involved in a programme of research that is exploring how 'physicality' and 'embodiment' can be brought into computer-based learning. By physicality is meant the movement of the body (e.g. gestures, posture changes) through an environment and manipulation of real world objects and artefacts. By embodiment is meant the way we encounter directly the physical, personal and social world. As mentioned earlier, both are normally an integral part of childhood development. We call our approach technology-mediated learning - one that considers how the new forms of pervasive technologies and mobile technologies that have begun to materialise during the last few years can be used, designed and assembled to augment everyday physical activities (e.g. walking, moving objects, manipulating things) using a diversity of digital representations (e.g. images, animations, sounds). Our approach is to find ways of focussing the learner's attention on their everyday activities with the physical world, and where we see the role of technology as one of augmenting, extending and amplifying these activities at hand. We argue that a main benefit of combining physical interactions with digital augmentation is the extent to which it promotes reflection in children (Rogers et al. 2002; Scaife, 2002), allowing them to combine and recombine the known and familiar in new and unfamiliar ways (see also Hoyles and Noss 1999). Furthermore, the combining of familiarity with
9. Digitally-Augmented Physical Worlds
175
unfamiliarity has been found to promote reflection in children that is critical in stimulating awareness and enhancing learning (e.g., Ackerman 1996; Piaget and Inhelder 1967). It is important to stress that we are not arguing against learning through desktop computerbased interactions, since these have now become central to modern day education (e.g. word processing, web browsing, problem-solving). What we are saying, though, is that this kind of computer-based learning has become a learning activity in itself, and is largely divorced from the everyday physical and social world that children engage with. Moreover, the learning benefits of education software packages can sometimes be unclear (e.g., Large et. al, 1994; Rogers and Scaife, 1997; Hoogeveen, 1997). A challenge facing us, therefore, is to consider how the new technologies can be exploited differently to provide more embodied learning interactions. In so doing, we claim that 'active' forms of learning can be promoted, where children are encouraged through their technology-based embodied interactions with the physical world to be creative and imaginative, to think for themselves and reflect on what is happening around them and the consequences of their actions.
2. Digital Augmentation How do we realise the potential of the new generation of pervasive technologies to stretch children's minds and, in so doing, support and extend active learning? In particular, how do we bring physicality into 'brains-on' learning? By this we mean combining physical movements, such as, manipulation with real world objects, gestures, and bodily posture changes, with higher order cognitive activities, like thinking, reasoning and reflecting. Our rationale for bringing together body and brain in this sense is based on fundamental developmental theory; effective learning takes place when meaning is taken from experience with the world, when children through their own experience discover what is "going on in their own heads" (Bruner; 1973, p.72). Physical engagement with something creates an involvement and activity in learning, which in turn, affects the cognitive and social interaction of the learner, in terms of their attention, inquisitiveness and reflection. Our approach is to consider how physical actions can become a more integral part of cognitive and social activities, through using digital augmentation, and in so doing, promoting learning activities that provoke children to reflect. Digital augmentation involves children interacting with the physical world in a variety of ways causing digital events to occur. We have been experimenting with a variety of new technologies, namely pervasive environments, handheld devices and wireless networking to create a range of 'digitallyaugmented physical spaces'. To interact with and in digitally-augmented physical spaces, involves children moving around them and/or exploring them, using a variety of tools and body movements. They can be created to be both indoors and outdoors - the key is to develop such spaces that exploit physicality in interaction, and which trigger various digital representations at appropriate and pertinent times, points and places, stimulating the children to decide what to do next in the learning activity, while also encouraging them to think and reflect. Thus, children are encouraged to drive their own learning and understanding, through the novel physicaldigital couplings. For this to work, however, requires designing open-ended and provocative learning or play activities, in the sense that it is not necessarily obvious to the children as to what to do or what to expect. Indeed, part of the active learning challenge is
176
9. Digitally-Augmented Physical Worlds
for them to have to discover for themselves what to do and what is possible through their physical interactions. Designing technology mediated active learning In this chapter, we describe in more detail what we mean by digitally-augmented physical spaces, showing how physicality can promote active learning and in particular, reflection in play and learning. We present a framework that we used to guide the design of our technology-mediated learning, followed by an account of three different learning experiences we developed. The first provides children with various ways of exploring and experimenting with mixing colour, for example, paint or light can be mixed to make different colour combinations using physical, or digital tools or a combination of both. The second is an adventure game, where children have to collaboratively discover as much as they can about a virtual imaginary creature, through interacting with various devices and aspects of the environments. The third provides a woodland field trip for children learning about ecology, enhanced by combining physical and digital information. A critical aspect of learning is for children to build their own understanding from what they already know, and taking from the world, new information with which they can expand their current knowledge, and reach more complex levels of understanding in ways which make sense to them. This construction of knowledge is based on an interaction between subject and object, through perpetual exchanges of thought and different kinds of experimental interaction (Piaget, cited in Holzer 1998). In addition, cultural background and social interaction is an inherent part of this knowledge construction, as a person's understanding and knowledge is grounded in their experience of the world, and is developed through social interaction and mediation of that grounding (Vygotsky, 1978). The concept of active learning, therefore, has several aspects, including, experienced-based learning; actively engaging in meaningful activities in the real world; collaboration, engaging in meaningful activity, taking part by talking about what is being learnt, and making it an active social interaction (Chickering and Gamson, 1987), and reflection. To make the learning process itself an active process, learners need to experience, construct, test, and revise knowledge (Thompson and Jorgensen 1989). They need also to interpret and transform it (Schomberg 1986), understand the content in context and create a personal meaning (Peterson et al, 1996). This involves reflecting in various ways in their activities. Furthermore, enabling explorative play within the 'real' world stimulates independent discovery, and in so doing, facilitates both the acquisition of information about, and experience with, the environment. In addition, exploration of different combinations of information can enhance creativity (e.g., Bruner 1985 cited in Clements 1995; Dansky and Silverman 1973). The advent of ubiquitous computing and pervasive environments offers great scope for supporting the various aspects of active learning. In particular, it offers opportunities for supporting 'physicality' in learning, enabling a variety of interactions with the real world. In so doing, such interactions can potentially enable a different kind of engagement with objects and the environment from those constrained by desktop machines. Physical interaction can, not only take place with the environment itself but also with the tools available in that environment. In turn, these tools can mediate interaction with the physical environment, changing the kinds of learning interactions that can take place.
9. Digitally-Augmented Physical Worlds
111
Thus, pervasive technologies have the potential to enable a more active, physical engagement with the environment, while also providing opportunities to digitally augment interaction in novel ways. One benefit of doing so is to provide different ways of thinking about the world than interacting solely with digital representations or solely with the physical world. Digital augmentation also provides the facility to convey information or experiences not possible in the physical world, for example, making the invisible visible. In turn, this can provide opportunities to encourage or even enhance further exploration, discovery, reflection and collaboration.
3. Conceptualising digitally-augmented physical spaces Designing computer-based learning and playing experiences to be physical requires us initially to find out what kinds of interactions children carry out in their physical environment, what kind of actions they perform and how they interact with different kinds of artefacts. This provides us with a basis from which to consider what kinds of physical actions and interactions maybe most amenable and appropriate for digital augmentation. It can also provide us with a starting point by which to conceptually understand how physicaldigital couplings can aid exploration and reflection in ways that are different from desktop software interactions. There are many kinds of physical actions and interactions and we do not wish to provide an exhaustive list here. Instead, we broadly categorise physicality into three main types, for the purpose of investigating how to support them digitally: (i) interaction with physical tools, (ii) physical movements, and (iii) combining artefacts with each other. (i) Interaction with physical tools These include actions like drawing or writing with a pen, chalk, or paintbrush and are highly familiar actions that children learn to do to and which 'externalise1 their cognition (Scaife and Rogers, 1996). A key question is how extending children's cognitive actions with physical tools can be further augmented through being coupled with new forms of digital content. A child may also use various tools (e.g., hammers, spades, wands) to cause certain effects to take place, that in turn can be coupled with digital content. (ii) Physical movements There are various kinds of movement children readily explore and engage in as part of their development. These include crawling, rolling around, dancing, walking and gesturing. We are interested in how these can be used to pervasively trigger various digital events (e.g., sounds, animations) in a way that becomes an integral part of a learning experience. (iii) Combining artefacts with each other A common kind of physical activity is placing one object besides, inside or on top of another (e.g., bricks) and inserting one inside another. A key question is how these highly familiar activities can be augmented with digital content in unexpected ways. We report here on three physically-augmented digital environments we have developed to explore how the three broad categories of physicality can be brought into playing and learning experiences. The first is the Chromarium, which explores primarily the first and third categories of physicality, augmenting the highly familiar activity of mixing colour with various physical-digital couplings. The second is the Hunting of the Snark, which focuses mainly on the first and second categories of physicality, investigating novel couplings between familiar actions and unfamiliar digital responses. The third is Ambient
178
9. Digitally-Augmented Physical Worlds
Wood, which explores all three categories of physicality, using a range of digital-physical couplings to support learning and reflection.
4. The Chromarium The Chromarium was developed as a mixed reality environment to enable young children, aged between 4-7 years, to explore different ways of mixing colour, using various physical and digital actions and combinations of the two (see Rogers et al., 2003). In addition to using physical paints for mixing colour, we transformed the physical activity of colour mixing using a variety of alternative physical-digital couplings, varying in their degree of physicality. One coupling with high physicality was mixing colour through combining two coloured blocks (see Figure 1). Each block has six sides, each with a different colour displayed on it. Each face is embedded with an RF tag; when one block is faced onto a table it is read by an RF tag reader triggering a digital animation of the colour of the face currently showing on the top of the block to appear on an adjacent display. When both blocks are placed on the table, an animation is triggered showing the two colours of the sides that are currently face-up being blended on the display. Thus physically turning the blocks and placing them onto the table results in another digital colour being mixed. An important difference from mixing colour using wet paints, is that colours can be rapidly and easily mixed and unmixed using the physical-digital coupling afforded by the blocks, which is not possible with wet paint.
Figure 1. Mixing colours using RFID tagged blocks
A 'low' physical condition was included; where colour mixing is done using a screen-based digital painting or digital lighting application (see Figure 2). Different coloured disks are moved onto each other, by using a touch screen, to mix the colours. The blended colour that results is shown appearing on the screen overlaying the combined digital disks. De-mixing and re-mixing is also relatively easy to achieve by dragging a disk away and replacing it with another one.
9. Digitally-Augmented Physical Worlds
179
Figure 2. Screen-based colour mixing
Another form of physical-digital coupling that was provided was to mix colours by using digital paint, resulting in a physical action (see Figure 3). Here the physicality is instantiated as an effect of a digital action.
Figure 3. Digital action giving physical effect
The digital-physical coupling is achieved through a digital image of a windmill, having alternative coloured sails being combined with a physical counterpart. When the digital sails are turned by touching them on the screen, the physical windmill sails are simultaneously rotated, creating the illusion of the physical sails changing colour (e.g., when yellow and blue sails are mixed in this way the illusion of green appears). A further condition was the use of torches, that had coloured lights, to mix colours. Owing to the different properties of light and paint, the result of mixing certain colours can be different. For example, mixing red, green and yellow results in the colour white appearing on a screen, when using lights, and brown when using paints.
180
9. Digitally-Augmented Physical Worlds
Type
Physicality
Wet paint mixing Torch light mixing Digital painting Digital lights Blocks and animation Windmill
High High Low Low High Medium
Coupling (causeeffect) Physical-physical Physical-digital Digital-digital Digital-digital Physical-digital Digital-physical
Reversable
Mixing
No Yes Yes Yes Yes No
High Low High High High Low
Table 1. Types of colour mixing in terms of different levels of physicality, couplings between cause and effect , extent of reversibility of actions and scope of possible exploratory actions. Thus, a number of digital and physical couplings were arranged to enable children to mix colour in novel ways, varying in degree of physicality either in the action or the resultant effect (see Table 1). The ways of mixing colour also varies along other dimensions, including the nature of the coupling, the ability to reverse actions easily and the scope of mixing afforded by the set-up. For example, the constraints of the digital-physical set-up meant that only limited colours can be mixed, based on the colours of the sails, in the physical windmill connected to the digital windmill display. One of the key questions we were interested in was whether high physicality has a different effect on the children's level and kind of exploration and reflection, compared with low physicality. As predicted, a main finding was that pairs of children explored, experimented and reflected much more with the physical blocks than they did with the equivalent digital representations. They rapidly combined and recombined colour, appreciating the reversible and immediate feedback provided by this physical-digital coupling. The children were also quite inventive and collaborative, sometimes piling the blocks on top of each other on the table and pressing them down hard to see if such actions changed the way the colours were mixed from the default setting. They also placed them side-by-side in the air to see if this had a different effect. A couple of the children placed their faces and hands onto the table to see if these would cause an effect and their colour to be mixed. With the digital-physical coupling (the windmill condition) the children marvelled at the physical effect they created and spontaneously came up with everyday theories of how this was caused. In contrast, in the digital-digital conditions the children tended to work alone when using the digital disks to mix colour. One child would drag the disks on top of each other while the other looked on. After trying a few combinations they would stop and wait for the next task to be given them by the facilitator. Thus, they were much less adventurous and exploratory in their actions. They were also less communicative and engaged with the task. In sum, the Chromarium study showed that the high physicality condition resulted in more active interaction, in the sense that more experimenting, spontaneity and glee were witnessed, coupled with explicit reflection on the why, how and what-if behind the children's actions. This suggests to us therefore, that physicality provided by tool
9. Digitally-Augmented Physical Worlds
181
manipulation offers the potential for more engagement and exploration, more opportunities for communication between children and more sharing and joint activity.
5. The Hunting of the Snark The Hunting of the Snark was designed as a digitally-augmented adventure game for children aged between 7-9 years (Rogers et al, 2002). A number of digitally-augmented activity spaces were developed, where different kinds and levels of physicality were designed to be a central part of the children's interactions in the game. In particular, various physical-digital couplings were designed to engage the children in novel ways. The goal of the game is to find and discover as much as possible about an elusive virtual creature, called the Snark, which is hidden in virtual space and appears digitally in a variety of physical places (e.g., the air, water, land). To do this, pairs of children have to interact with the virtual Snark, by walking around in a cave where it is sleeping, feeding it in a well where it is swimming, and flying with it in the air.
Figure 4. Using the Snooper to collect virtual clues
To begin the game, the children have to collect a number of physical tokens that will allow them to interact with the virtual Snark. The function of the physical tokens is to provide an explicit mechanism for triggering a physical-digital event. To collect the tokens, the children use a specially designed 'Snooper' tool, which is a PDA hooked up with ultrasonic indoor positioning sensors that can detect hidden virtual objects (Randell and Muller, 2001). The children have to move around a physical space, using the Snooper tool to find them. When in close proximity of a hidden token, a digital representation of it appears on the PDA screen, enabling the children to then find a physical counterpart (see Figure 4). The various tokens consist of food objects (to feed the Snark), musical stones (to enter the cave) and a key (to enter a clothes box).Feeding animals (e.g., ducks, fish) in water is a very familiar activity that the game capitalised on. The children readily took to this physical action by placing the physical food tokens into a feeding chute adjacent to the virtual well (see Figure 5). Depending on the type of food token (e.g., sweets, meat, vegetable) the Snark shows a digital emotional response of disgust or delight on the water's surface. The Snark itself does not appear, just its response in the form of a simple animation.
182
9. Digitally-Augmented Physical Worlds
Figure 5. Feeding the Snark at the well
To find out more about the Snark's personality (e.g., shy, cheeky) the children have to explore a darkened cave. By walking around the floor, their foot movements are detected by a series of pressure sensitive pads. Depending on which part of the cave floor they stand on different types and levels of sounds (e.g., forest sounds) are played; the Snark responds accordingly via emitting its own sounds. Thus, a high level of physicality is involved in this interaction, involving whole body movement to elicit a digital response from the Snark.
Figure 6. Flying with the Snark
To find out more about the emotional state of the Snark, the children fly with the Snark, flapping their arms in front of a screen, with the expectation that the Snark will come out from hiding and virtually fly with them in the digital sky (see Figure 6). To enable this to happen, a pair of wearable 'flying' jackets were built, embedded with multiple detection sensors and accelerometers, that gathered data on pairs of children's arm movements. Again, physical body movements (this time arm movements) were required of the children to trigger a digital visual representation of the Snark to appear; the type of emotional response (e.g., sadness, excitement) being determined by the way the children co-ordinated their arm movements.
9. Digitally-Augmented Physical Worlds
183
In sum, the way physicality was instantiated in the Snark game, was through getting the children to interact with a virtual creature, through a number of physical movements (manipulation, gestural, positional), using a variety of digitally enhanced objects, devices or clothing. In so doing, the children discover different things about the Snark, such as its likes and dislikes and its personality. Importantly, they learn that the State of the Snark changes depending on what they do and how they interact with the Snark in the digitally-augmented activity spaces. With this information, they are able to reflect and create their own understanding of the novel physical-digital couplings - which is very different from how they might do so if reading a story about the Snark in a book or interacting with a software adventure game. The findings from a number of pairs of children playing the Snark game (Price et al, 2003) were indicative of a highly engrossing and engaging experience. The children were never quite sure what would happen next and what the nature of the next physical-digital coupling might be. Similar to the high physicality condition in the Chromarium, the children spontaneously generated theories about the unfamiliar couplings, especially when the effects of their familiar physical actions were contrary to their expectations. This led them to creatively experiment and explore other ways of interacting with the Snark, and, in turn to collaborate much more with each other. In addition, they demonstrated the ability to piece together different bits of information and to understand at different levels of abstraction.
6. Ambient Wood The Ambient Wood was designed as a learning experience in the form of a field trip for pupils aged 10-12 years studying ecology. Digitally-enhanced physical tools were developed, using wireless networking with mobile and handheld technologies, to enable children to learn about the habitats in a physical woodland both during and after their explorations of it. Digital augmentation was also provided through more pervasive means; the children's physical movement and location were used to trigger sounds from various parts of the wood and images to appear on a handheld PDA screen at particular locations in the wood. The forms of physicality provided in the learning experience are shown in table 2. They involve the children walking around the wood and discovering things about the habitats they come across, using their naked eyes and ears, together with various tools and through the use of pervasive technologies, that provide contextually-relevant digital information to their ongoing exploratory activities. These include sounds, images and animations of animal and plant behaviour and at a higher level various biological processes - that are normally invisible to the naked eye or ear. (i) Physical movements in the wood In the physical wood, the children explore the woodland habitats, themselves, by observing different things in the real world, by seeing, touching, hearing and smelling what is around them. Exploring the physical environment in this way provides much physical engagement.
184
Type of physicality
9. Digitally-Augmented Physical Worlds
Method/device
(i) Physical movements Naked eye and ear (ii) Interaction with Probing device physical tool Computer display (iii) Interaction with physical tool (iv) Physical Pinged digital movements information (v) Combining artefacts Tagged objects
Coupling (cause-effect) Physical-physical Physical-digital
Digital media None
Physical-digital
Dynamic visualisation Holistic visualisation Images, sounds
Physical-digital
Animations
Physical-digital
Table 2. Types of physicality experimented with in the Ambient Wood, showing method/device used, the nature of the coupling between the cause-effect of an action and the form of digital augmentation Additionally, in the 'ambient' wood, the children can interact with the physical wood in a variety of ways, resulting in a number of physical-digital couplings being triggered. The goal of augmenting the wood in this fashion was to encourage the children to integrate their physical discoveries of the environment (e.g., spotting a caterpillar on a thistle) with the more abstract knowledge they have learned in the classroom, together with the digital representations collected and observed through the digital augmentation (e.g., the interdependency between a caterpillar and a thistle in a wood for a given ecosystem). (ii) and (iii) Interaction with physical tools One of the tools provided for the children to explore their environment, at a more abstract level, was a probing tool. This was developed to collect light and moisture readings from the environment. The tool is a simple to use and lightweight device, with a light sensor and moisture detector prongs attached to it, enabling alternate readings of light and moisture (see Figure 7). The readings can be taken anywhere and immediate feedback is provided on an attached PDA device (see Figure 8). Abstract representations in the form of relative levels of the two variables are displayed on the PDA screen each time a child takes a probe reading.
9. Digitally-Augmented Physical Worlds
185
Figure 7. Using the Probing device to take moisture readings
Figure 8. PDA showing light and moisture readings
The children's probe readings are also tracked, using GPS, and recorded and then rerepresented to them in the form of an agglomerate dynamic information visualisation, that can be examined in more detail in a classroom setting. In this more conventional learning environment, a PC is considered to be an appropriate tool to use by the children to interact with and hypothesise about their different sets of collected data. The purpose of providing this kind of digital representation is again to provide a link between the abstract data and the physical activity of collecting it, but to set it up in a way that enables the children to reflect (i.e. to stand apart from the physical activity) on how the different patterns of light and moisture they have collected, affects what lives where and why in a part of a habitat. The visualisations also provide a sense of ownership and a personal relationship with the data, that can facilitate their ability to recall where they were in the wood for the various projected data points and what they saw, found and heard there. Having a more intimate relationship with the abstract data, in the sense of knowing how they were physically created, thus can trigger strong associated memories. Furthermore, by having a richly builtup picture ready-to-hand, a better grounding can be available by which to understand the higher order interdependencies that arise in the physical habitats. In so doing, it can provide
186
9. Digitally-Augmented Physical Worlds
better leverage by which to think about how and why such visualisations can be useful in performing more abstract and complex tasks on the data. A main function of this physical-digital augmentation, therefore, is to provide strong dynamic links over time, representation and physical activity that can aid the complex learning activity of reasoning about the interdependencies in a given habitat. Firstly, it provides immediate feedback for the ongoing physical activity of probing the wood, providing a representation of light or moisture levels on the handheld PDA. Secondly, it provides delayed feedback, re-representing the information in a more abstract form that can be discussed and reflected upon later in a classroom setting. Thirdly, it provides cumulative feedback, enabling children to perceive more holistically and abstractly the combinatorial distribution of their collected data from their probing. Each of these builds on the other, providing an integrated set of representations that the children can use to build up and integrate with their own understanding of how the physical world works. (iv) Physical movements Another form of physical-digital interaction that we capitalised on in the Ambient Wood project was the provision of contextually-relevant digital information (e.g., a thistle dying, a bird singing, a caterpillar eating). Children's movements in proximity to where such physical events might occur triggered sounds to be emitted in the wood via wireless speakers hidden in the trees and/or images with sound to appear on their PDA. In creating a digital event to occur in this ambient way, our goal was to prime the children to pertinent aspects of the wood, according to their location in the wood, such that they might see for themselves the physical agent of the digital event (e.g., a thistle, a caterpillar) and reflect upon its behaviour and underlying processes in the habitat. To encourage interpretation and reflection the children reported back to a remote facilitator (located elsewhere in the wood) what they had seen or heard and discuss it in relation to other aspects of the habitat (see Figure 9).
Figure 9. Receiving contextually relevant digital information
This mode of physicality is similar in many ways to that experimented with in the Hunting of the Snark game. A difference, however, is that the unconstrained space of the wood means that the children never know when or what body movements will trigger something to happen. In the Hunting of the Snark game, the indoors space was much more
9. Digitally-Augmented Physical Worlds
187
constrained, and so the children quickly worked out that flapping their arms or walking over the confined space in the cave would cause various digital events to happen. In contrast, in the ambient wood, the children need to walk in the proximity of a pinger (a movement sensing device) to trigger a digital event and then it is short-lived. A number of pingers were strategically hidden in 'key' parts of the wood, being geographically spread out. For most of the time, therefore, their walking movements remain undetected. Thus, serendipity and surprise play a much greater role for how the ambient information is discovered. Minimising the amount of pervasive information available was a deliberate design decision, as we did not want the children to be too focussed on locating the hidden pingers in the wood - otherwise it could have turned the activity more into a treasure hunt. In so doing, the children can instead orient to the physical activities of looking, testing and listening, observing the life in the physical habitat while receiving the occasional pervasive ping of digital information. (v) Combining artefacts Another form of physical-digital interaction we used was tagged objects. After exploring the wood, and discussing their findings, the children are required to make predictions about what would happen to various processes in the habitats if certain alien organisms were placed in the various habitats. Two alien RF tagged objects were provided for this purpose, presented in Petri dishes; a lifelike plastic spider and some fungi. The children have to predict what organisms will survive and what will die, if, either or both are added to the part of the woodland they have explored. To determine if their hypotheses are correct they test them back in the woodland, through using a special periscope device that has attached to it a receptacle (a disguised RF tag reader) in which the Petri dishes, containing the organisms, are placed (Wilde et al, 2003). A digital animation is played on the viewer of the periscope for the three different outcomes, depending on which of the organisms (or both) are placed in the receptacle (see Figure 10).
Figure 10. Experimenting using the Periscope
Another way of combining artefacts that was exploited was the provision of physical tokens, in the form of RF-tagged paper cut-outs of organisms in the wood. The goal was to get the children to think about their experiences when apart from the physical activity in a classroom setting and to reconstruct their findings in the form of a pictorial representation for the different seasons than the one they had explored the wood. In so doing, a key task was for them to externalise the relational links between the kind of habitats and the
188
9. Digitally-Augmented Physical Worlds
organisms they supported. Various kinds of digital feedback were provided at different stages of this task, providing hints as to what works and the provision of resultant animations of combining the tokens in certain ways (Marshall et all; 2003, in press). Findings from a study of different pairs of children exploring the ambient wood were like the, Hunting of the Snark game, indicative of a highly engaging experience, that the children readily took to (Price et al, 2003). Moreover, our deliberate decision to let the activity be 'physically-led' by the children rather than task-based provided many opportunities for discovery and exploration. Never quite knowing what to expect next had the effect of captivating the children and maintaining high levels of concentration and motivation. When 'being there' exploring the wood, they frequently generated ideas and chatted with each other about how to find out about the habitat and why certain things lived where they did. When 'being apart' in the classroom setting, the children were not at all self-conscious and talked openly and freely about their data collections. Using the various digital visualisations and animations provided, they also reflected on why certain parts of the habitats had different readings and what the implications of this were for the organisms living in that part of the habitat. Thus, it appears that the provision of a medley of physicaldigital couplings, varying in kind of physicality, offered much scope for bridging the gap between physical experiences and the learning and reasoning about abstract processes - in ways, we would argue, that goes beyond what has been possible to achieve through the provision of desktop educational software.
7. Conclusions Our research has shown how digitally-augmented physical spaces can be designed to exploit the interactional capabilities enabled by pervasive technologies, to support learning in quite different ways than has traditionally been the case with desktop-based PC interactions. Most significantly, the technologies provide opportunities for designing a new genre of technology-mediated learning that combine physicality and cognitive processes. In so doing, they can foster reflection, interpretation and creativity. A key question this raises is how do such arrangements of physicality and digital augmentation promote more active forms of learning? In what ways, say, does moving through a physical space, manipulating physical objects or combining them - causing various digital events to occur - get children to think more about what they are experiencing, doing and importantly, how this relates to what they know and what they do not know; and what the significance of the coupling they have experienced is to what is happening around them and their previous experiences? In particular, how do the children make sense of and integrate the series of physical-digital encounters they have within the physical environment they are in while also abstracting relevant knowledge from them? We suggest that one of the key aspects of interacting in digitally-augmented physical spaces is to raise the awareness of the children as to what they are doing in them. Another core aspect is that they provide a rich experience allowing children to make explicit bridges between their various perspectives and understandings of the physical and digital worlds. Another features is the anticipation and contemplation that is triggered when experiencing couplings between highly familiar physical actions and unfamiliar effects. The degree of authenticity of the learning experience and the amount of collaboration that results can also be greater — both of which are considered in the literature to be important aspects of active learning. Below we briefly describe each of these aspects:
9. Digitally-Augmented Physical Worlds
189
Awareness Physical interaction in the digitally-augmented physical spaces enables awareness at different levels: firstly, awareness of the objects being manipulated and their functionality, focussing attention on the activity or task at hand; secondly, awareness at a multiple perceptual level, enabling access to more information through different senses, providing a richer basis for reflection about the 'world' and thirdly, focusing attention on contextually relevant information at a given time, drawing attention to 'highlighted' aspects of a physical world in particular locations. Experience The experience of interacting in the digitally-augmented physical spaces is more than just visual or multimedia (as is largely the case with desktop interaction). It can entail all senses. Experiencing the digital-physical couplings through more varied modalities, simultaneously or separately, offers a greater diversity of perceptual information on which to reflect about the experience, the environment, or the discoveries that are being made. Anticipation Many of our physical-digital couplings were designed based on familiar physical actions and unexpected effects. For example, walking passed a plant (e.g., a thistle) results in an ambient sound being played, and in so doing making an invisible behaviour in a habitat visible (e.g., butterfly sipping nectar). The juxtaposition of the familiar with the unfamiliar provokes children in ways in which normally taken-for-granted couplings (between familiar action and familiar effect) are often overlooked, causing them to reason how and why such links are possible. Exploration The spaces allow for high levels of exploration and discovery. Many of them can be designed to enable a number of different combinations of actions and interactions to be experimented with and in so doing allowing for creative exploration. Authenticity Another aspect of digitally-augmented physical spaces is that they can provide children with the means by which to interact with the physical environment, that is personally meaningful for them, allowing them to reflect on their own experience in relation to the learning experience. This form of interaction was most evident in the Ambient Wood learning experience, where children had the opportunity to physically collect data that was then re-represented digitally to them and which they could manipulate and hypothesise from in more abstract ways. Collaboration Compared with static screen collaboration, digitally-augmented physical spaces can support more diverse forms of collaboration, between children and others. The interactions are very much part of the ongoing learning or playing experience and as such can be commented or questioned as part of the ongoing activity. More collaboration involves more verbal engagement, promoting a greater exchange of ideas and suggestions. In conclusion, our research has shown how designing computer-based interactions as an integral part of digitally-augmented physical spaces provides much scope for supporting active learning. Our research has shown how technology-mediated learning experiences can be designed to encourage children to engage in a range of familiar and pervasive physical actions and interactions, where digital information is triggered, obtained or made present as part of the ongoing learning or playing activity in hand. In so doing, children can become much more engaged and creative participants in their learning experience.
190
9. Digitally-Augmented Physical Worlds
Acknowledgements This research is funded by the Engineering and Physical Sciences Research Council (EPSRC) as part of the EQUATOR IRC, and was inspired by the late Mike Scaife. We would like to thank our research partners from the University of Bristol, University of Nottingham, University of Southampton and the Royal College of Art. We thank the schools and all the children from Sussex and Nottingham who have participated in this research.
References Ackermann, E. (1996) Perspective-Taking and Object Construction: Two Keys to Learning. In (eds) Y. Kafai and M. Resnick, Constructionism in Practice: Designing, Thinking and Learning in a Digital World Lawrence Erlbaum, NJ. Ananny, M. & J. Cassell (2001). Telling Tales: A new toy for encouraging written literacy through oral storytelling. Presentation at Society for Research in Child Development, Minneapolis, MN, April, 2001. Breslow, L. (1999) New Research Points to the Importance of Using Active Learning in the Classroom. Teach Talk Articles in the Faculty Newsletter, Vol. XIII, No. 1, September/October. Clements, D. (1995) Playing with Computers, Playing with Ideas. Educational Psychology Review 7(2), 203-207. Bruner, J. (1973) Going Beyond the Information Given. New York: Norton. Dansky, J. L., and Silverman, I. W. (1973) Effects of Play on the Associative Fluency in PreschoolAged Children. Developmental Psychology 9, no. 1. Forrester, D. and Jantzie, N. (2000)http://www.ucalgary.ca/~gnjantzi/learning theories.htm [March 2003] Hall, W. (1994) Ending the Tyranny of the Button, IEE Multimedia, 1(1), 60-68, Spring 1994. Hoogeveen, M. (1997)Towards a theory of effectiveness of multimedia systems. International Journal of Human-Computer Interaction. 9 (2), 151 -168. Holzer, S. (1998) From Constructivism ... To Active Learning http://www.succeed.ufl.edu/innovators/innovator 2/innovator002.html [March 2003] Hoyles, C. and Noss, R. (1999) Playing with (and without) words. Proceedings of the seventh Euopean Logo Conference Eurologo '99 Sofia, Bulgaria, 18-29. Ishii, H. and Ullmer, B. (1997) Tangible Bits: Towards Seamless Interfaces between People, Bits and Atoms. CHI 1997: p. 234-241 Large, A., Beheshti, J., Breuleux, A., Renaud, A. (1994) Multimedia and Comprehension: A Cognitive Study. Journal of the American Society for Information Science, 45 (7), 515-528. Morris, J. (1999) The Digital Lunarium: learning through web.media.mit.edu/~joanie/ti99/dig-lunarium.pdf [March 2003]
physical
manipulation
Papert, S. (1980) Mindstorms: Children, computers and powerful ideas New York, Basic Books.
9. Digitally-Augmented Physical Worlds
191
Piper, B. Ratti, C. and Ishii, H. (2002) Illuminating clay: a 3-D tangible interface for landscape analysis. CHI 2002: p. 355-362. Price, S., Rogers, Y., Scaife, M., Stanton, D. and Neale, H. (2003) Using 'Tangibles' to Promote Novel Forms of Playful Learning. Interacting with Computers, Special Issue: Interaction design and children, May 2003, Vol. 15/2. Randell, C. and Muller, H. (2000) Context Awareness by Analysing Accelerometer Data In (eds) B. Maclntyre and B. lannucci,, The Fourth International Symposium on Wearable Computers IEEE Computer Society, October 2000 p. 175-176. Randell, C., and Muller, H. (2001) Low Cost Indoor Positioning System. In (eds) Gregory D. Abowd, Ubicomp 2001: Ubiquitous Computing, p. 42—48. Springer-Verlag, September 2001. Resnick, M., Martin, F., Sargent, R., and Silverman, B. (1996) Programmable Bricks: Toys to Think With. IBM Systems Journal, Vol. 35, No. 3-4, p. 443-452. Resnick, M., Martin, F., Berg, R., Borovoy, R., Colella, V., Kramer, K., and Silverman, B. (1998) Digital Manioultaives: New Toys to Think With. In Proceedings of CHI '98 pp. 281-287 Resnick, M., Berg, R., and Eisenberg, M. (2000) Beyond Black Boxes: Bringing Transparency and Aesthetics Back to Scientific Investigation. Journal of the Learning Sciences, Vol. 9, No. 1, pp. 730. Rogers, Y. and Scaife, M. (1997) How Can Interactive Multimedia Facilitate Learning? Proceedings of First International Workshop on Intelligence and Multimodality in Multimedia Interfaces. Rogers, Y., Scaife, M., Harris, E., Phelps, T., Price, S., Smith, H., Muller, H., Randal, C., Moss, A., Taylor, I., Stanton, D., O'Malley, C., Corke, G., and Gabriella, S. (2002) Things aren't what they seem to be: Innovation through technology inspiration. DIS 2002 Designing Interactive Systems Conference A CM 2002 Rogers, Y., Scaife, M., Gabrielli, S., Smith, H. and Harris, E. (2003). A conceptual framework for mixed reality environments: Designing novel learning activities for young children. Presence, Vol 11, No 6, pp 677-686. Rogoff, B., (1990) Apprenticeship in Thinking: Cognitive Development in Social Context. New York: Oxford University Press, 1990. Saxe, G., Gearhart, M., and Guberman, S. (1984) The social organisation of number development. In B. Rogoff and J. Wertsch (eds) Children's learning in the Zone of Proximal Development, San Fransisco, Jossey Bass. Scaife, M. (2002) External Cognition, Innovative Technologies and Effective Learning. In (eds) P. Gardenfors and P. Johansson, Cognition, Education and Communication Technology. Lawrence Erlbaum Associates. Soloway, E., Guzdial, M., Hay, K., (1994) Learner- Centered Design: The Next Challenge for HCI, ACM Interactions, April. Stanton, D., Bayon, B., Abnett, C., Cobb, S and O'Malley, C. (2002). The effect of tangible interfaces on children's collaborative behaviour. In Proceedings of Human Factors in Computing Systems (CHI 2002) ACM Press. Tapscott, D. (1998) Growing Up Digital: The Rise of the Net Generation. New York. McGraw Hill. Underkoffler, J. and Ishii, H. (1998) Illuminating light: an optical design tool with a luminoustangible interface. In CHI' 98 Proc., ACM Press, New York=., 542-549.
192
9. Digitally-Augmented Physical Worlds
Vygotsky, L. S. (1978) Mind in Society: The Development of Higher Psychological Processes (eds) M. Cole, V. John-Steiner, S. Scribner & E. Souberman. Harvard University Press, Cam, Mass. Wood, D., and O'Malley, C. (1996) Collaborative learning between peers: An overview. Educational Psychology in Practice, 11(4), 4-9.
YVONNE ROGERS is professor of information science and Informatics at Indiana University and adjunct professor of cognitive science. She also retains her chair in computer science and artificial intelligence at the University of Sussex, UK. She is internationally known for her work in the fields of Human-Computer Interaction and Computer-Supported Cooperative Work and has published widely in both. She is interested in exploiting new technologies - especially ubiquitous, pervasive and tangible technologies - in innovative ways to support learning. Her research focuses on augmenting everyday, learning and work activities with interactive technologies. In particular, she designs external representations, especially dynamic visualisations, to support more effectively 'external cognition'. Previous positions include assistant professor at the Open University and a senior researcher at Alcatel telecommunications company. She has also been a visiting scholar at UCSD and a visiting professor at Stanford University, at Apple Research Labs, and at the University of Queensland.
SARA PRICE is a Research Fellow in the Interact Lab at the University of Sussex. With a background in Cognitive Psychology her primary interest is on the role of external representations and tools in mediating learning. Her current research is with EQUATOR (funded by the EPSRC) exploring how combined physical digital environments may promote novel forms of playing and learning, and encourage new ways of thinking. Her main interest lies in exploring ways in which ubiquitous, mobile and tangible computing can support effective learning.
10
194
"What somebody would call storyboarding, and what I call project definition, is for me the real key for generating sustained flow in a spread out course or project. If teachers don't help to formulate the idea of what they are going to do in terms of a story board or a frame, the students won't get into a state of flow, except for the one unique student who is exceptional. Most of the others need structure. Some is provided by the teacher's scaffolding and orchestration and the rest must be built by themselves." Daniel Schneider
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
195
10,
Learning Together Through Collaborative Portal Sites Daniel K. Schneider
Today exists an increasing interest for so-called "active" and "rich" pedagogies that originate in various socio-constructivist schools of thought. The hope is to create deeper, more integrated and more applicable knowledge. We also want students become better general problem solvers and better group workers. Finally there is a pressure to make learning more interesting and even more fun. However, experiments made with learnercentered "new pedagogies" have shown that automatic results are not guaranteed. Good pedagogical design is crucial to their success. The efforts to make "new pedagogies" effective require the use of structured scenarios where the teacher has to fulfill a triple role of facilitator, manager and "orchestrator". A pedagogical scenario (also called a "storyboard") is a sequence of phases within which students have tasks to do and specific roles to play. Without technology they become quite cumbersome to orchestrate and regulate. Moreover, more advanced pedagogical functionalities can simply not be achieved without the help of computers. In other words, modern and active pedagogies are more successful if the teacher can profit creatively from information and communication technology (ICT). We will focus in this chapter on scenarios that are appropriate to higher secondary and university education.
1. Pedagogical Design In our understanding of learning we stress the importance of knowledge construction based on previous knowledge (Piaget's constructivism [15]) and interaction with the social environment (Vygotsky [26]). This leads to various pedagogical strategies like "projectbased", "problem-based", "enquiry-based" or "case-based", upon which we cannot expand here. All of them feature authentic and rich contexts, rich student activities and a certain amount of collaborative work. "The reason that Dewey, Papert, and others have advocated learning from projects rather than from isolated problems is, in part, so that students can
196
10. Collaborative Portal Sites
face the task of formulating their own problems, guided on the one hand by the general goals they set, and on the other hand by the 'interesting' phenomena and difficulties they discover through their interaction with the environment" (Collins et al 1989: 487). Powerful learning environments that aim at the development of general problem skills, deeper conceptual understanding and more applicable knowledge include according to Merrienboer and Pass (2003) the following characteristics: "(1) the use of complex, realistic and challenging problems that elicit in learners active and constructive processes of knowledge and skill acquisition; (2) the inclusion of small group, collaborative work and ample opportunities for interaction, communication and co-operation; and (3) the encouragement of learners to set their own goals and provision of guidance for students in taking more responsibility for their own learning activities and processes." Interaction among subjects has an important place since it generates extra activities like explanation, disagreement, mutual regulation which in turn trigger extra cognitive mechanisms like knowledge elicitation, internalization, reduced cognitive load (Dillenbourg 1999:6). There are different degrees and qualities of interaction among members of learning community: (1) In "Collaborative work" partners work together on the same task, either synchronously or in frequent asynchronous interaction. True collaborative pedagogies are quite difficult to implement, since peers should have more or less the same level, be able to perform the same actions, find a common goal and work together (Dillenbourg 1999). (2) In "Cooperative work" partners split the work, solve sub-tasks individually and then assemble the partial results into the final output. These designs are easier, but the benefit is only interesting if the global task - i.e. building a common artifact like an encyclopedia about certain topic - stimulates individual work. (3) In "Collective work" each works alone on his task, but shares results and problems with the others, and therefore shares inspiration, exchanges help and so forth. "Collective activities" can be added on top of each other strategy in order to boost individual or group performance. Finally, collaborative and cooperative designs can be mixed, e.g. collaborative work in small groups can be combined with a cooperative scenario at the class level. Pedagogical effectiveness is not guaranteed if the teacher simply asks students to do projects and to learn together. We advocate semi -structured pedagogical scenarios that define an orchestrated sequence of learning activities. Such pedagogical "scripts where students and tutors have to play as actors play a movie script are mostly sequential, at least from theh students perspective. However, unlike movie scripts, there is Figure 1: A careful equilibrium between liberty room for improvisation and exploration. and guidance is needed. Some tasks should be defined as mere goals, e.g. at some point the teacher can ask students to hunt down and formulate definitions of the objects they will have to study. The teacher has to respect a harmonic equilibrium between the freedom that is necessary for intellectual development and motivation on one hand and certain guiding principles on the other. Over-regulation will
10. Collaborative Portal Sites
197
have negative effects on crucial factors like development of general problem-solving, metacognition capacities, motivation, etc. that we have defined above as majors pedagogical goals. Finally, we also should point out that "old teaching" by content transmission plus exercising remains more efficient for teaching basic vocabulary, concepts and skills. We will not come back on this matter here but just insist that various pedagogical strategies have to meet pedagogical goals. In this chapter we are interested by the question of how we can teach rich, grounded and applicable knowledge. Activity-based, collaborative, and construction-based pedagogies can be implemented at three levels: (1) the micro-level, i.e. smaller pedagogical scenarios or projects which can be components for larger projects, (2) long term projects, i.e. project-based classes and (3) the general study environment favouring student initiative and community building on which we will come back later. While micro activities (lasting only over a single or a few lessons) can not reach the same goals as true project-based teaching can, they can nicely complement traditional instruction and are often the only realistic alternative in today's organization of the school and university system. We first will examine particular instructional design issues at the level of smaller scenarios. Pedagogical scenarios / small projects Projects should be broken into smaller scenarios so that students are challenged to master as much of a task as they are ready to handle. Such scenarios can also be introduced as "mini-projects" within more traditional set-ups. Each scenario evolves in cycles (or phases) with typically the following ingredients (in whatever order): 1. 2. 3. 4.
Produce (write/draw/build) Deposit (sharing) Look (discovery) Discuss (feedback and exchange)
Figure 2: The basic do-deposit-look-discuss loop.
As figure 2 shows, resources, tools and products play an important role. Each time a student does something, there should be a product (even as small as a little message) that is deposited somewhere and that can be looked at and discussed. Let's have a look at a simple example. Imagine that for a given purpose, students need references for a project. We can turn this into a pedagogical activity with a scenario that include the following steps: 1. The teacher introduces the theme, gives clues and asks students to consider the different aspects of the subject (Discuss) 2. Students search the web with various search engines and bookmark the links they find interesting (Look, Deposit) 3. Students then try to work out a certain amount of categories and sub-categories for this theme (Look, Do, Deposit) 4. The results are put in common and a hierarchy is worked out (Look, Do, Discuss)
198
10. Collaborative Portal Sites
5. The approved categories are entered in a common space (e.g. the classroom wall, a sheet of paper or an electronic links management system) (Deposit) 6. Students classify, enter and describe their links (Do, Deposit) 7. Teacher provides an evaluation (Discuss) As this example clearly shows, most active, constructive and collaborative pedagogies do not necessarily need any special tools, but work can be made more efficient and more powerful by adopting some support technology. Walls in a classroom run out of space, paper is lost and collaboration within the classroom is under heavy time constraints and therefore "home work" lacks the sort of support that classroom activities do have. Content needs to be managed, knowledge exchange must be organized, discussion tools must favour exchange of arguments, projects must run, and generated knowledge must be managed. Technology is primarily a medium to organize activities. The teacher's manager role is to make sure that such loops are productive, e.g. that the students produce something, that it is task related, that they engage themselves in metareflection (look critically at their own work) and that they discuss and share with others. The teacher' s facilitator role is to help students with their tasks, e.g. help them to select resources and tools, explain difficult concepts and procedures, "debug" when they are stuck etc. The teacher's orchestrator role is to implement (or most frequently also to create) the scenarios or scripts as they are also called. This means basically to define a scenario as a sequence of clearly identifiable phases in a way that learner's focus on smaller amount of tasks at the same time and that these tasks are not too difficult to be solved at some point. As we said before, such a scenario should not be "over-scripted", the student should in general be its own master of the tasks and tasks should have some flavor of authenticity. Along similar lines, the teacher should not directly interfere with student's products, but only give feedback and evaluation and let the student fix things himself. Defining a scenario therefore is a workflow design problem where the goal is apprenticeship, i.e. what the student has learnt from performing a set of activities. What kinds of activities typically happen in such a workflow approach? (1) Gathering and distribution of information: Teachers and learners share resources and the activities are designed to help them gather information and make it available to all. (2) Creation of collaborative documents: Students can write definitions, analyse cases, solve problems, write documents and create illustrated documents together around specific themes. (3) Discussion and commentaries around productions: Learners identify together facts, principles and concepts and clarify complex ideas. They formulate hypothesis and plan solutions, make links between ideas, compare different points of view, argue, evaluate, etc. (4) Project management related activities: Learners can decide work plans, share tasks and form groups, decide a schedule, etc. Teachers can distribute and regulate tasks. Let's now have a look at a typical more complex set-up for project-based learning. Project-based courses Project Based Learning is a teaching and learning model (curriculum development and instructional approach) that emphasizes student-centred instruction by assigning projects. It allows students to work more autonomously to construct their own learning, and culminates in realistic, student-generated products. More specifically, project-based learning
10. Collaborative Portal Sites
199
can be defined as (Synteta. 2003): • Engaging learning experiences that involve students in complex, real-world projects through which they develop and apply skills and knowledge • Learning that requires students to draw from many information sources and disciplines in order to solve problems • Learning in which curricular outcomes can be identified up-front, but in which the outcomes of the student's learning process are neither predetermined nor fully predictable • Experiences through which students learn to manage and allocate resources such as time and materials. Projects are complex endeavours involving many different activities. In particular students have trouble for (a) initiating inquiry, formulate coherent research questions; (b) define a research project; (c) direct investigations; find resources, (d) manage time; keep deadlines, estimate time needed to do a task, (e) collaborate and give feedback; articulate work of others and give regular feedback, (f) follow-up the project; revise products. In addition to the difficulty of setting clear goals for various phases, students have trouble relating data, concept and theory. Therefore, a teacher should orchestrate a project into several more or less sequential scenarios that in turn can be broken down to smaller phases (we will show an example later). This will insure that learners will focus on smaller sub-problems, will do things in the right order (i.e. define research goals in the beginning of the project). While it is the role of the teacher to orchestrate (define both boundaries and steps for student activities), a certain amount of fuzziness is absolutely desired for several reasons. First all students must be trained early on to cope with these sort of situations, because they are current in "real life", i.e. fuzziness is part of authenticity. Second, only when students are put in situations where they have to re-construct knowledge (not just to reproduce it) are they able to constitute deep, grounded, connected and applicable know-how. Of course, a project course also can contain more "classic" learning activities, e.g. how to use a technical "language". The general study environment The community factor is particularly important in open and distance learning situations but relevant for any set-up. Functioning virtual communities share a certain amount of practices, common goals, common language and some idea of members helping each other". In our case, the core of a learning community is the class and can quite easily be enhanced with the help of collaborative Internet portals because social interactions can be mirrored and reified (there are traces left that can be inspected) and students also can meet synchronously or asynchronously when they are not in class. In addition to support the class-as-community, we suggest to open portals to the outside world and to let profit other classes from the resources and work produce, to show other teacher's how "you did it". It is also possible and highly beneficial to have outsiders (like domain experts) participate. E.g. the portal used to teach some academic matter could also be promoted as a resource site to the whole world. External users may at some point start to participate and therefore enhance the teacher's and student's own information and collaboration web. It is even conceivable to plan the other way round. A teacher in academics could also start by building up a community site for a given topic and then include students' activities inside to make it
200
10. Collaborative Portal Sites
more "lively". Intensive pedagogies, flow and creativity It is very important to us that teaching generates enthusiasm, favours concentration and also creativity, which are very distinct but somehow interconnected phenomena. Lloyd P. Rieber (1998) convincingly argues that learning process itself - and not just the result should be interesting, if one seeks higher motivation among learners. "Serious play" or "hard fun" are intense learning situations where learners engage large amounts of "energy" and time and that do provide equally intensive pleasure at certain moments and which have been identified as "flow" or "optimal experience" by Mihaly Csikszentmihalyi in 1990. Flow situations have been mainly noticed and studied in play or artistic creation and are defined as states of happiness and satisfaction that arise when "carried" by an automatic and spontaneous activity. It is interesting for teachers to know that "flow states" go along with the impression of discovery and creation and boost performance in conjunction with important cognitive efforts. Conditions in which flow happens are characterized in the literature by an optimised level of challenge, a feeling of control adapted to the learner, a touch of fantasy, and feedback of the system. There are multiple lessons that we could draw for the design of learning environments. An open, active and project- based learning is favourable to trigger challenge, curiosity, and it lets the student have some control, e.g. change things. Other "flow" theory principles are known from more "behaviourist" instructional designs. Open and active learning should not be "programmed", but the teacher still can insure that at least some tasks are very affordable and lead to quick results and more importantly that frequent feedback is provided by the system, by co-learners or by the teacher (whatever appropriate). Creativity is a far more complex issue and its relation to flow is not obvious. "Optimal experience" as described by gamers or programmers enhances without doubt productivity, but does not necessarily entail creativity. According to Feldman (1994), there are many "creativity variables" at three different levels: (1) the social field, e.g. support networks; (2) the domain (symbol systems of knowledge) and (3) the individual including intellectual traits (presence of ideas, complexity of thought, reflexivity), personal traits (e.g. preference for complexity, capacity to provide sustained efforts) and cognitive structures (e.g. expertise). It is clear that education cannot influence on all these variables. But it certainly can have a positive influence on individual dispositions that already exist. It can act upon conditions, i.e. on educational tasks and the general learning environment like the "class spirit" with the help of specially designed technology that we will introduce later.
2. Tools ICT have support potential for most of the functions provided by an educational system. Several pedagogical-technical models are currently available and sometimes in competition. Examples are "open-resource-based learning" (using simple web technology), neoinstructionalism (e-leaming platforms), collaborative learning (using computer-supported collaborative learning systems or groupware systems), tele-teaching (using increasingly sophisticated conferencing systems). The history of the pedagogical use of Internet shows a confrontation of two schools of thought, one that favours open and activity-based teaching
10. Collaborative Portal Sites
201
and an other that focuses on the digitalisation of traditional knowledge transmission. Early adopters of the web in 1993 were advocates of innovative project-based "teaching and learning" and they saw the new technology as a chance for change. But quickly web-based training systems inspired by traditional CBT software started appearing on the market and form the core of today's so-called "Learning Management" or "e- learning" systems. They focus on delivery of content and have little interest for active and rich pedagogies. Open-ended, creative and active pedagogies can profit from almost any Internet technology, as long as students can also be producers (not just readers and button pushers). Besides traditional tools like HTML pages and forums, there exist quite a number of interesting tools like participatory content management systems or collaborative hypertexts like Wikis. However, there isn't any "rich" platform with an offering that covers most of our needs. In other words, we do have educational software to deliver course- ware but we don't need it really. Wanted are tools that support students to solve more complex and open-ended tasks. Our current work aims to provide affordable support for innovative scenarios described by various socio-constructivist schools of thought and we will outline a partial, but operational solution that is available now. ICT as thinking tools Basic technical requirements for active and rich pedagogies are not extremely high. Interesting results can already be achieved by providing the following sort of functionalities (1) Access to rich information sources by various means, e.g. browsing, searching by categories or popularity, searching by keywords; (2) manipulation of various types of information contents (including annotation); (3) rich interactions between actors; (4) integration and knowledge management. Active pedagogies assign a different role to ICT and in particular to the status of documents. Learners generally select the documents they need by themselves from a larger choice (including Internet and libraries). Moreover, they actively participate in the production and the annotation of documents, some of which can be reused later. Such a constantly changing environment therefore also requires awareness tools that put forward what has changed, what is new, what is popular, what is exciting, etc. A computer should mainly be a facilitating structure, a thinking, working & communication tool and not a content transmission device. Accordingly, most student and teacher activities should be supported by computational tools and lead to new "contents". We do not ask ourselves how to convey contents and how to control reading and exercising (as in old-style "elearning"), but how we can support various knowledge productions and exchange tasks. Community, Collaboration and Content Management Systems Simple Internet technologies (web pages, forums and e-mail) have been successful in education because they answered basic needs for information exchange, communication and collaboration and because teachers have control. There are however four drawbacks: (1) Maintaining static web sites (including the student's pages) is time-consuming, (2) simple discussion systems like forums or mailing-lists do not do very good knowledge management. (3) More sophisticated scenarios are badly supported and (4) there is no glue
202
10. Collaborative Portal Sites
for putting "things" together. Community web sites actually face quite similar problems and seem to have found at least a partial answer. Within the last two years an impressive number of what the authors coin C3MS (Community, Content and Collaboration Management Systems) have sprung into existence. Inspired by personal weblogs (also called blogs, which are increasingly popular journaling systems), news systems, simple content management systems and various popular groupware applications, they offer a modular system for configuring interactive community web sites. In addition, most of these systems provide documented extension mechanisms allowing third party persons to contribute modules with additional functionalities. C3MS systems are a form of Web portals. They gather a variety of useful information and communication resources into a single, 'one-stop' web page. A portal therefore is a collection of objects (information bricks) and services (operation on these bricks) and it can be configured for the needs of a specific community. When the user works with a specific resource, e.g. a collaborative hypertext, only a part of the interface changes. A portal therefore is a kind of "cockpit" where the central views changes, but the other instruments stay in reach. Table 1 shows a non-exhaustive list of standard tools available in a typical C3MS portal system and how the can provide support for various functions that a pedagogical information and communication system should provide: Table 1. Function Content management
Knowledge exchange Exchange of arguments Project support Knowledge management
Community management
Functions and tools of the portal
C3MS modules (tools of thejiortal) News engine (including a organization by topics and an annotation mechanism) Content Management Systems (CMS) Collaborative hypertexts (Wikis) Image albums (photos, drawings, etc.) Glossary tool or similar Individual weblogs (diaries) News syndication (headlines from other portals) File sharing (All CMS tools above) Forums and/or new engine Chats Project management modules, Calendars FAQ manager Links Manager ("Yahoo-like") Search by keywords for all contents "Top 1 0" box, rating systems for comments "What's new" (forum messages, downloads, etc.) Presence, profile and identification of members Shoutbox (mini-chat integrated into the portal page) Reputation system Activity tracing for members Event calendar News engine
C3MS bricks We use the term "C3MS brick" for a module (component) that takes care of a specific task, can be easily separated from others, can be configured and administered, can be combined and orchestrated with others and all this through the main portal environment. These
10. Collaborative Portal Sites
203
building bricks for educational scenarios are described in detail in the "TecfaSEED Catalogue" available on our support site (Schneider et al. 2003). Table 2 presents as an example the "news engine" which is usually the C3MS brick by default shown to users. Table 2. Generic name(s) Software names (Post nuke centric) Functional Description Structural Description Pedagogical interest Construction process Other Notice Support for activities
Functional description of the news engine
News/Articles/Topics/Sections News, Submit_News, Story Submission Module, Topics Newsletter, NewsPortal, PN Submit News Submit news, display the news on the index page, post new articles or stories or topics on the site. Functions: Submit, comment, edit, delete, rate, search, browse, moderate This is a core module of most portals but there exist also some 3rd party ones with special features Interact by providing new information (to start a story, a project, an activity), comment information of others, asynchronous debate, present an expert's view on a theme Exists by default in the main menu Can be commented Brainstorm, IntroWork, SendFeedBack, SubmitStory, SubmitComment (see table for explanations)
C3MS bricks mostly are little, but powerful tools to manage smaller bits of information and that allow the community to contribute with comments and sometimes votes. In addition, various applications provide self-ordering and awareness mechanism to the users, e.g. what is new, what is popular, etc. Therefore such portals are particularly useful to manage informally generated knowledge, e.g. the result of educational activities. Portals usually have incorporated search engines, and provide functionalities for rating information so that good information "floats" to the top. For more structured information, e.g. web links, hypertexts etc. there exist special applications that allow users to make quick updates (instead of going through the process of editing HTML files and uploading them). A simple C3MS brick usually offers insert - categorize - annotate - evaluate - sort - search functionality. Such features define the core of a "living documents" or "knowledge management" system and are essential to support student activities engaged in complex pedagogical scenarios. Besides these predominant and simple tools, more complex applications can be embedded into these portals. E.g. our team has developed an "ArgueGraph" (Chakroun 2003), a Computersupported collaborative learning (CSCL) discussion tool according to a model developed by Dillenbourg; "PESC" a pedagogical scenario tool inspired by Moodle (Dougiamas 2002); and ePBL, a pedagogical project management module (Synteta 2003).
204
10. Collaborative Portal Sites
3. Instructional Implementation Scenario Planning Pedagogical storyboarding with a C3MS follows a simple principle. The teacher creates a pedagogical scenario or a sequence of scenarios for a larger project where he describes different phases of the work process. Each phase contains at least an elementary activity that in turn should be supported by a tool (portal brick). An elementary activity is something like "search on the Internet", "insert a link", "make a comment", "coedit a text", Figure 3: Scenario planning and C3MS brick selection "vote for something", "enter an item to a glossary". It is needless to say that external tools can also be used, but at least heir products (e.g. drawings) can be inserted into the portal. We classify activities (basic scenarios) according to the following categories: 1 . Gathering and distribution of information: Teachers and students share resources and the activities are designed to help them gather information and make it available to all. 2. (Co)creation of documents: Here the students can write definitions, analyse cases, solve problems, write documents and create illustrated documents around specific themes. 3. Discussion and commentaries around productions: Students identify together facts, principles and concepts and clarify complex ideas. They formulate hypothesis and plan solutions, make links between ideas, compare different points of view, argue, evaluate... In table 3, let's have a look at an example of the scenario called "references list" that we introduced before. Table 3. The technical-structural definition of the "reference list" scenario Title References list Web search, classifying, conceptualisation, synthesis... Goals 1 1 years old students and more Public Description The students have to work on a theme they don't master for a project. They have to create together a list of web sites that will help them work in a later phase. These sites will have to be described and classified From several days to several weeks Duration Steps The teacher introduces the theme, gives clues and asks students to consider the different aspects of the subject. ("IntroWork" or "Brainstorm") Students search the web with various search engines and bookmark the links they find interesting ("SearchWeb" or "KeepReference) Students then try to work out a certain amount of categories and subcategories for this theme. ("CreateCategories") The results are put in common and a hierarchy is worked out ("CoEdit")
10. Collaborative Portal Sites
205
The approved categories are entered in the portal ("CreateLinkSpace") Students classify, enter and describe their links ("SubmitLinks" or "CommentLinks")
Each scenario described in the catalogue is composed of a certain amount of steps that can be described in terms of generic educational activities, which we labelled with a tag like "BrainStrom" or "SubmitComment". Technical "C3MS bricks" can in their turn, support most of these labelled generic activities. A teacher can therefore plan educational scenarios with the help of a more abstract vocabulary that will help him to choose from a set of supporting technology. Here is an example entry from the catalogue: CoEdit: Creation and modification of collaborative documents. Available C3MS Bricks: • "Wiki": Creation of collaborative documents that can be edited by all members by using a simple markup language. • ContentExpress: Creation and editing of "pages" though webworms and insertion into a menu structure. Less adapted to collaborative work. Table 4 shows a partial list of generic educational activities that are typical components of pedagogical scenarios and that can be implemented with C3MS portals. Table 4. Label Brainstorm
Some elementary generic activities Short description
Everyone says/writes what he knows or imagine on a subject. Before starting an activity students try to figure out all they already know so that they can integrate knew knowledge better. Creation of collaborative documents. They all can modify a unique CoEdit document. CommentLinks Insertion of commentaries under the links entered in the portal. Useful to give hints on web sites content. See "CreateLinkSpace" CreateCategorie Determining categories for a theme, entering categories in an interaction space. s EditGallery Use of the gallery module to display images, photographs and commentaries. Configuration and use of the glossary to display words, expressions and EditGlossary their definition. The teacher can enter the words to be defined and the students can enter the definitions of the words or expressions whose meaning they had to find out. Modification of an already displayed text EditStory EditSummary: Creation of a specific displaying space on the portal Creation of a quiz, poll, survey... EditVote: Discussions on forums or in the news engine (add comments to stories) Interact Giving one's opinion on the value and interest of the displayed links by RateLinks: voting. The links are then classified. SendFeedBack Evaluation and commentaries upon productions by the teacher or experts (during and/or after the activities). The feedback can be individual or collective. Displaying of the best productions according to chosen criteria, which can ShowBest be quite motivating. SubmitComment Commentaries, reactions, complements to a text or any presented material. Can be done an unlimited amount of times SubmitLinks Insertion of a link in the link space, which will have been configured - see "CreateLinkSpace" and "CreateCategories" SubmitQuestion Insertion of a question to be answered or a word or expression to be defined SubmitStory Insertion of a text to be commented or to which a following or commentaries will be associated, exposition of cases... Vote for a project, an idea, etc. VoteFor
10. Collaborative Portal Sites
206
A complex scenario template at high-school level Let's examine now a larger, but not too complex scenario template that illustrates PROJECT "study wild life" the basic principles of scenario planning Main activities (scenarios) that could happen in a biology class a high school level. Imagine a class where students 1 .earn how to use a portal have to study wildlife of the area. One could Make a common glossary (including 2 imagine that each student can select an inks to resources) animal for study (including more "exotic" 3 Define research subjects genres like insects and fish) and that for Make research plans (including each animal a certain amount of options 4 •esearch goals) remain open, e.g. study of habitat, 5 -ield work behaviour with humans, reproduction, etc. Each project should be defined individually, 6 but the very general approach could remain similar for all participants as expressed in F igure 4: Scenario planning and C3MS brick figure 4. Then, there should be a certain s election amount of collective activities, like the construction of a glossary that defines essential terms. If the teacher considers that students will better understand terms if they search, write and discuss them themselves, then he can look at our template and adapt it to his own needs. Our template for the glossary activity can be found in figure 5. It is important to state again that we only make suggestions regarding the different phases and that we do not even suggest a single mapping to elementary activities nor a single mapping of elementary activities to a technical module. In other words, the teacher must be in control throughout the whole design. Educational technologists should only offer "half-baked" solutions. Ideally, teachers have to adapt a pedagogical-technical implementation to the task, their conceptual and technical skills and to available technology. Also, a compromise must be made between selecting the best tools for each task and not to overwhelm the students with GLOSSARY activity Phases 1 jarticipants identify nteresting "words" 2
agree on a jrovisional list
(scenario) Generic activities IntroWork, Brainstorm EditGlossary oP EditPage
3
search for informatior Search Web, and share links EditLink
4
synthesis and editing CoEdit
5
eacher feedback
CheckWork
6
Jditing of final jefinitions
CoEdit
Generic activities CoEdit
description
Available C3MS bricks
creation of Wiki, collaborative ContentExpress documents GlossaryTool generate
Brainstorm deas
Wiki, News Engine, forums, Bulletin Boards
Figure 5: Glossary scenario, possible generic activities and available bricks
10. Collaborative Portal Sites
207
too many tools. Figure 5 shows some of the "open decision space" teachers may have. After examining the situation he may for instance come up with the following solution (table 5). As one can see, our hypothetical teacher selects 3 tools (Wiki, Links manager and News engine): Table 5.
An instantiated glossary activity Instructions to students Tools
Phase 1 Participants identify interesting "words"
Wiki
2
agree on a provisional list
Wiki
3
search for information and share links synthesis and editing
Google, Links manager
4
Wiki
5 teacher feedback
News engine
6 editing of final definitions
Wiki
After discussion in the classroom, each student has to select three terms and enter them to the wiki as homework (first come, first goes) In the classroom, the list is discussed and cleaned up and each student will receive 3 items to work on. Each student has to produce 4 links (day 1) and comment 2 other links (day 2 of homework) Each students receives 2 links and has to edit them. Students are encourages to link to other items and external links. Teacher writes a feedback article, which is also discussed in class. Students make final modification to their work and will be evaluated on this.
This example illustrates the structure of exploratory scenarios. Generally speaking, a teacher should think about the following set-up that reflects the principles of pedagogical workflow introduced earlier: • Activities should start with some sort of conditioning that will generate curiosity, interest, motivation and also show the interest of technology in our case. The initial classroom discussion and the perspective of publishing a nice glossary on the Internet should do this. In addition, entering 3 words on a Wiki is not very difficult and will make students familiar with the particularity of this tool • Activities should give space to discovery by induction and therefore include exploration, search for information, experimentation and formalisation of working hypothesis that can be confronted to the others. Activities in phase 3 partly implement this. • Learners should be active and creative, even when they are involved in seemingly simple tasks like glossary making. They should discuss and cooperate with their pairs. Our glossary scenario has some "built-in" collaboration requirements. • Feedback is important for each student activity. We also suggest a formal evaluation of the final product (including a score). The teacher may also give bonus points for cooperative behaviour, e.g. forum messages or helpful comments for the other's work.
A note on scenario planning, execution and evaluation There are many different schools of thought regarding scenario planning, but some general advice can be formulated: Most scenarios require a preparation phase. In larger projects students may encounter several preparation activities throughout the project. When possible students should choose their subjects, working strategies and formulate their goals. It is very difficult and not always desirable to predict the detailed development of scenarios. The teacher should prepare and master a certain number of path breaking inputs and advice
208
10. Collaborative Portal Sites
according to the needs of the learners. These rather retroactive inputs differ from the more proactive strategy of a more instructional design approach. Teacher's input should sustain learner's activity and should not always be formulated as directive but as advice. The tone for all presented tasks should consider that. Students may not carry out the expected actions. In such cases new regulatory inputs must be generated, e.g. the teacher could post an article with advise to the news board. He may even formulate new pedagogical goals and adapt his script accordingly. Regulation is a subtle and individual intervention and needs to be adapted to every single learner if possible. It is one of the most important guiding tools. The evaluation part of scripting is not meant to be solely a classical evaluation at the end of the scenario that in our opinion should also happen in order to provide a clear reward. However, it is more important, that the work (i.e. the 'visible products' of activities) produced by the learning community should be evaluated (compared, commented,..) by all participants, learners and teachers. It is crucial to design the way such comments and criticism should take place during the scenario. Learning how to give feedback to other learners is by the way a pedagogical goal by itself, since it is an ambitious task to implement a good feedback culture within a learning community. Finally we would like to recall again the most important principle: Do not overscriptl Students need some space of liberty, do have to formulate goals and finds, do have to make errors. Otherwise they will not develop general problem solving capacities, i.e. metacognitive capacities, which is a clearly stated goal of active, and rich constructivist pedagogy. As a corollary, teachers must expect breakdowns and reasons leading to opportunistic scenario adaptation.
Example of a "light-weight" Internet Activity for children Our team did and does participate in several so-called Internet activities. The projects we support are mostly run by active teachers or non-governmental agencies and concern extracurricular activities like "water", "ecology of polar regions" or children's rights. We shall briefly report here on an annual two-month activity we run with our "Terre des Hommes" partners and that are open to any class wishing to participate. In the 2002 edition concerning the particular topic of "migration" we mainly worked from classes around Geneva and from Burkina Faso. The portal was designed to support the following activities from which the teachers could choose or combine: • A discussion forum to initiate dialog between different nationalities. Different topics were created according various lines of reflection determined by the core group of Terre des Hommes volunteers and participating core teachers. • Article sections contained stories about concrete migration experiences and were open to discussion • A quiz section allowed testing knowledge about migration and legal programs. Teacher's could submit their own quiz (including ones produced by their own class) • Pupils could submit their own experience as stories • A poems tool allowed to publish and comment poems • A photo album was meant to present classes to each other or to show pictures and drawings of other interest • In addition to these interactive tool, the portal contained various structured
209
10. Collaborative Portal Sites
information These types of activities are not particularly original, but some teachers did profit from the occasion to create some longer structured activities and to integrate the "children's rights" theme into a curricular context, e.g. the French class. It is important to mention that most teachers in our area are only used to produce web pages and to use simple "threaded forums" with their classes. Therefore, our own goal was to make teachers familiar with the idea that there exist a variety of little interactive tools appropriate for different tasks, and that different tasks could be integrated into bigger and richer scenarios. Most teachers newly exposed to this kind of activities were only able or keen to implement shorter activities and preferably with the forum. However, once they have been trough an experience, they usually try to orchestrate richer scenarios and also can count on the advice by the increasing amount of more experienced teachers. In conclusion, while such Internet activities do not necessarily provide enormous pedagogical benefits, they nicely prepare both teachers and pupils for more complex activities.
Example of a project-based course ICT supported project-based courses can nicely be set up in a "blended situation", where face to face teaching is mixed with distance teaching. The methodology and techniques we report here are developed and studied by P. Synteta (2003). We estimate that the methodology is ready for usage, although progress in several areas can and will be made. The course we will briefly describe here concerned "exotic hypertexts" and was given in a mixed format by he author. It lasted 6 weeks, with a few initial half days in classroom and a 2 hours presentation of the projects at the end of the course. The public were 12 graduate students in educational technology of many different backgrounds. The students had a large freedom for choice of subjects within the general theme and basic requirements were to produce a research plan, to respect of task schedules, to participate in mandatory collective work (include diary writing), then to execute the research plan and to produce a draft of a paper presenting results. There were three major pedagogical goals: (1) Learn something about a specific topic related to more exotic hypertexts (Topic Maps, MOO spaces, Wikis, RDF/RSS syndication, etc.); (2) Leam XML; (3) learn how to run exploratory projects. Table 6 shows the skeleton of the major students activities. Within each activity a certain number of tools had to be used by the students. Table 6.
1 2 3 4 5 6 7 8 9 10 11
Major phases of the Staf-18 course on "exotic hypertexts"
Activity Get familiar with the subject project ideas, Q&R Students formulate project ideas Start project definition Finish provisional research plan Finish research plan Sharing audit audit Finish paper and product Presentation of work
Date 21-NOV-2002 29-NOV-2002 02-DEC-2002
Imposed tools (products) links, wiki, blog classroom news engine, blog
05-DEC-2002 06-DEC-2002
ePBL, blog ePBL, blog
11-DEC-2002 17-DEC-2002 20-DEC-2002 10-JAN-2003 16-JAN-2003 16-JAN-2003
ePBL, blog links, blog, annotation ePBL, blog ePBL, blog ePBL, blog classroom
210
10. Collaborative Portal Sites
The course first starts with a "wake up activity" where students had to enter resources into the links manager, and enter a few definitions in the wiki. Project ideas have first been discussed in the classroom. Classroom activity also include some traditional teaching, i.e. several introductory lectures + questions. In a next step, students had to formulate projects ideas as articles. Once they starting working on a project students had to use a special purpose project tool named ePBL, which stands for "Project-Based e-learning" (Synteta 2003). In particular, they had to define research plans with a specially made XML grammar. Required information concerned overall aim of the project, research goals and questions, work pages etc. Student could upload these files to server by the means of a versioning system. Since students had to work with a validating editor (of their own choice) the XML grammar reinforced conformance of projects to some norms. More importantly, the grammar acts as scaffolding or thinking tool helping the students to produce and structure ideas. Contents of the uploaded project file are automatically parsed and summary information is made available in a students/teacher cockpit. Students were asked at regular intervals to update the project file (including workpackage completion information). Teachers then use the cockpit to annotate the project with comments and to enter a more formal evaluation. After each such audit the teacher also posted a summary article to the portal. At the end of the course, the students had to write a paper, again by using an XML grammar from which an electronic book containing all the work has been produced. In addition to the above-mentioned main activities, other interactions were carried out. Sometimes articles about a course-related topic were posted (even spontaneously by students). The portal also has support forums (both technical and conceptual). It displays RSS news feeds summarizing news from other interesting sites. Some side blocks contain awareness tools (who is connected, who passed by, new messages in forums, etc.). A shoutbox (mini-chat) was used to reinforce the feeling of being "present" and for short messages from the teacher. Other less popular modules include a calendar and chat rooms. Lastly, after each activity, students had to make a diary entry (personal weblog) which gave the teacher important information about encountered difficulties. Students also have used the weblog and the wiki as personal sounding board. The news engine was used as complementary regulation tool besides the ePBL project definition and monitoring application. It was to used to announce activities (at least one / week) and to provide feedback regarding activities or observations (namely major difficulties found in weblogs or forum messages). The news engine therefore is a "heart-beat" tool that gives "pulse" to the whole process, which is very important. Results of this and the two other experiments with other teachers were very encouraging. We found that all students defined interesting projects (either some exploratory empirical studies or some technical developments) and that they came up with interesting results. The quality of the final paper in this specific course wasn't very good generally, but then only a draft has been required and we hardly could ask more in 6 weeks. We found that by using this design, students worked harder, respected deadlines much better and met the pedagogical goals. Class spirit was quite extra-ordinary and we shall comment on this in the next section. Not surprisingly, teacher involvement was a very critical variable. Constant pressure, but also rapid feedback and availability of both the teacher and his teaching
10. Collaborative Portal Sites
211
assistant was judged to be highly positive in student interviews we carried out. We are therefore quite happy to claim that this quickly outlined design seems to be a good instance of the teacher as facilitator, manager and "orchestrator" paradigm. There were of course difficulties encountered in our Staf-18 course. In particular, working with an XML grammar at the very beginning of their studies was both a culture shock and a technical difficulty for most students. They never encountered structured text before and had difficulties to adapt. They also had initial difficulties to work with several portal tools at the same time and to participate in collective knowledge sharing and confrontation activities. However, since activities were mandatory and tools were gradually introduced they very quickly (after about 2 weeks) felt even "at home" in the portal, and really appreciated to learn together, a subject we will look into now.
Community, flow and creativity boosting with C3MS portals While as we showed before, C3MS portal provide rich functionalities for pedagogical "story-boarding" they have been designed first of all as community portals and are therefore ideally suited to boost collective learning, creativity and optimal experience. First, the portal should be a rich information space for "domain support" and it should encourage students to add their own contributions. Such a space also encourages exploration. Typical tools are links managers, wikis, news engines and RSS feeds that keep users up-to-date about articles posted to other interesting portals or individual weblogs. Intellectual support is provided via forums, annotations and articles. Student productions are always accessible to all (including visitors) and therefore provide for recognition. One could manage activities by using various standard tools like articles, forums and the calendar or with special tools. In our experience, it has been shown that Figure 6: The virtual environment argument: C3MS support for students are more like to the optimal experience and critical "creativity" variables contribute to an environment if they own an identity. In the student's partly automatically generated home page on the portal one can see their contributions, read public parts of their personal weblog and conversely each production in the portal is signed with a clickable link to the author. A successful teaching by projects pedagogy needs to provide strong emotional support and it is therefore important to encourage spontaneous, playful interaction and corners for humour that will augment quality of on-line life and contribute to class spirit. Tools like the shoutbox or a little quotation box can do wonders. Lastly, but not least, a personal weblog (diary) can stimulate meta-reflection, in particular if the teacher requires that students write
212
10. Collaborative Portal Sites
an entry after the completion of each activity. Portals should also be designed in the spirit of true virtual environments that have drawn a lot of attention in the last decade. Let's discuss a few features outlined by Dillenbourg et al. (2002): A pedagogical virtual environment (VE) is constructed virtual information space built with the appropriate tools as outlined above. A VE is also a social space, where pedagogical interactions take place. Different spaces become places and the places are populated (Dieberger) and configure social activities. This is not actually true of portals, but it is possible to see at least who is connected, who "passed by lately" and who did what. The same holds true for its geometry, virtual space is not truly represented, but "traces" left by students are step in this direction. Students in a portal are not only active information users and exercise executors as in elearning, but they do co-construct the environment. Virtual environments are multi-purpose, they do not just provide a container for specific activities, they can even be used inside the classroom and provide a number of functionalities that support multiple pedagogies, even traditional content transfer and quizzing if needed. Our short discussion shows that C3MS systems provide a lot of affordances. But experience with interactive collective environments shows that technology itself does not necessarily provoke the emergence of rich interactions. In order to turn in "alive" the teacher really has to integrate at least some collective tasks into the pedagogical scenario (like coconstruction of dictionaries, sharing of web links, posting of great ideas found during project execution, argumentation about certain concepts, etc.). Furthermore, the teacher has to insist that all communication (except face-to-face) happens inside the portal, e.g. he has to refuse to answer questions by E-mail and insist that students use the forum. According to experience, only about 1/4 of all learners spontaneously use the community features of a portal, but an other half can be quite easily be convinced by designing appropriate scenarios. Once the space starts building up and once they have been through peer assistance and emotional support they start to develop "a feeling to be at home". Of course, to make this happen the most important variable is teacher engagement. He has "to be there". This is the reason, why we do not the use the very popular term of "learner-centred pedagogy" for our approach. The pedagogies we advocate are very much teacher-centred as well. The teacher's role as facilitator, manager and "orchestrator" is far more prominent than the one he has as simple content presenter and exercise monitor in "traditional" pedagogy.
4. Innovation and Change Management Technology and teachers Technology, in order to be acceptable by the teacher community should appeal to teachers with different levels of technical competence and different levels of "activeness". We discriminate four levels of use with respect to how they appropriate learning technologies: (1) Reusing: Teachers who appreciate ready-to-use material. In our case, this is a scenario that has been instantiated with content. (2) Editing: Teachers who feel the need to modify the content of a scenario they appreciate. On the technical side, module administration skills are needed (3) Designing: Teachers who compose completely new scenarios by reassembling basic components. This requires modules selection and installation skills (4)
10. Collaborative Portal Sites
213
Programming: Some teachers like to program and we can expect them to develop modules. Of course, the same teacher could borrow objects at levels 1, 2 or 3 at different times. We believe that teachers ought to be able to work according to their technical skills, to their personal investment, to what is available. Our local strategy is also to train local ICT support persons in the school system; to sponsor Internet events led by specially motivated actors (teachers, NGOs, etc.) where other teachers can "just participate" to whatever degree they wish; and to propagate the use of portals for community purposes (e.g. school web sites). We are aware that this "new language" which involves both new pedagogical behaviours and appropriation of collaborative Internet tools can be only very gradually (over several years!) be introduced to a larger audience. "Best case" examples that are implemented by teacher's themselves without much help from a research laboratory can help a lot, in particular if they have been run by local teachers.
Teacher communities We also should point out that community portals are becoming popular in other contexts. Increasing familiarity with this tool and perception of its general usefulness for "real life" will help introducing it to education (like the successful use word processors for creative writing). Success stories of new technologies in education are often related to the teachers' ability to insert it into existing knowledge. Teachers must have an operational awareness (von Glasersfeld) in addition to operational control and they can gain it either by participating in school portals or teacher's portals (that are preferably run by teachers themselves. Teacher's active in these portals are now aware of this technology and are much more ready to use them with their own students. Portal technology is also a tool for networking between communities. Automatic news syndication (RSS feeds) allows members of one portal to be aware of what happens in other portals or even individual weblogs. This way, communication flows are insured but the teachers remain in control over their own portal, they are therefore much more motivated to engage in online activities and they remain in contact with their peers and can receive and provide advice.
Difficulties While conducting our field experiments, which are run according to collaborative, design principles we ran into many difficulties. Let's now examine four major issues we identified: (1) The modern interactive Internet that makes use of complex "cockpits" is largely unknown to education. Few users (teachers and learners) have "portal literacy" and they only use a fraction of the offered functionalities, i.e. our first strategy is to install portals, even if the task does not require it. This way, users get familiar with the typical layout of a portal even if they only use the news engine. The strategy to help teachers to run teacher portals is related, i.e. we insist that teachers should experience themselves the technology. Initially, most teachers wish to work with very minimal configurations, but quite soon they start to experiment with additional tools, e.g. a links manager or a wiki. They also can be quite enthusiastic about "fun tools" like a random quotation engine, a shoutbox, or minisurveys. After 2-3 years of occasional exposure to portals, most teachers will be familiar with the portal concept and they are ready to think about complex scenarios. Finally, we offer technical help including hosting and training courses. According to our experience it
214
10. Collaborative Portal Sites
takes at least an intensive training week to train a teacher for this sort of technology. (2) Pedagogical scenario planning ("story boarding") is unusual and very few teachers can "spontaneously do it". Resources like the TecfaSEED catalogue are useful, but they are by no means enough. Either formal training or open and patient support on a per- needed basis is quite essential. However, teaching ICT and new pedagogies to teachers in a classical way is fairly useless (it has been done not very successfully over the last decade). It is more about helping them to fix themselves innovative pedagogical goals and then to assist them to implement something. Again, peer-to-peer support is crucial and we therefore support teacher-run teacher-portals and teacher-led initiatives as much as we can. (3) The worst obstacle is time. Organization of school life into isolated lessons above primary school level and the absence of "project courses" in the curricula inhibit longer lasting activities. There are a few possibilities to "beat time". One is to encourage crosscurricular activities where several teachers participate in the same project. An other strategy tries to integrate extra-curricular activities into mainstream activities, e.g. combine foreign language teaching with a participation in an Internet project. In some countries, teachers at high school level can run specialization courses where they have quite a lot of freedom. The most creative experiments we have observed did happen in biology classes where complex Wiki activities have successfully been conducted (e.g. Notari 2003). (4) Despite all these difficulties, interesting experiments happen at all levels of school. But unfortunately many teachers face administrative resistance, e.g. they face hostile and/or incompetent PC managers or Internet ports censored by network administrators meaning that teachers can't use creative applications like Swiki servers or MOOs. In some areas there exist forms of censorship. To add contents to official school servers, teachers have to pass it through a reviewing committee and this takes a lot of time. Sometimes it is strictly forbidden to put children's pictures on the Internet, and which denies them identity and therefore motivation. Finally, it sometimes is very difficult to host teacher-selected portals on official servers. Our solution here is quite simple, we either host these projects or we teach them how to install portals with a private provider. We just add a short comment about the university level where the issues concern simply pedagogical training and resources. Changing a teaching strategy and running high quality project-based courses require a lot of investment that does "not pay" in career terms. In addition, recent programs that sponsor ICT in university education with quite substantial grants require that funds are funnelled into content production and accreditable distance teaching (main stream e-learning) instead of making existing programs more attractive. The solution is provide incentives for creative teachers and to convey the message that good teaching can be intrinsically rewarding, e.g. by the possibility to leverage ideas from interesting student projects.
10. Collaborative Portal Sites
215
5. New Opportunities We do believe that there are new opportunities for rich and active pedagogical scenarios. Community, Content and Collaboration Management Systems (C3MS) present functionalities that teachers are keen to have, like news/ comments, forums, simple CMSs, wikis and others. These tools offer support for the accumulation, organization and display of contents as well as many forms of user Figure 7: The layer approach to the design of a teach interaction. This allows to create rich ing platform pedagogical "workflow" scenarios. In addition, a well-configured C3MS portal is a community engine that transforms a pure work tool into a collective and collaborative "place" that boosts class dynamics. In our opinion, a pedagogical portal should have a "clear focus" but "fuzzy edges" (Rieber 2001). Often, one associates new rich and open pedagogies with "learner-centred". We believe that being "learner-centred" is not sufficient, since mainstream "eContenf'-centred elearning also rightly claims to be learner-centred, since students work at their own speed. Good learner-centred pedagogies are also very teacher-centred, since the role of the teacher can become very complex and demanding. Let's recall the three principle roles that we attribute to the teacher-designer of structured, but active, open and rich educational scenarios: • His role as a manger is to ensure productivity, i.e. that learners do things. • His role as a facilitator is the help them in their choices and to suggest resources and tools that will help them to solve problems and get tasks done. • His role as an orchestrator is to create "story-boards", i.e. to break down projects into scenarios, and scenarios into phases. He also may decompose problems into manageable sub-problems or alternatively encourage and help students to do so themselves. It is very important to respect a principle of "harmony", to find equilibrium of different pedagogical strategies and tactics and not (and we insist on this) to be tempted by overscripting. In our philosophy, a teacher should think of himself primarily as a "landscaper" who uses ICT to build places where learners can "sculpt" according to some rule and with as much help as appropriate.
216
10. Collaborative Portal Sites
Towards a scenario and modules economy? Since C3MS systems have a modular and an extensible architecture they can be adapted/combined/ configured to many specific usage scenarios". Our hope is to create some sort of educational modules economy with the PostNuke platform in order to gain an initial experience in this area and that later (in 5-10 years?) may lead to more formal standards (including the system's architectures).
Figure 8: Towards a scenario and modules economy for educational scenarios
Figure 8 shows the model of what we hope may become a scenarios and portal modules economy.
A well-trained teacher can configure portals and its "tools" according to his own needs. He can also hunt down new modules. He can re-purpose tools, e.g. he could use quizzes which are normally used for assessment as discussion openers. He can also suggest to the increasing number of technical support people to develop new tools. Since this technology is focused on "orchestration" and not content delivery, we believe that it will spread in the nearer future with almost the same ease as web pages did, but it will bring new functionalities. Teachers should have control over their environment and they can share their experience within teacher portals using the same technology and both fit the C3MS philosophy. Finally, C3MS may be a chance to promote the open and sharing "Internet Spirit" to education, which is threatened by the philosophy of the closed so-called "educational platforms". According to our initial experience, and despite many difficulties like administrative hurdles, the time it takes to accommodate new pedagogical strategies, the disputable ergonomics of some software that we will have to overcome - teachers who engaged themselves "love it" and their students too.
10. Collaborative Portal Sites
217
6. Practical Information
The TecfaSEED teacher's catalogue and the TecfaSEED community portal This portal is a bilingual (English, French) centre for exchange and collaboration on socioconstructivist teaching & learning with the Internet. You are encouraged to submit News, use the forums, add or consult web links, download software and make use of the wiki or any other application. We also provide a limited form of conceptual and technical support in the forums. URL: hitp:/,'!cctasccd.unme.ch/'door
Selecting the right software for your needs In the absence of standards for rich and active pedagogies, we suggest to adopt one of the following solutions: 1. A technology-savvy teacher interested by modern server-side technology should try to install and to run his very own C3MA portal. At the time of writing the "PostNuke" is a good compromise. Possibly on a machine that is available in his school, else with a private provider. A variant is to have it installed by someone in the organization or some Internet enthusiast that will do it for little money 2. Ask around if the school system supports a community portal (for schools) and use this system. You may even think of re-purposing the heavy enterprise portal you may have access to (e.g. Lotus/Domino, IBM Websphere etc.). However, this entails negotiation with some central informatics department. 3. As a last resort, re-purpose the functionalities of an e-learning platform. Currently, we repeat, there is no perfect "off the shelf platform for the kind of pedagogies we advocate. We do not know yet the full potential of C3MS like PostNuke or Zope, and we hope that systems with better functionalities and better ergonomics will emerge. One major limitation of using C3MS portals seems to be the lack of provision for integration (and in particular data-flow) between applications, which are required for more complex Computer Supported Collaborative Learning (CSLS) scenarios. Another limitation concerns management of contents, activities and people over time. However, we think that C3MS are a first step to support a large amount of rich pedagogical scenarios and in addition they can be used for other interesting purposes such as teacher community support or school portals.
A note on e-learning software and standards Learning management systems (main-stream e-learning software) can hardly be used to implement rich pedagogical scenarios. The currently dominant e-learning framework of IMS/ADL/SCORM mainly implements simple pedagogical sequencing (presentations of materials), meta-data and quizzing. Despite serious theoretical foundations (e.g. Gagne), most of what can be seen on the market is simple "shovelware" that often lacks the sort of interactivity that we had with CBT back in the '70s. Progress achieved in the last few years
218
10. Collaborative Portal Sites
mainly concerns modular management of vaguely defined "reusable learning objects" (RLO) plus ease of delivery via the web of course. Interestingly enough IMS adopted in 2003 a new pedagogical markup language and standard which is called "learning design" (LD) (Koper 2003). It does have the potential to describe very rich and diverse pedagogical situations. However, it is yet unknown if industry or an academic consortium will provide a full implementation and we can not predict if there will be one day a standardized pedagogical platform that supports rich and active pedagogical scenarios. In conclusion, given the current state of affairs outlined above, we strongly recommend investing into "street standards" like the PostNuke C3MS. It does have its deficiencies, but a large user community sustains it and it does have a stable API, which favours a "modules economy". Its LAMP (Linux-Apache-MySQL-Php) architecture allows dozens of parttime developers to program interesting modules within a reasonable time frame. What will happen if a system likes PostNuke dies? Well, it already happened twice that the whole core programming group left for other projects. Others took it up. Since such systems are open source and free, there will be upwards-compatible forks. E.g. eNvolution (created by dissenters) runs about 98% of all compliant PostNuke modules. This and the fact that there are no alternatives in sight are the strong arguments in favour of popular open source collaborative portal systems.
Acknowledgements This chapter reports on collective work done by the Swiss SEED research team at Tecfa, namely: Pierre Dillenbourg, Catherine Frete, Fabien Girardin, Stephane Morand, Paraskevi Synteta. I also would like to express my thanks to Barbara Class, Francois Lombard, Charlie Lowe, Alan McCluskey, Michele Notari, Elodie Sierra for interesing input. Confrontation with other SEED projects also has been extraordinarily inspiring to all of us. Let's quote Chronis Kynigos' "half-baked solutions" or Ulrich Hoppe's "orchestrations". Last, but not least, I would like to thank the open source community that brings us tools like PostNuke. I hope that we can pay it back sometimes by bringing Internet creativity to education. This research and field work is part of the SEED project (European 1ST Programme No IST-2000-25214) and sponsored by the Swiss Federal Office for Education and Science (No OFES: 00.0287).
References [1] Bielaczyc, K, Collins,A.(1999). "Learning Communities in Classrooms: A Reconceptualization of Educational Pratice", in Reigeluth, C. (ed) Instructional-Design Theories and Models, Vol II, London: Erlbaum. [2] Chakroun, M. (2003), Conception et mise en place d'un module pedagogique pour portails communautaire PostNuke, Insat, Projet de fin d'etudes. URL: http://tecfaseed.unige.ch/users/mourad/arguegraph/ArgucGraph.pdf [3] Collins, A., Brown, J. S., & Newman, S. E. (1989). Cognitive apprenticeship: Teaching the crafts of reading, writing, and mathematics. In L. B. Resnick (Ed.), Knowing, learning, and instruction: Essays in honor of Robert Closer (pp. 453-494). Hillsdale, NJ: Lawrence
10. Collaborative Portal Sites
219
Erlbaum Associates. [4] Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. Harper and Row, New York [5] Dieberger, A. (1999) Social connotations of space in the Design for Virtual Communities and Social Navigation. In Munro, A., Hook K. & Benyon D. (Eds), Social Navigation of Information Space, pp. 35-54. Springer: London [6] Dillenbourg P. (1999) What do you mean by collaborative learning?. In P. Dillenbourg (Ed) Collaborative-learning: Cognitive and Computational Approaches, (pp.1-19). Oxford: Elsevier. [7] Dillenbourg, P., Schneider,D., Synteta,V., (2002) "Virtual Learning Environments", Proceedings of the 3rd Congress on Information and Communication Technologies in Education, Rhodes, Kastaniotis Editions, Greece, 3-18. [8] Dougiamas, M. & Taylor, P.C., Interpretive analysis of an internet-based course constructed using a new courseware tool called Moodle, Curtin University of Technology, URL: http://www.ecu.edu.au/conferences/herdsa/main/papers/nonref/pdf/MartinDougiamas.pdf [9] Feldman, D.H., Csikszentmihalyi, M. Gardner, H., (1994) Changing the world, A Framework for the Study of Creativity, Westport: Praeger [10] Guzdial, M. et al. (2000) A Catalog of CoWeb Uses, GVU Tech Report 00-19. URL: http://coweb.cc.gatech.edu/csl/24 [11] Guzdial, M., Rick, J., and Kehoe, C. (2001) Beyond Adoption to Invention: Teacher-Created Collaborative Activities in Higher Education, URL: http://coweb.cc.gatech.edu/csl/24 Journal of the Learning Sciences [12] Koper, R., Olivier, B. & Anderson, T. (2003), IMS Learning Desing Information Model. IMS Global Learning Consortium, http://www.imsglobal.org/learningdesign. [13] van Merrienboer, J.J.G. and Pass.F. (2003) Powerful Learning and the Many Faces of Instructional Design: Toward a Framework for the Design of Powerful Learning Environments, in De Corte, E. et al. Powerful Learning Environments: Unraveling Basic Components and Dimensions, Amsterdam: Pergamon, 3-20. [14] Notari, Michele (2003), Scriping strategies in computer supported collaborative learning envrionments, unpublished master thesis, TECFA, FPSE, University of Geneva. [15] PiagetJ., (1967), La pschologie de I'intelligence, Paris: Armand. [16] Reigeluth, C. M. (Ed.). (1983). Instructional-design theories and models: An overview of their current status. Hillsdale, NJ: Erlbaum. [17] Rieber, Lloyd. P., Smith, L., & Noah, D. (1998). The value of serious play. Educational Technology, 38(6), 29-37, [En ligne] Adresse URL : http://itechl.coe.uga.edu/~lrieber/valueofplay.html [18] Rieber, L.P. (2001, December). Designing learning environments that excite serious play. Paper presented at the annual meeting of the Australasian Society for Computers in Learning in Tertiary Education, Melbourne, Australia.
220
10. Collaborative Portal Sites
[19] Scardamalia, M., Bereiter, C., McLean, R., Swallow, J., & Woodruff, E. (1989). Computer supported intentional learning environments. Journal of Educational Computing Research, 5, 51-68 [20] Schneider, D., Synteta,V., Frete C., (2002) "Community, Content and Collaboration Management Systems in Education: A new chance for socio-constructivist scenarios?". Proceedings of the 3rd Congress on Information and Communication Technologies in Education, Rhodes, September 26th-29th 2002 [21 ] Schneider, D. Dillenbourg, P., Frete, C., Morand, S., Synteta, P. (2003), TECFA Seed Catalog, URL: http://tecfa.unige.ch/proj/seed/catalog/, Draft version. [22] Schneider, D. et al. (to appear) "Conception et implementation de scenarios pedagogiques riches avec des portails communautaires", Les Communautes Virtuelles Educatives. Pour quelle Education ? Pour quelles cultures ? Second Colloque de Gueret, 4,-6 juin 2003. [This is a text with a similar aim as this chapter] [23] Synteta, P.(2002). Project-Based e-Learning: The model and the mehod, the practice and the portal. Accepted PhD proposal, University of Geneva, Geneva, Switzerland. URL: http://tecfa.unige.ch/perso/vivian/ [24] Synteta, P. (2003). Project-Based e-Learning in higher education: The model and the method, the practice and the portal. Studies in Communication, New Media in Education, (pp. 263-269). http://tecfa.unige.ch/perso/vivian/ [25] von Glasersfeld, E. (to appear) "Radical Constructivism and Teaching" Scientific Reasoning Research Institute, University of Massachusetts, URL: http://www.umass.edu/srri/vonGlasersfeld/onlinePapers/html/geneva/ [26] Vygotsky, L.S. (1962) Thought and Language. Cambridge, MA: MIT Press. [27] Wilson, B. & Lowry, M. (2001), Constructivist Learning on the Web, in Burge,L. (Ed.),Learning Technologies: Reflective and Strategic Thinking. San Francisco: Jossey-Bass, New Directions for Adult and Continuing Education. URL: http://ceo.cudenver.edu/~brenLwilson/WebLearning.html
DANIEL K. SCHNEIDER is senior lecturer and researcher at TECFA, a research and teaching unit in the faculty of psychology and education, University of Geneva. Holding a PhD in political science, he has been working in educational technology since 1988 and participated in various innovative pedagogical and technological projects. He has been a prime mover towards the introduction of creative pedagogical strategies and ICT technologies. His current R&D interests focus on modular, flexible and open Internet architectures supporting rich and effective educational designs. Within TECFA's "blended" master program in educational technology, he teaches courses on educational information & communication systems and virtual environments. He also organizes training courses for teachers and webmasters.
11
222
"We should be able to turn these workshops and the ideas emerging from them into a research program spanning the spectrum of foundational issues to their implications and from that to real change in education. What does it mean to say that a learning experience is intrinsically motivated? If we agree that flow is a construct, a purposeful construct, then what are the different purposes and the target moves? Making the construct democratic, something that everybody has now and then, may trivialise it. I argue that we should see flow as something very special and I am interested in the role that technologies can play in generating flow. My intuition is that often the technology disturbs the flow experience." Ulrich Hoppe
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
223
11
Collaborative Mind Tools H. Ulrich Hoppe
Computer-Supported Collaborative Learning (CSCL) technology can be more than just an application of computer mediated communication by providing "computational objects to think with" in collaborative environments. This kind of technology is not only of interest for virtual learning applications but also for face-to-face classrooms with networked ubiquitous computing facilities, such as big interactive screens (whiteboards), pen based tablet computers. The "Cool Modes" framework makes use of these ingredients to support collaborative modelling activities in science and mathematics lessons. It supports educational workflow between different levels of scale in the size of learning groups and through a variety of cooperation modes.
1. Dimensions of technology-supported collaborative learning Over the last fifteen years, we have observed an increasing interest in supporting not only individual learners but also learning groups as part of research and development in the field of technology enhanced learning. This new theme has led to constituting "ComputerSupported Collaborative Learning" (CSCL) as a research field of its own right. The most prominent CSCL scenarios have been centred around virtual learning groups using computerised communication and cooperation facilities. The support technologies rely heavily on the Internet, including both local and wider range networking scenarios with applications such as mailing and threaded discussion tools, shared archives such as BSCW/BSCL (cf. [1]), multi-user chats and MOOs [2], as well as tools for synchronous shared workspace communication [3]. It is a misconception to identify CSCL scenarios only with "virtual learning" in the sense of distance learning scenarios. This article will particularly elaborate on face-to-face settings with synchronous co-construction in shared workspaces. From a technical point of view, the same networking technologies can be used in remote learning scenarios as well as, e.g., in classroom settings. Of course, the functional and usability requirements for supporting face-to-face as compared to remote co-construction are significantly different in the sense
224
11. Collaborative Mind Tools
that remote environments require more explicit coordination (turn-taking), floor control and awareness mechanisms. In addition to the distinction between co-located and remote learning settings, also the difference between synchronous and asynchronous applications has to be considered. Asynchronous sharing relaxes organisational and time constraints on the arrangement of learning scenarios. Table 1 shows the well-known 2 x 2 classification theme for groupware application with examples from the learning domain for each category. Indeed, each one of the quadrants stands for specific lines of research into CSCL. Yet, the distribution of research efforts is not homogeneous over all the quadrants: The most special and least dealt with is certainly the co-located / asynchronous condition. Here, we would find fixed installations of digital pin-boards or other types of smart objects that facilitate communication in "passing-by" mode. The other three quadrants are more populated with examples, probably with remote applications being the most frequent. Yet, with recently grown interest in mobile and wireless applications around PDAs and tablet PCs the first quadrant is catching up [4, 5]. Table 1: Classification of cooperative (learning) technologies Same time (synchronous) "computer- integrated Same place (co-located, face-to-face) classroom" group level I-III "virtual classroom", Different place tele-lecturing (remote) group level II-III
Different time (asynchronous) Digital pin-boards, smart objects group level II-III Shared archives, threaded discussion tools group levels IH-IV
Dillenbourg [6] has stressed the importance of group size ("scale") as a distinctive factor in the characterisation of group learning scenarios. Table 1 indicates dependencies between scale and the other dimensions, based on examples discussed in the CSCL literature. The categorisation of scale is based on the following four levels: • Level I: individual activities (though not CSCL in a genuine sense this category is important to complete the picture); • Level II: small groups of 2 to 5 (these have been the main target of "shared workspace" technologies); • Level III: large groups, e.g. a whole class or face-to-face course; • Level IV: learning communities, characterised by anonymity between group members which requires "thematic" interaction in asynchronous mode. So far, CSCL research has mainly addressed the needs of these different scales separately, but of course it is also important to facilitate "learning trajectories" between scales. Incompatible technology still leads to disruptions when a learner brings results of individual studies at home to the classroom, or when small group work in classroom is to be aggregated and discussed in a plenary classroom session. The technology and examples described in the sequel are mainly located at (or in between) levels I-III, yet in the future we have to consider "educational workflows" which span over all levels.
2. "Collaborative Mind Tools" - a synthesis of two paradigms To a large extent, CSCL technologies have been centred around synchronous and asynchronous information exchange, e.g. on conferencing techniques and sharing of
11. Collaborative Mind Tools
225
resources and materials as well as on digital archives. Accordingly, in CSCL computer mediated communication has been in the foreground as compared to techniques involving the semantic processing of computerised representations. On the other hand, originally motivated by the limitations of conventional individualised computer tutors, there is another tendency to focus on interactivity in learning environments using rich and powerful "computational objects to think with". This gave rise to the development of interactive cognitive tools or "mind tools" [7], which were essentially based on the direct manipulation of visual objects by the user-learner but also based on the computational processing of related symbolic objects and representations. Typical examples are visual languages for argumentation and discussion as well as visual tools for simulation and scientific modelling. In the co-construction of scientific models, learners can engage in cognitive and social processes that promote collaborative knowledge building. Accordingly, there is a new challenge in providing modelling tools in a collaborative, distributed computing framework. This can be achieved through shared workspace environments that enable a group of learners to synchronously co-construct and elaborate external representations. Concerning the domain content of these representations, two ends of a spectrum have to be distinguished: on the one hand, System Dynamics models or Petri Nets provide a complete semantic definition of all objects and thus allow for "running" the models as simulations. In contrast, less structured representations are used for conceptual modelling, design, argumentation or discussion support. Here, a full semantic interpretation of the shared objects and their relations is not available. Applications of this type have been reported for group discussions [8, 9, 10] and scientific argumentation [11]. Only more recently, scientific model building with operational representations has been considered for CSCL environments [12, 13]. In general, the co-constructive creation or use of shared visual representations is considered to be an important facilitator for creative processes in group working and learning scenarios. Shared workspaces with visual objects enrich human-human communication by opening a new channel: communication through the artefact. When jointly creating and manipulating artefacts, the co-learners' language based interaction is complemented by an external medium providing inherent constraints. Whereas language utterances rely on individual interoperation "in the head", actions on the object level have directly observable results and consequences for future actions. A basic function supported by shared visual representations is externalisation. Following Nonaka [14], externalisation plays an important role in "organisational knowledge creation", namely in that it supports the transition from tacit, individual to explicit knowledge. According to Hoppe & Plotzner [15] shared workspace environments support the following types of cognitive processes within the learning group: coordination of individual contributions or action through external constraints of the shared workspace, reification of contributions as manipulable objects, "mise en relation " (in a Piagetian sense) by visually relating individual contributions to each other using a spatial metaphor, reuse of group results (e.g., for reflection, comparison, further elaboration). Along very similar lines, Suthers and Hundhausen [16] distinguish the following three essential roles of shared external representations in co-constructive activities: initiating negotiations of meaning, serving as a representational proxy for purposes of gestural deixis, providing a foundation for implicitly shared awareness (group memory).
226
11. Collaborative Mind Tools
In summary, the notion of "collaborative mind tools" stands for a new tendency to extend CSCL technology beyond language centred computer-mediated communication towards richer environments providing "computational objects to think with" - now jointly.
3. The Cool Modes environment In line with the idea of collaborative mind tools, the Cool Modes ("Collaborative Open Learning, Modelling and DEsigning System" [13]) platform has been conceived as a tool environment. This was done not only to support co-constructive activities as such but also to facilitate the provision of different modelling tools and languages for the educational designer and developer. The main difference between Cool Modes and comparable systems such as Belvedere [11] or SEPIA [9] is the potential of adding and even mixing different representations with different degrees of computational semantics. The range of languages includes free-hand annotations and drawings available on a special layer in any workspace, argumentation graphs and concept maps, as well as specific modelling languages for Petri Nets and System Dynamics. Cool Modes allows the use of multiple workspaces represented in different windows that can be arranged freely. Each workspace consists of a number of transparent layers containing objects such as hand-written strokes, images and other media types. Four predefined layers with different functionality exist by default, one for a background image, one for hand-written annotations and two for visual language objects embedded in graph structures. The language elements are provided on "palettes" in the tool menue of the Cool Modes environment. These palettes can be dynamically added and removed. At any time, the user selects one of the currently loaded palettes to manipulate the active workspace. Figure 1 shows an example of a System Dynamics model built with Cool Modes. System Dynamics is a visual specification technique for numerical modelling which has originally been used to represent models of population growth and natural resources [17]. The System Dynamics palette in the tool window to the right consists of three types of nodes and two different types of edges. There are stock nodes (rectangle) which can hold and collect values, there are constant nodes (circle) which represent a fixed value and there are/7ow rate nodes (diamond) which can be used to control the flow between two stock nodes and to calculate intermediate results. A rate node changes its colour to show if it is used to calculate intermediate values (light green) or if it is used to control an actual flow (dark green). There are two types of edges: InfoEdges and FlowEdges. As the names indicate, a FlowEdge is used to actually transfer certain quantities of "substance" from one stock to another, according to the value calculated in the connected rate node. An InfoEdge is used to provide read access to values of other nodes (without changing these). The example model simulates the spread of a disease in a population.
//. Collaborative Mind Tools
227
Figure 1: A System Dynamics model with hand-written notes The collaboration support integrated in Cool Modes relies on a communication server called MatchMaker [18]. Without going into technical details, it is important to know that MatchMaker implements a replicated synchronisation model in which all the connected or "coupled" workspaces or objects are fully operational on a data level in their local computing environments. This implies that models will not be lost when a joint session is finished, and it also allows for partial coupling limited to specific objects or layers in a workspace. Figure 2 shows two workspaces with Petri Nets for which the model layer is coupled, but the handwriting layer is not.
Figure 2: A Petri Net in a shared workspace with private annotations Cool Modes facilitates educational workflow between individual work, small groups and the whole classroom (levels I-III). Imagine the following scenario: There is a group of
228
11. Collaborative Mind Tools
learners in a room; each learner has a computer of his or her own and runs Cool Modes. They start to discuss a topic chosen by the teacher and use Cool Modes to support the discussion and to make (shared or private) annotations. Having agreed on the basic constraints of the problem, they arrange themselves in sub-groups, each group starting a MatchMaker session of its own. The learners solve the problem by building a model in collaborative or individual mode using the various palettes available in Cool Modes. The learners (or the teacher) can make annotations to the models - either hand-written or by using the discussion palette. After a while, the teacher will call the students to finish their small group session and to concentrate on the electronic whiteboard. MatchMaker will be used to connect a Cool Modes workspace on the whiteboard screen to workspaces of one or more groups. Based on the classroom discussion models will be improved and "handed back" to their authors. In our current practice, Cool Modes results are stored in the file system or on a BSCW archive. In both cases, the results are only available for a determined group of users who know about the context and history of these documents. A new development will allow for storing learning results in a database associated with metadata which are partly generated from the lesson context and partly assigned explicitly. This will allow for sharing learning objects in broader community of learners (level IV).
4. Principles for introducing digital media to the classroom In our vision of classroom applications, we see multi-functional and multi-representational tools such as Cool Modes as digital, active extensions of the chalkboard and paper & pencil. The tools should ideally be used in networked ubiquitous and potentially mobile computing environments to support modelling, interactive presentation and group discussion in a variety of educational scenarios, including traditional lectures (presentation) as well as tutorials and collaborative work in small groups. The use of the tools should not be conceived as a special learning mode defined by "working on or with the computer" but as an instrumental extension of a specific scenario. One of the most frequent learning scenarios is still the classroom. Our idea of a "computerintegrated classroom" has been practically elaborated and put into practice in the European NIMIS project [19]. The most evident and concrete result of NIMIS is a classroom installation which features special hardware such as an interactive whiteboard and penbased tablets embedded in the pupils desks in a networked environment with educationally motivated groupware functions. Although the target group of the NIMIS project were "early learners" (4-8 years old), certainly too young for using symbolic modelling tools, we are now extending this approach to elder age groups including high school and academic environments.
11. Collaborative Mind Tools
229
Figure 3: Scenes from the NIMIS classroom in Duisburg Generalising from the NIMIS experience, we can formulate principles for introducing networked digital (rather than "computers") in the classroom: (a) Unification of media and learning activities on a digital platform Educational media of the future will be unified on a digital platform. Negroponte's vision of ,,Being Digital" [20] is particularly fruitful for educational scenarios since integrated digital media facilitate a free flow, re-use and recombination of the materials and products of learning in a classroom. New types of collaborative activities arise from these scenarios. However, the objects of learning are not only digital ones: Easy transitions between the physical and the digital world are facilitated and become a subject of learning processes. (b) Supporting the classroom as a whole through an integrated networked infrastructure Connecting learners in the NIMIS sense goes beyond providing Internet access in a computerised classroom. Intranet facilities are seen as prior to Internet access. Integration, i.e. connectivity and inter-operability, fosters the communal aspect of the classroom and collaborative learning by giving flexible access to classroom resources for teachers and students and by facilitating a high degree group awareness. (c) Design for reflection In the NIMIS perspective, two types of reflection in learning environments are considered. As an implicit result of the educational design of learning environments, learners have access to previous results and learning episodes as objects in the environment (e.g. through the visualisation of problem solving trajectories or through the provision of object repositories with versioning). This kind of reflection is an interactive process on the part of the human participants, such as learners, tutors or teachers. Secondly, certain types of analysis and interpretation on the part of the machine are also possible. The basis is a general architecture that includes history transcripts and plug-in facilities for intelligent monitoring and diagnosis [21]. However, we do not intend to build systems in which the learning process is under control of the machine; monitoring and analysis will provide local and partial feedback to learners or it can serve as decision aid for tutors or teachers.
230
11. Collaborative Mind Tools
(d) Priority of pedagogy over technology Certain technological scenarios might lead to more centralised control or to higher shares of individualised learning as opposed to partner or group work. Such potential changes in social and other aspects originating from the inherent logic of the technology without a clear pedagogical justification should not be accepted. The design of educational scenarios should in first place be based on pedagogical premises and objectives. (e) Consequences for teachers' roles and competence We believe that tomorrow's teachers will have to fulfil the role of classroom information managers. Already today, particularly primary school teachers act as managers of rich distributed classroom activities and of a variety of resources. With the help of advanced technologies, certain routine tasks such as the detection and correction of individual errors may be partially left to a computerised support system. This will enable teachers to concentrate even more on aspects of knowledge management and on supporting special needs. Given the ease of use of the new technology, there is no need for spending more efforts than today on system-specific ICT training for teachers. However, classroom information management will become a new prominent issue in teacher training and in teachers' professionalism. It will involve aspects of knowledge processing and representation, the design of learning materials and group scenarios for collaborative learning and new technology supported methods for reflection and analysis of classroom experience.
5. Recent experience and perspectives Within the European SEED project (IST-2000-25214), the Collide research group in Duisburg is currently testing new forms of using digital media in the classroom with a group of associated teachers. This activity is based on the premises that we accept the given curriculum and do not introduce new computer orientated content (1), that we do want to maintain, maybe enrich, each teacher's grown teaching style and preferences (2), but that, together with each of the teachers, we want to achieve a richer and more integrated form of using interactive digital media in the classroom (3). As for (3), our central focus is on the expressive and productive function of media as opposed to their container function in "content delivery". In these learning scenarios, we are using specific extensions of the Cool Modes environment, which have been designed together with teachers. One such example is a Cool Modes palette that supports the creation of experiments in stochastics using dice, lotto urns and other "generators". E.g., it also allows for representing the well known "birthday problem" (Fig. 4). The specific question is to determine the probability of finding at least two people with the same birthday in a group of size N (here: N=32). The experimental work can be arranged in small groups working in one environment or in synchronous shared environments. Results can be shared between groups by coupling container objects and result tables. Figure 5 illustrates the use of the stochastics micro-world in a probability course (mathematics) in 9th grade at the Elsa-Brandtsrom-Gymnasium, Oberhausen. One part of classroom allows for small groups working in a round table arrangement with embedded computer screens. Graphics tablets facilitate hands written annotations. The other half of
I I . Collaborative Mind Tools
231
Figure 4: "Birthday experiment" the classroom has a U shaped arrangement of tables around an electronic whiteboard with front projection. The network includes the computers in the round table working area as well as the "teacher's computer" on the whiteboard.
Figure 5: Modelling tools for stochastics Figure 6 was taken in the same classroom during a biology course in 12th grade on System Dynamics (exponential and logistic population growth, consumption of natural resources, interacting populations such as predator-prey models). In the concrete situation, the teacher works on a big interactive display using the Notelt free hand annotation tool. A scanned-in image of the development of coffee production in Brazil has been loaded into a Notelt page. The data show a periodic pattern and an increase over time. The teacher wanted to construct a linear approximation of this overall increase. Since Notelt does not provide parameterised geometrical shapes but only free hand input, he took a ruler designed for the chalkboard to draw a straight line on the electronic board. The result was perfectly OK, but the teacher articulated afterwards that he felt uneasy using the physical device as an add-on to the
232
11. Collaborative Mind Tools
Figure 6: Ad hoc use of an analogue device over a digital representation digital representation. He thought that his "fuzzy" way of achieving the goal was inferior to a "clean" computer operation. Yet, what is really bad about this blend of the digital and the physical-analogue? The result is on a digital level and thus maintains the full potential of being electronically archived, reused, distributed, multiplied and post-processed. As for the input process, a line-drawing operation would not necessarily have been better than this "brute force" method. The analogue device is clearly visible to the audience, and, as for the adjustment of the straight line, it offers more degrees of freedom than a computerised line drawing operation which usually requires fixing one point first or just allows for parallel movement of a given line. In all our previous experience, we had not seen this specific combination of digital and physical tools. From the point of view of media integration, this episode is a good example of crossing the physical-digital barrier. On a more general level, this is an example "digital mimicry" as the introduction of digital artefacts as a substitute of existing analogue media (e.g., an interactive whiteboard for a chalkboard or a digital piano for an acoustic one). In educational settings this enables teacher to transfer their professional skills without much loss. In further phases, the valueadding function of the digital technology (esp. media integration and re-use) has to be discovered and appropriated. In the sense of "ubiquitous computing" and "calm technology" [22], computers may get out of sight as explicit determinants of the learning situation. Still, inter-operability is more crucial than ever. Only through interoperability we can exploit value-adding function such as media integration and flexibly scaleable re-use. Not only must we design for multiple users with potentially different roles, we also have to consider that the activities will take place in open technical environments with various software and hardware components, which cannot be controlled in a predefined way.
11. Collaborative Mind Tools
233
Acknowledgements I want to express my gratitude and appreciation for the teachers who have creatively adopted our somewhat strange suggestions for transgressing the well known and accepted educational usage of computers. The outcome was not predictable but we feel encouraged by the results. Thanks also to all members of the COLLIDE research group for making this experience possible. The work reported here has been supported by the European Commission under the contracts ESPRIT 29301 (NIMIS) and IST-2000-25214 (SEED).
References [I]
Stahl, G. (to appear). Groupware goes to school: Adapting BSCW to the classroom. International Journal of Computer Applications Technology (IJCAT), 19. "Current approaches for groupware design, implementation and evaluation" (issues 3/4).
[2]
Dillenbourg, P., Jerman, P., Schneider, D., Traum, Buiu, C. (1997). The design of MOO agents: Implications from an empirical CSCW study. In B. du Boulay, R. Mizguchi (eds.), Artificial Intelligence in Education - Knowledge Media and Learning Systems (Proc. of AI-ED '97, Kobe, Japan, August 1997) (pp. 15-22). Amsterdam et al.: IOS Press / Omsha (IBN 90 5199 353 6).
[3]
Plotzner, R., Hoppe, H.U., Fehse, E.; Nolle, C., Tewissen, F. (1996). Model-based design of activity spaces for collaborative problem solving and learning. Proceedings ofEuro-AJED '96 (pp. 372-78), Lisbon, September 1996. Lisbon: Ed. Colibri (ISBN 972 8288 37 9).
[4]
Roschelle, J., Pea., R. (2003). A walk on the wild side: How wireless handhelds may change computer-supported collaborative learning. International Journal of Cognition and Technology. 1 / 1, 145-168.
[5]
Pinkwart, N., Hoppe, H.U., Milrad, M., Perez, J. (2003). Educational scenarios for the use of Personal Digital Assistants. Journal of Computer Assisted Learning. 19 (3), 383-391.
[6]
Dillenbourg, P. (1999). Introduction: What do you mean by "collaborative learning"? In Dillenbourg, P. (ed.), Collaborative Learning - Cognitive and Computational Approaches (pp. 1-19). Amsterdam: Elsevier (ISBN 0 08 043073 2).
[7]
Jonassen, D. (2000). Computers as mindtoolsfor schools: Engaging critical thinking (2nd edition). Upper Saddle River, NJ: Prentice-Hall.
[8]
Conklin, J., Begemann, M. L. (1987). gIBIS: A hypertext tool for team design deliberation. In Proceedings of Hypertext'87 (pp. 247-251). Chapel Hill, North Carolina.
[9]
Streitz, N., Haake, J., Hannemann, J., Lemke, A., Schuler, W., Schtttt, H. & Thiiring, M. (1992). SEPIA: A cooperative hypermedia authoring environment. In Proceedings of the 4th ACM Conference on Hypertext(ECHT'92) (pp. 11-22). Milano, Italy.
[10] Hoppe, H.U.; Gassner, K.; Miihlenbrock, M.; Tewissen, F. (2000). Distributed visual language environments for cooperation and learning - applications and intelligent support. Group Decision and Negotiation (Kluwer), 9 (3), 205-220. [ I I ] Suthers, D., Weiner, A., Connelly, J. & Paolucci, M. (1995). Belvedere: Engaging students in critical discussion of science and public policy issues. In Greer, J. (ed.,),
234
11. Collaborative Mind Tools
Proceedings of the 9th World Conference on Artificial Intelligence in Education (pp. 266-273). Washington DC, USA. [12] Joolingen, W. R. van (2000). Designing for collaborative discovery learning. In G. Gauthier, C. Frasson, K. VanLehn (eds.), Proceedings of 5th International Conference on Intelligent Tutoring Systems (pp. 202-211), Montreal (Canada). Berlin, Heidelberg: Springer (ISBN 3 540 67655 4).. [13] Pinkwart, N., Hoppe, U., GaBner, K. (2001). Integration of domain-specific elements into visual language based collaborative environments. In M.R.S. Borges, J.M. Haake, H.U. Hoppe (eds.), Proceedings of CRIWG 2001 (International Workshop on Groupware) (pp. 142-147). Los Alamitos, California: IEEE Computer Society Press (ISBN 07695 1351 4). [14] Nonaka, I. (1994). A Dynamic Theory of Organisational Knowledge Creation. Organization Science 5(1), 14-37. [15] Hoppe, H.U. & Plotzner, R.: Can analytic models support learning in groups? In Dillenbourg, P. (ed.), Collaborative Learning - Cognitive and Computational Approaches (pp. 147-168). Amsterdam: Elsevier (ISBN 0 08 043073 2). [16] Suthers, D., Hundhausen, C. (2003). An experimental study of the effects of representational guidance on collaborative learning processes. The Journal of the Learning Sciences 12(2), 183-218. [17] Forrester, J. W. (1968). Principles of Systems. Waltham, MA: Pegasus Communications. [18] Tewissen, F., Baloian, N., Hoppe, H.U., Reimberg, E. (2000). "MatchMaker": SynchronisingObjects in Replicated Software Architectures. In Proceedings of CRIWG 2000 (International Workshop on Groupware) (pp. 60-67). Los Alamitos, California: IEEE Computer Society Press (ISBN 0 7695 082). [19] Hoppe, H.U.; Lingnau, A.; Machado, I.; Paiva, A.; Prada, R.; Tewissen, F. (2000) Supporting collaborative activities in computer-integrated classrooms - the NIMIS approach. In Proceedings of CRIWG 2000 (International Workshop on Groupware) (pp. 94-101). Los Alamitos, California: IEEE Computer Society Press (ISBN 0 7695 082). [20] Negroponte, N. (1995). Being Digital. New York: Vintage Books. [21] Muhlenbrock, M., Tewissen, F., Hoppe, H.U. (1997). A framework system for intelligent support in open distributed learning environments. In B. du Boulay, R. Mizguchi (eds.), Artificial Intelligence in Education - Knowledge Media and Learning Systems (Proc. of AI-ED '97, Kobe, Japan, August 1997) (pp. 191-198). Amsterdam et al.: IOS Press / Omsha (IBN 90 5199 353 6). [22] Weiser, M. & Brown, J.S. (1997). The coming age of calm technology. In P.J. Denning, R.M. Metcalfe (eds.), Beyond Calculation - The Next Fifty Years of Computing (pp. 75-85) New York: Copernicus (Springer) (ISBN 0 387 94932 1).
I I . Collaborative Mind Tools
235
ULRICH HOPPE holds a full professorship for "Collaborative and Learning Support Systems" at Gerhard-Mercator-Universitat Duisburg, Germany. His research group COLLIDE has been engaged in several European projects in the area of advanced computational technologies in education. In 1998, COLLIDE initiated the NIMIS project on developing innovative classroom technology for early learning in the framework of a European program on "Experimental School Environments". The specialties of NIMIS are characterized by combining a computer-integrated classroom environment ("roomware") with new interaction techniques (pen-based input, speech output) and intelligent analysis and support. With an original background in mathematics and educational technology (Masterequivalent "Staatsexamen" in Mathematics and Physics from Marburg University, 1978; PhD in Educational Technology from Tubingen University in 1984), Ulrich Hoppe has been working for about ten years in the area of intelligent user interfaces and cognitive models in HCI (Fraunhofer Society Stuttgart, 1984-87, GMD IPSI Darmstadt 1987-95), before he re-focused his research on intelligent support in educational systems and distributed collaborative environments.
This page intentionally left blank
(N
238
"There are some kids, and some adults, who despite a massive amount of really interesting teaching just don't grasp the structure of numbers. If you do 101 minus 26 they do not have an idea about what it is. Is it a graphical thing? Is it a visual thing? Is it an action thing? And if you can't get these very basic ideas, about numbers, there is an enormous amount of things you just cannot do. There are a whole lot of careers that you just cannot have and a whole lot of life chances that are simply lost. It is a deep emotional problem too. I just feel that with all these wonderful technologies we ought to be able to help people with a new perspective so that they get new entry points for learning about numbers that are better suited to them. I also wonder whether we could learn more about the brain through neuroscience that could help us?" Celia Hoyles
A Learning Zone of One's Own M. Tokoro and L. Steels (Eds.) IOS Press, 2004
239
12.
Playgrounds for A New Mathematics1 Celia Hoyles
New computational forms are pervading the social and economic lives of individuals and nations alike. The notation systems or representational infrastructures we use to present and re-present our thoughts to ourselves and to others, and to support reasoning and computation, shape what can be known and how it can be known. A central challenge of mathematical learning for educators is therefore to exploit these new systems to create a more learnable mathematics.
1. The role of computational media and the separation of outcome from process As recounted in Shaffer & Kaput (1999), the development of computational media required three elements: the existence of discrete notations without fixed reference fields (that is, the idea of formalism), the creation of syntactically coherent rules of transformation on such notations, and a physical medium in which to instantiate these transformations outside the human cortex and apart from human physical actions. Hence in the 20th century a profound shift occurred, from operable notation systems requiring a suitably trained human partner for execution of the operations, to systems that run autonomously of a human partner. The devolution of execution to the machines means more than this: not only do the machines now do mathematical execution, it implies that any consequential appreciation of what the machines do must itself be based on mathematical principles. If an individual does not have the means formally to relate his or her intellectual model of the mathematical principles with those inside the machine, then appreciation of the model must necessarily be partial. Of course, this does not mean that such models need to be expressed in the same languages as used inside the machine. Quite the reverse. It means that we have to find
Some parts of this paper are adapted from Kaput, Noss R, & Hoyles (2002)
240
12. Playgrounds for A New Mathematics
ways to help people to capture the dynamics or the mechanism of the system, so they can follow the consequences of particular actions while maintaining a realistic sense of the structures of relationships between them. The representational system in which the structures are expressed is usually transparent to an expert. But when constructing something new, a learner needs to be aware of the processes of construction and to think explicitly about the representational system itself—it needs to be simultaneously both transparent and opaque. This 'co-ordinated transparency' (Hancock, 1995) represents a synthesis of meaning and mechanism, a situation in which fluency with and within the medium can temporarily be replaced by a conscious awareness of its (usually invisible) internal structures. We share a vision of a mathematics curriculum that assumes mathematical understanding should be built around the construction and interpretation of underlying mechanisms, the quantitative and semi-quantitative models that describe phenomena in the 'real' world or the world of emerging mathematical ideas, where students explore mathematical technologies and analyse processes in contexts that show how they can be used and in particular why they work in the way they do (see Hoyles, Morgan, Wodhouse, 1999). We now turn to two examples to illustrate the synergy of knowledge and representational infrastructure and how both are shaped by the available technology.
2. Developing a sense of mechanism We focus here on a corpus of work from the Playground project (see www.ioe.ac.uk/playground) which set out to explore new ways to express mathematical relationships, bringing children into contact with mathematized descriptions of their realities at ages much younger than we would normally countenance with static technologies. Our central focus was to open possibilities for children (aged 4 to 8) to design, construct and share their own video games. We designed computational environments for children to build and modify games using the formalisation of rules as creative tools in the constructive process. They built their own executable representations of relationships—in effect, they were what we would describe as programming. We called these environments 'Playgrounds'. We worked with two programming systems, ToonTalk—an animated programming language (Kahn, 1999)—and Imagine, a concurrent object-oriented variant of the Logo programming language (Blaho et al, 1999: note, at this point, the language was named "OpenLogo"). Each of our two Playgrounds represented a layer we built on top of these platforms, incorporating elements that allowed multiple entry points into the ideas of formalising rules. Here, 1 concentrate on our work with ToonTalk. Our objective had a strong epistemological rationale. The challenge was to find ways for young children to use non-textual means to express and explore the knowledge which underpins the genre of video games: what it means for objects to collide, how 2-dimensional motion of an object (or a mouse, or a joystick) can be thought about, the construction of animation, and the hundreds of little pieces of knowledge which go to make up the workings of video games. We saw this as an instantiation of the much broader class of knowledge, 'developing a sense of mechanism'.
12. Playgrounds for A New Mathematics
241
Our choice of video games built on established work by, for example, Kafai (1995) in that it was a particularly attractive platform for many children. ToonTalk is a world in which animations themselves are the source code of the language; that is, programs are created by directly manipulating animated characters, and programming is by example (see Cypher, 1993). A full description can be found in Kahn (1999). Robots are trained to carry out tasks inside houses (defining the body of a method). A user trains a robot by entering its "thought bubble" and controlling it to work on concrete values (see Figure 1). The robot remembers the actions in a manner that can easily be abstracted to apply in other contexts by later removing detail from the robot's thought bubble. Message passing between methods (robots) is represented as a bird taking a message to her nest.
Figure 1. A robot is trained to add one value to another
The nature of the platform is paramount. Our choice of ToonTalk implied that any layers we built above it had to mesh with the metaphors of the platform. Our aim was to design a permeable abstraction barrier between ready-made pieces of open code with multi-modal representations (we called these 'behaviours': some examples are given below) and the ToonTalk language itself lying underneath. This stands in marked contrast to some modern programming languages such as Java and C++ which by default enforce these abstraction barriers and do not allow programmers using predefined objects to discover their underlying implementation. But the crucial dimension dictating the design of the playground layer was that of openness. At any level of granularity, an element should be decomposable into smaller pieces down to the lowest level of the animated ToonTalk programming language. Indeed as we began to see children decomposing the games and sharing their parts across sites and countries, it became clearer that we were working in a design paradigm akin to component software architecture (CSA). While some (but not all) of the component community are concerned to a greater or lesser extent with the adaptability of their components, for us it was central. We are concerned with designing software for investigating mechanism; individual components therefore need to have intuitive windows
242
12. Playgrounds for A New Mathematics
into their workings and a means for modification. (For more information on the role of behaviours in playgrounds, see Hoyles, Noss and Adamson, 2002) In the design of an environment where the opening of mechanisms is the primary objective, it is desirable not only that pieces are easily opened but also that they afford access to their workings through an intuitive interface. In traditional CSA, the user interacts with the interface model provided by the architecture but not the implementation of individual components. In our open component model, we require both. Users should be able to work at several levels simultaneously: a) composing components where necessary as wholes relying solely on the interface for component manipulation and b) opening a component to reveal the source code whenever modification or inspection of the component is desired. To facilitate this, we need to ensure that components inter-operate at a technical level and also that manipulation at interface and implementation levels is made intuitive by a high degree of semantic interoperability. In other words, if users are to use, share and manipulate components in the construction of larger pieces of software, consistency of interface and multiple ways of accessing the functionality become important criteria in their design. At a general level, therefore, our challenge was to find ways to give ready-to-use 'code' to children while simultaneously leaving open both the code itself (how it worked) and the way in which it fitted with the platform level (how it inter-operated with other bits of code). But this raised a further, more subtle challenge. In accessing different levels of the 'playground', children would interact in different ways with the system; and different children would prefer to interact at different levels and in different directions. The design process was therefore, not only complex, but also necessarily iterative, in that we needed to try to mesh our design with our emerging knowledge of what children could do with it.
3. The design of behaviours
At the simplest level, a behaviour is a ready-made piece of programming code that can be (re-)used, inspected, or combined. As summarised above there were three principal rationales which prompted us to design behaviours, and which led us to recognise the importance of an intermediate level between raw programming code and higher level constructions in the Playground. They are: 1) a requirement for portable re-usable code that can be shared across collaborating communities 2) the provision of dynamic representations of the code's functionality 3) the combination of representation and functionality in single elements. These rationales led us to describe the three main features that we demanded from our behaviours. Visible functionality: it is easy to see what things do, and equally, to see how they do it. As an illustration of what we mean, consider the following two levels of exactly the same code: • At the level of mechanism, a robot waits for his box to contain a number and the text "yes" and then adds 1 to the number (and then waits again). • At the level of behaviour, the picture moves up so long as the shift button is held down.
12. Playgrounds for A New Mathematics
243
We tried to interpret visibility in the broadest sense, by representing the behaviours of objects in multiple ways, and we incorporated visual cues aiming to make clear what did what by labelling them using multi-modal representations of an object's characteristics and functionality; that is visually, textually, aurally and by using text-to-speech mechanisms, as illustrated in Figure 2.
Figure 2: A behaviour.
Interoperability: This refers not only to the interface and technical aspects of intercommunication between objects but also to the ability of a user to use code pieces from different sources. The former we refer to as technical interoperability and the latter as semantic interoperability. We needed to make it easy to inspect processes, modify them, and import and export them to other objects. Although we aimed for much more than merely hooking pieces of code together, or inserting them in new locations, we did regard this as a crucial activity. In fact, our first task was to modify ToonTalk so that we had a clear and consistent mechanism for importing and exporting robots (programs) allowing the construction of a common language that could be used across games, classrooms and across the net, providing a high degree of semantic interoperability. To achieve this goal, modularity of design became a major aesthetic which assisted us in this respect. In the example below (see Figure 3), the behaviours on the back of a bouncing ball can be simply picked up and placed on the back of a mouse picture, the mouse then will behave exactly as the ball did.
244
12. Playgrounds for A New Mathematics
Figure 3: Illustrating interoperability Permeability: this is shorthand for the permeable abstraction barrier referred to earlier: our goal was to make it as easy as possible to jump up or down levels of abstraction, in terms of the level of detail with which it is necessary to interact, the self-containedness of a piece of code, and the extent to which it is usable in other contexts. This permeability between layers meant that while it was possible to stay at one level and construct games through combining behaviours, the next level down was always available for inspection and change. Several factors contribute to this functionality. Firstly, behaviours are part of the object they control; secondly, behaviours are constructed to be modular. In Figure 4, the behaviour 'When I hit something, I make a sound' is clearly represented, and separated from the other behaviours.
Figure 4: Modularisation of behaviours
Finally, the lowest level in the system - the ToonTalk code itself - is always exposed.
12. Playgrounds for A New Mathematics
245
It might be that some readers will be wondering what this has to do with mathematics. Our reply is that building video games requires articulating how objects interact; how rules are expressed formally and their implications, and that this is a central aspect of what it means to think mathematically in the computational era. At a more detailed level, this claim breaks down into two sub-claims. First, it is about children learning that there are rules with implications for what is modelled and what is observed, and these rules are something over which they have some control. Second, these rules embody and are built upon a previouslyconstructed representational infrastructure that offers extraordinary power to those who master it. Over the Playground project we collected numerous examples of children collaboratively taking apart a scene and exploring how it worked, why it worked and how it could be changed. Often, a teacher was involved: in fact, a key aspect of the claim is that such an approach was more teachable than other programming environments, as the things that matter are visible and can easily be manipulated and the granularity of the pieces could be customised by the teacher for the learner. All this was true and exciting, but a year or so after the Playground project, I reflect upon the huge potential of the work, the rich data we collected and analysed to illustrate this potential, but - to be frank- the rather little impact this work has had on the future of learning beyond the boundaries of the small community of school and students that formed part of our project. We have to learn from our successes but analyse the limitations of our work in so far as they influence mainstream learning for most children. Maybe we were just 'too far' from what could be acceptable in schools; dare we say just too ahead of our time?
4. Geometrical Modelling I now turn to what might appear to be a more conventional mathematics problem to again illustrate the interacting knowledge and the representational infrastructure instantiated with available technology. A British mathematics textbook presents the following problem (Figure 5). In the game of rugby, points are scored by carrying the ball across the "try line" defended by the opposing team. Suppose this happens at point C. Then, extra points may be scored (the try is "converted") if the ball is successfully kicked between the goal posts A and B; for this kick, the ball can be placed at any point K on the "kick line" through C perpendicular to the try line. The problem asks "where should the kick be taken from?", and translated into conventional mathematical terms this becomes "for what choice of K is the angle AKB at a maximum?" In di Sessa, Hoyles and Noss (1995) we described how at a workshop, the participants—all expert with different types of software, including Logo, StarLogo and dynamic geometry— were asked to solve this problem by whatever means they chose. The idea was to try to expose the kinds of mathematical thinking which different people chose to express in different software.
12. Playgrounds for A New Mathematics
246
c
A
B try line
K kick line Figure 5. The rugby kick problem.
What was this problem about? The most common view was that it had to do with maximising angles. But this interpretation did not map onto a unique set of mathematical techniques. The group who worked with paper and pencil adopted the familiar strategy of drawing diagrams, putting in unknowns and setting up equations. They did not experience the need to use the computer as scaffolding for their work: they knew an approach and were content to stick with what they perceived as a standard maximisation exercise (express AKB as a function of position K, and find the stationary point of this function using calculus). Try line
Figure 6. A way to parameterise the Rugby problem
The way the problem was to be modelled was instantiated in a process of labelling that made explicit what was to be ignored and what was to be the focus of attention - the need to find a relationship between 0 and d (see figure 6). After this step, the solution, d2 = p(p+l) for maximum 0, was relatively straightforward — assuming knowledge of, and skill at manipulating algebra and calculus tools! Some used Cabri Geometry, and read the problem as a geometrical one: therein lurks Euclid's "subtended angle" theorem, and the solution is given by finding the circle through A, and B such that the kick line is a tangent to it—a straightforward task for dynamic geometry (Noss & Hoyles 1996, pp. 241-43). Thus those who chose to use Cabri activated
12. Playgrounds for A New Mathematics
247
a web of connections to geometrical objects and their properties. Their problem was concretised by these meanings as they become operational in the Cabri setting. We observe that not only was the workshop fascinating to the participants at the time but also that the reflections on the workshop and its differing outcomes became the source of many fascinating accounts. Mitch Resnick and Uri Wilensky decided to work on the same problem using their favourite tool, StarLogo. StarLogo comprises thousands of turtles, each of which can control a computational process with the result that multiple processes can be controlled in parallel. There are many features of StarLogo but of principle interest here is that turtles can detect and distinguish other turtles along with certain features of the world over which they move. The StarLogo group, rather than positioning the rugby problem in geometry or calculus, connected it to probability and statistics — a matter of personal preference perhaps, but a preference surely structured by the design of the system they wanted to use. The story of what happened next is taken up by Resnick: "We imagined hundreds of rugby players standing at different points along the line perpendicular to the try line, and we imagined each player kicking hundreds of balls in random directions. To find the point X that maximises the angle AXB, we simply needed to figure out which of the hundreds of rugby players scored the most conversions (by successfully kicking balls between the goal posts A and B). The reason is clear: If each player is kicking balls in random directions, the player with the largest AXB angle will score the most conversions. This strategy is an example of what is sometimes called a Monte Carlo approach. It was quite easy to write the StarLogo program to implement this strategy. We used turtles to represent the rugby balls. To start, the program put thousands of turtles/balls with random headings at random positions along the perpendicular line. Then, the program 'kicked' all of these turtles/balls, moving them forward in straight lines. Finally, the program took all of the turtles that successfully went through the goal posts and moved them back to their starting points on the perpendicular line. The point with the most surviving turtles is the point that maximises the angle AXB ". (Resnick, 1995, pp. 39-40) What a different view of the problem and one that stunned the workshop with its simplicity! Clearly this solution draws on a very different web of mathematical ideas than the previous two. It is also more closely connected to the original rugby setting, a setting which both the other groups had suppressed the moment they had mathematised the problem by isolating the relevant variables, or locating helpful theorems. Perhaps a more general solution too: Resnick points out, 'If extra constraints were added to the problem, such as a wind blowing across the field, or a limitation on the distance a rugby player can kick a ball, it would be quite easy to adjust the StarLogo program to take the new constraints into account. It would be more difficult to adjust traditional geometric analyses' (ibid. p. 40). From a conventional mathematical viewpoint, this probabilistic approach lacks generality, because it does not produce a formula for the distance CK in terms of the variables in the problem (the distances CA and AB). But, Wilensky argues:
248
12. Playgrounds for A New Mathematics
" (...) the notion of mathematical generality itself is relative both to the underlying supporting technologies, the potential contexts of application and the social purpose accorded the mathematics education. In a computation rich world, is an algebraic formula more general than an algorithm? Is an analytic solution more general than a probabilistic approach? •Seeing this problem as a school mathematics problem ... causes us to immediately ignore the real world conditions of the problem [e.g. effects of wind, variable strength of the kicks]... The analytic solution is brittle, it works only under the idealised conditions of the problem. In contrast, the modelling solution can be easily modified to take account of [different real world conditions] and seeing the different results ". (Wilensky, 1996)
Though Wilensky's general point has force, this probabilistic approach by itself cannot answer questions like, "why is there a position of maximum subtended angle? how does this position vary with the relative positions of C, A and B?" It will solve the problem and provide insight into a method of solution without explaining - but does the first method explain? What can we learn from these three solutions? The first point is that it would seem clearly ridiculous to order the solutions and the competencies displayed in any hierarchy of mathematical sophistication; in each case different connections are made, different meanings are evoked, each coherent and defensible within their own framework. Why was this activity so successful in the workshop? Because the computer focused our attention on the diversity of knowledge, style and solution. There was also a strong commonality lurking in the subtle domain of feelings and aspirations. Everyone was open to, respected, and was fascinated by the different approaches and epistemological revisions; everyone held the same deeply-rooted affective response to mathematics based on the notions of challenge, empowerment, and enjoyment; everyone wanted to share and compare ideas in an open way either in their construction or in their communication; and everyone adopted a playful approach, tinkering with and experimenting in a culture of building tools to help in the process of conceptualisation. The environment was genuinely collaborative and constructive but also energised by tremendous tenacity tinged with the spice of competition - some worked all night on their solutions, determined to show the power of their preferred software! We do not know the extent to which these affective factors determined the diversity of approaches, but we are fairly sure that their absence would at least have dampened them. The choice of medium clearly mediated the range of meanings and connections likely to structure the interaction, and likely to emanate from it. The choice was not arbitrary but determined as much by familiarity and expertise as by suitability. Papert's (1996) later discussion of the "rugby problem" situated it within the larger context of an exploration of the conventional ("school maths") and alternative domains that lie in the "space of mathematics educations". He is emphatic that: "This episode is not primarily about software tools. It is about the status of ideas on an empowerment/disempowerment dimension". (ibid, p. 113)
1 2. Playgrounds for A New Mathematics
249
Papert contrasts the Cabri and StarLogo approaches with that of his own devising: "[the StarLogo method] is quintessentially computational but leaves one with an answer that is mathematically unexplained and poorly connected to geometrically related situations ... [in the Cabri method] the computer is used to find an answer in a pre-computational conceptual space. In both cases the computer used as a tool effectively leads to a solution but in neither does the computational environment make the mathematics more perspicuous... The goal is to use computational thinking to forge ideas that are at least as "explicative" as the Euclid-like constructions (and hopefully more so) but more accessible and more powerful, (ibid, p. 115-6) My personal approach to the problem ... took the form of a "heuristic drive" to break my thinking out of the [kick line] and to visualise the subtended angle over the whole plane. My sense of connection was to potential fields and to questions about visualising them by devices such as drawing equipotential contour lines ". (ibid, p. 114) So Papert devises a little program (in Logo), to repeatedly choose random points in the plane of the "rugby pitch" and to colour each point on a spectrum according to magnitude of the angle subtended there from AB: "the contour lines slowly emerged like a developing Polaroid photo. Only a few minutes were needed to recognise the emergence of circles; after ten minutes the graph was crisp and visually as well as conceptually pleasing". (ibid, p. To reiterate Papert' s powerful insight in my own words: the way we (researchers, teachers and students approach a problem is deeply shaped by salient ideas in mathematics at the time. These ideas do of course develop salience in ways that shape and are shaped by the emergence of new tools: that is if we (schools, teachers and policy-makers) allow them to do so.
5. Conclusions In this chapter I have attempted to show how mathematicians and mathematics educators are beginning to turn their attention to defining newly empowering representational infrastructures for learning. In the past, beginning with writing itself (Kaput, 2000) more powerful representational infrastructures have been a source of intellectual and mathematical power, but at a cost of learnability and hence access. Hence they tended to remain the province of an elite minority who were inducted into their use. New computational media offer the opportunity to create democratising infrastructures which will redefine school mathematics (for a fuller discussion of these issues, see Noss and Hoyles, 1996). Viewed optimistically, these will exploit the processing power of the new media while at the same time ensuring that students maintain an intuitive feel of the central knowledge elements at work and how they relate to each other. Yet if the power and potential of computers are to be exploited in school mathematics, attention must be paid to this level of representational infrastructure.
250
12. Playgrounds for A New Mathematics
Our challenge is to focus attention on the design and use of representational infrastructures that intimately link to students' personal experience. This is a necessary step if we are to move away from a 19th century school mathematics concentrating on isolated skills based on static representational systems in a tightly-defined curriculum (with only a minority able to engage in independent problem solving). To achieve our goal we need to undertake systematic research into the potentials and pitfalls of incorporating these new structures into our curriculum. We must acknowledge and exploit the 'overhead' of introducing technology into teaching and learning mathematics - rather than ignoring it as so often has been the case in the past. We must also not underestimate the challenges that will have to be faced even to begin to put these ideas into practice in a small way let alone one that is more systemic. Our vision of future learning is that children will master new kinds of knowledge by exploring how they perceive that things work, and how (roughly) to build a model of their developing knowledge, interpret the outcome of a model and change a model. By this means they will learn that phenomena have complex and interrelated causes and relationships. Most crucially, by designing tasks and activities for teachers and students to engage in exploring and debugging mathematical ideas we might offer a more accessible, interesting epistemology for all.
References Blaho, A., Kalas, I. and Tomcsanyi, P. (1999). OpenLogo—a new implementation of Logo. In: Nikolov, R., Sendova E., Nikolova, I. and Derzhanski, I. (1999) Proceedings of the Seventh European Logo Conference. 95-102. Cypher, A. (ed)(1993). Watch What I Do: programming by demonstration. Cambridge, MA: MIT Press. DiSessa, A. (2000). Changing minds, computers, learning and literacy. Cambridge, MA: MIT Press. Hancock, C. (1995). The medium and the curriculum: reflections on transparent tools and tacid mathematics. In DiSessa, A.A., Hoyles, C. & Noss, R. (eds) Computers and Exploratory Learning. Berlin Heidelberg: Springer-Verlag. pp. 221-241. Hoyles, C., Morgan, C. & Woodhouse, G. (Eds). (1999). Rethinking the mathematics curriculum. London: Palmer Press. Hoyles, C., Noss, R. and Adamson, R. (2002) 'Rethinking the Microworld Idea. Journal of Educational Computing Research, 27, 1&2, 29-53. Kafai, Y.B. (1995). Minds in Play: Computer game design as a context for children's learning. Lawrence Erlbaum Associates.
12. Playgrounds for A New Mathematics
251
Kahn, K. (1999). Helping children learn hard things: computer programming with familiar objects and activities. In Druin, A. (ed) The Design of Children's Technology. San Francisco: Morgan Kaufman Publishers Inc. 223-241. Kaput, J. (2000). Implications of the shift from isolated, expensive technology to connected, inexpensive, ubiquitous, and diverse technologies. In M. O. Thomas (Ed.) TIME 2000: An international conference in Mathematics Education, (pp. 1—25). University of Auckland, New Zealand. Kaput, J, Noss, R. & Hoyles, C. (2002). Developing New Notations for a Learnable Mathematics in the Computation Era. In English, L (eds) Handbook of International Research in Mathematics Education, Lawrence Erlbaum. Chapt.4 pp. 51-75 Noss, R., & Hoyles, C. (1996). The visibility of meanings: modeling the mathematics of banking. International Journal of Computers for Mathematical Learning 1,1( 17), 3-31. Noss, R. & Hoyles, C. (1996). Windows on Mathematical Meanings: Learning Cultures and Computers. Dordrecht: Kluwer. Papert, S. (1996). An Exploration in the Space of Mathematics Educations. International Journal of Computers for Mathematical Learning 7,7(17), 95- 123. Resnick, M. (1995). New paradigms for Computing, New Paradigms for thinking in In DiSessa, A.A., Hoyles, C. & Noss, R. (eds) Computers and Exploratory Learning. Berlin Heidelberg: Springer-Verlag. pp. 31-44 Shaffer, D. & Kaput, J. (1999). Mathematics and virtual culture: An evolutionary perspective on technology and mathematics education, Educational Studies in Mathematics, 37, 97-119. Wilensky, U (1996). Modelling Rugby: Kick First. Generalize later International Journal of Computers for Mathematical Learning 7,7(17), pp 125- 131
252
12. Playgrounds for A New Mathematics
CELIA HOYLES has been Professor of Mathematics Education at the Institute of Education, University of London since 1984. She was awarded the top first class honours Mathematics degree in 1967 at Manchester University, and her PhD in Mathematics Education in 1980. Hoyles has been the director of over a dozen research projects concerned with primary and secondary mathematics, particularly in relation to the use of computers, students' conceptions of proving and proof, and mathematics used in the workplace. She was a member of the ESRC Research Grants Board 1994-1998 and was elected Chair of the Joint Mathematical Society of the United Kingdom in October 1999. She is a member of the Advisory Committee on Mathematics Education that speaks for the mathematics community to U.K. Government on policy matters. She has been serving as Dean of Research and Consultancy at her university since June 2002.
13
254
"Computer programming is a new way to extend thinking. Just as paper and pencil enable you to think differently by providing an extension of your memory and by providing a way to organize and communicate your thoughts in ways that are fundamentally different compared to pure verbal communication, computational media provide new ways of thinking. You don't merely express an idea in computer language, but the computer can realize the idea turning your ideas into running simulations, interactions, games, algorithmic computations, or whatever." Ken Kahn
A Learning Zone of One's Own M. Tokoro and L Steels (Eds.) IOS Press, 2004
255
13.
ToonTalk - Powerful Building Blocks for Computer-Based Learning Environments Ken Kahn
Many of the attributes of good computer-based learning environments such as support for collaboration, creativity, and community are widely agreed upon. Interfaces should be seamless. Content should be challenging without being frustrating. Authoring support should be flexible and powerful. But what about universality in learning environments? Should learning environments support the specification and execution of arbitrary computations? And if one does support universality, should it be for more than just expert users? I argue here that universality for all is both desirable and attainable. Furthermore, by universality I mean more than a theoretic equivalence to a Turing Machine, but rather an elegant and cognitively appropriate model of general computations. The classical notion of universality only addresses the ability to compute the values of functions, ignoring the ability to effectively use the input and output devices typically associated with desktop computers. Ideally, learners should be able to describe computations, including those involving the display, mouse, keyboard, or sound card. And they should be able to describe those computations in a language that is well-suited to their cognitive abilities. A computer becomes whatever software instructs it to be. A learning environment that does not restrict the range of software that can be created and run is able to exploit all that a computer can be. Most learners can find something that is personally compelling among the literally millions of different kinds of things a computer can be. Software is a fundamentally new kind of medium. It is a medium where one expresses ideas that are given life by machines called computers. Ideally, educational software should be transparent. To be transparent, software needs to be composed of modules that can be understood and changed by students as well as authors. Previous attempts to provide universal universality (i.e., fully general computational tools for everyone) have had limited success. Logo [Papert 1980] and Smalltalk [Kay 81] both strive to be general purpose programming languages for all. To
256
13. ToonTalk
master these languages or their successors (e.g. Boxer [Boxer 02], StarLogo [Resnick 1997], and SqueakfSqueak 02]) requires a set of skills that many students fail to acquire. To program in these languages one needs to be fluent in a set of computational abstractions (e.g. procedures, variables, and conditionals). And programs are constructed in the realm of abstract variables and not concrete values. Everyone, however, can master tools capable of constructing arbitrary computations. (By "everyone" I mean students who are capable of learning other school subjects.) ToonTalk ([Kahn 96], [Kahn 02]) is a programming environment that greatly lowers the difficulty of programming without sacrificing power or expressibility. It does this by replacing computational abstractions with tangible concrete analogies and by supporting the ability to program with examples and subsequently remove details to obtain generality.
1 A Brief Introduction to ToonTalk ToonTalk started with the idea that perhaps animation and computer game technology might make programming easier to learn and do (and be more fun). Instead of typing textual programs into a computer, or even using a mouse to construct pictorial programs, ToonTalk allows real, advanced programming to be done from inside a virtual animated interactive world. The ToonTalk world resembles a modern city. There are helicopters, trucks, houses, streets, bike pumps, toolboxes, hand-held vacuums, boxes, and robots. Wildlife is limited to birds and their nests. This is just one of many consistent themes that could underlie a programming system like ToonTalk. A space theme with shuttlecraft, teleporters, and so on, would work as well, as would a medieval magical theme or an Alice in Wonderland theme. The user of ToonTalk is a character in an animated world. She starts off flying a helicopter over the city. (See Figure 1.) After landing she controls an on-screen persona. The persona is followed by a dog-like toolbox full of useful things. (See Figure 2.)
Figure 1 - Flying over the City
Figure 2 - Followed by the Toolbox
13. ToonTalk
257
The ToonTalk city is where all computations take place. Most of the action in ToonTalk takes place in houses. Homing pigeon-like birds provide communication between houses. Birds are given things, fly to their nest, leave them there, and fly back. Typically, houses contain robots that have been trained to accomplish some small task. A robot is trained by entering his "thought bubble" and showing him what to do. Robots remember actions in a manner that can easily be generalised so they can be applied in a wide variety of contexts. (See Figure 3.)
Figure 3 - Training a robot to double a number
258
13. ToonTalk
2 Abstract versus Concrete A robot behaves exactly as the programmer trained him. In computer science terms, this training corresponds to defining the body of a method in an object-oriented programming language such as Java or Smalltalk. A robot can be trained to send a message by giving a box or pad to a bird; spawn a new process by dropping a box and a team of robots into a truck (which drives off to build a new house); perform simple primitive operations such as addition or multiplication by building a stack of numbers (which are combined by a small mouse with a big hammer); copy an item by using a magician's wand; - change a data structure by taking items out of a box and dropping in new ones; or terminate a process by setting off a bomb. Computational Abstraction
ToonTalk Concreteness
computation
city
a running program
actor, process, or concurrent house or back of picture object an independent activity or behaviour method or clause
robot
the smallest coherent program fragment guard or method preconditions
thought bubble
conditions before running a pro gram fragment method actions or body
actions taught to a robot
a sequence of actions message or array or vector a container of items
box
13. ToonTalk
comparison test
259
set of scales
testing if something is more than something else process spawning
loaded truck
the creation of a new activity
process termination
bomb
the termination of an activity
constants
number, text, picture
basic elements
15T992! Toon.Ta.Hci
channel transmit capability or bird message sending a way to send messages channel receive capability or nest message receiving a way to receive messages persistent storage or file
notebook
a place to store things permanently 1
Salily 2
Table 1 - Computer Science Terms and ToonTalk Equivalents The fundamental idea behind ToonTalk is to replace computational abstractions by concrete familiar objects. Even young children quickly learn the behaviour of objects in ToonTalk. A truck, for example, can be loaded with a box and some robots. (See Figure 4.) The truck will then drive off, and the crew inside will build a house. The robots will be put in the new house and given the box to work on. This is how children understand trucks. Computer scientists understand trucks as a way of expressing the creation of computational processes or tasks.
260
13. ToonTalk
Figure 4 - A truck being loaded with robots and a box 3. ToonTalk as a Learning Environment As Papert [Papert 1980] and others have observed, computer programming can be a fertile ground for learning general thinking skills. These include problem decomposition, component composition, explicit representation, abstraction, debugging, and thinking about thinking. In addition, educators can build levels of software upon ToonTalk to help students learn other subjects. The Playground Project [Playground 01] explored the use of ToonTalk to enable young children to build their own video games. In doing so the children acquired an understanding of the mechanics of a complex artefact (a computer game). They learned how to play creatively with rules to make their own games. The children became good at taking apart a computational entity and creatively reassembling it in new ways, sometimes combining parts from different games. This work relied upon the fact that things built in ToonTalk are modular and transparent. A child, for example, can take a simple Ping Pong game, remove the bouncing ball, and flip it over to see the robots that give the ball its behaviour. If structured well, the ball's behaviour is composed of sub-behaviours for bouncing off of other objects, for bouncing off the edges of the game area, and for its initial movement. A child can then find a scorekeeping mechanism, watch how its robots work, and change it to reflect the way she wants the score to be kept in her game. The modified scorekeeping program can then be added to the back of the ball. She then may decide her game should have bouncing fish and place all of these behaviours as a component on the back of a picture of a fish. If she wants lots of fish in her game she can copy the fish and the behaviours on the back are copied as well. She then need only drop her fish on an appropriate background picture. The Playground Project also explored collaborative game design. Children were able to exchange games with other children (even in other countries) simply by giving their games to ToonTalk birds whose nests are on other computers. The birds then deliver the games (and associated messages) to the other children. These children then play, analyse, modify and review the games and send them back in the same manner. A new three-year European research project began September 2002 called WebLabs. Its goal is to repeat the success of the Playground Project with a focus on science and math instead
13. ToonTalk
261
of computer games. ToonTalk is providing the infrastructure for building "transparent modules" for learning about specific topics like gravity, randomness, robotics, infinity, or bouncing. Children are using these modules to create their own simulations and games to explore these scientific topics. The children then embed their ToonTalk creations into active essays [Resnick 96] that are web reports that include runnable programs. They then publish their reports to facilitate collaboration and peer review.
4 Learning ToonTalk Among the subjects that can be learned while inside ToonTalk is ToonTalk itself. Receiving expert instruction from a teacher or a textbook is just one way to learn ToonTalk programming. ToonTalk provides four fundamentally different ways for children to learn ToonTalk on their own: 1. Free play. An open-ended, unconstrained, rich environment to explore and create things. 2. A puzzle game. A sequence of puzzles that gradually introduces the elements of ToonTalk and techniques for building programs. 3. Pictorial instructions. Sequences of pictures that show how to build programs. 4. Demos. Narrated demos showing various elements of ToonTalk and construction techniques. A Safe and Self-revealing environment Proponents of constructivism [Papert 1993] argue well for the position that the best, deepest, longest-lasting learning happens when the learner discovers and constructs the knowledge herself. ToonTalk has a "free play" mode designed to accommodate this kind of learning. Exploratory learning is best supported by an environment that is safe and self-revealing. An environment is safe to explore if novice actions will not cause any permanent damage. For example, in ToonTalk there is a character named Dusty that acts like a hand-held vacuum. A beginner exploring ToonTalk might pick up Dusty and vacuum up something important. However, Dusty doesn't destroy things, and he can be used in reverse to spit out all the things he has ever vacuumed up. A self-revealing environment is designed so that an inquisitive explorer can discover what objects exist and how they behave. ToonTalk, for example, contains boxes. Even very small children discover on their own how to move boxes, how to put things into them, and how to take things out of boxes. Good animation and sound effects help greatly in making an environment self-revealing. If a user holds something over an empty compartment of a ToonTalk box, she sees that part of
262
13. ToonTalk
the box wiggle and change colour in anticipation. If she then clicks the mouse button, she sees an animation of the item leaving her persona's hand and falling into the compartment and hears an appropriate sound effect. If a force-feedback joystick is connected to the computer, she even feels the weight and other properties of objects. It is very difficult to make a completely self-revealing environment. For example, in ToonTalk, boxes can be joined together and broken apart by actions that are often not discovered by children on their own. To completely explore ToonTalk, children need help. ToonTalk includes an animated talking guide or coach named Marty to help a child explore ToonTalk. (See Figure 5.) Marty keeps track of what actions a user has performed. He also is aware of what item a user is holding or pointing to and tries to suggest an appropriate action in the current context. For example, a child holding a box who has put things in and out of boxes but hasn't joined two boxes together will hear a suggestion from Marty about how to join boxes. Some children send Marty away, preferring to explore without any help. Others can be seen trying Marty's suggestions one after another. Children react to Marty differently depending upon whether he communicates by talk balloons, as in comics, or uses a text-to-speech engine to actually speak. For some children, reading is a slow and burdensome task. Ideally, a self-revealing environment should also be incremental. An incremental environment may feel open-ended and rich but is designed so that certain objects or actions can be discovered only after others have been mastered. This helps reduce confusion and frustration that often results from the initial explorations of a rich and complex environment. Popular video games such as Nintendo's Super Mario Brothers® are excellent examples of incremental self-revealing environments. When a player starts these games she finds herself controlling an on-screen character. Initially all she needs to do is move. Soon she sees some coins and by walking into them they are acquired. Soon after there are coins that are not reachable without jumping, and the player experiments with a small set of buttons on the controller to discover how to jump to get those coins. After hours of play, the player has discovered a wide variety of actions her persona can perform and the properties of many different objects in the environment. Some video games have on-screen characters that reveal some of the harder-to-discover game elements. Such characters were the inspiration for Marty, ToonTalk's guide. Puzzle sequences as tutorials A carefully designed sequence of puzzles can be very effective pedagogically. Many computer and video games use puzzles as an effective and fun tutorial. Lemmings® and The Incredible Machine® are two good examples. The idea is to present a sequence of puzzles that introduces new elements or actions one at a time in a simplified or constrained environment. A series of puzzles is more appealing to most children when it is embedded within a narrative adventure. The ToonTalk puzzle game starts with a brief "back story". An island is sinking, and a friendly Martian named Marty happens to be flying by and rescues
13. ToonTalk
263
everyone. He is nearly finished rescuing them when he crashes and is hurt. The player volunteers to rescue Marty. Because he is hurt (you can see his arm in a sling and his bruises), Marty can't get out and build the things needed to fix his ship. So he asks the player to make things for him. The player goes nearby where the components she needs can be found. She has to figure out how to use and combine them. When stuck or confused, the player can come back to Marty, who provides hints or advice about how to proceed. If a player is really stuck on a particular problem, then Marty gives her detailed instructions so she can proceed to the next puzzle. Note that getting advice or hints from Marty fits the narrative structure since Marty knows what to do but is too badly injured to do it himself. In order to fix Marty's ship, the first job is to fix the ship's computer. The computer needs numbers and letters to work. The goal of the first level is to generate the numbers needed. The culmination of the level is the construction of a program that computes powers of two (1, 2, 4, 8, and so on to 2 to the thirtieth power). The next level involves the construction of a program that computes the alphabet. The task after that is to fix the ship's clock. Solving these puzzles involves measuring time, mathematics, and some new programming techniques. At one point the player has constructed a number that shows how old she is in seconds. And the number changes every second! It is instructive to look at some puzzles in detail. The first real program a player builds is in the ninth puzzle. Marty needs a number greater than one billion for the computer. The player needs to train a robot to repeatedly double a number. Several of the earlier puzzles prepare the player for this task: 1. The first three puzzles introduce numbers, addition, and boxes (data structures). 2. In the fourth puzzle Marty needs a number greater than 1,000 (see Figure 5). When the player goes next door on the floor is just the number 1 and a magic wand that copies things (see Figure 6). The trick to this puzzle is to repeatedly copy the number and add it to itself, thereby doubling it each time (see Figure 7). In addition, the magic wand has a counter that is initially set to 10. After ten copies it has run out of magic and won't work any more. This helps constrain the search for a solution. The solution requires the player to repeat the same action 10 times. 3. In the eighth puzzle, the player is introduced to robots and builds her first program. This puzzle is very simple. Marty needs a box with two zeros in it. When the player goes next door she sees a robot with a magic wand and a box with one zero in it (see Figure 8). The wand is stuck to the robot and can't be used to copy the box. Most players discover that you can give the box to the robot (and those that don't, do so soon after getting hints from Marty). The player trains the robot to copy the box and drop the copy. Giving the robot the box activates him. He repeats what he was trained to do and copies the box.
264
13. ToonTalk
Figure 5 - The injured alien introducing the fourth puzzle
Figure 6 - TTie initial state of the fourth Figure 7 - Using the Magic Wand to copy a number during the fourth puzzle puzzle These early puzzles are designed to simplify some programming tasks. For example, the player doesn't need to know how to terminate the training of a robot. When the limit on the number of steps the robot remembers is exceeded, his training is automatically terminated. Similarly, the counter on the magic wand ensures that the robot will stop after the correct number of iterations. In later puzzles, arranging for robots to stop when a task is completed becomes the player's responsibility.
Figure 8 - The initial state of the eighth puzzle
13. ToonTalk
265
By the time the player starts the ninth puzzle, she has performed the prerequisite actions and must combine them properly to train a robot to repeatedly double a number. The player is presented with a robot holding a wand good for 30 copies and a box with a 1 in it. Because the player has only the robot and the box to work with, and because of the limitations imposed on the robot, this otherwise overly ambitious early programming example can be solved by most players with few or no hints. And yet the constraints do not make the puzzle trivial: experimentation, thinking, and problem solving are necessary to solve the puzzle. A good series of puzzles leads a player step by step where the puzzle designer wants to go. The players don't feel as if they are being led anywhere but have the illusion that they are in control. The puzzles constrain the set of objects that can be used and how they can be used so that the player has only a few choices. If designed well, the puzzle sequence can be challenging without being frustrating. Even among those children for whom the puzzle game is well suited, there is variation from those that want to figure out everything themselves to those who very quickly want hints. In Toontalk, if you come to Marty empty handed or with the wrong thing, he will give you a hint. Each time you return during the same puzzle you get a more revealing hint until eventually you get detailed instructions from Marty on how the puzzle should be solved. This behaviour accommodates a wide range of learning styles from independent problem solving to following directions. A good puzzle sequence has a "self-testing" character. ToonTalk puzzle number 15, for example, is a difficult programming task for novices — generating a data structure containing 1, 2, 4, 8, and so on up to 1,073,741,824. The prerequisite knowledge for constructing such a program was acquired in solving puzzle 9 (constructing a program to compute 2 to the 30th power) and puzzle 13 (constructing a data structure filled with zeros). These puzzles in turn rely upon having learned in earlier puzzles how to double a number and how to train robots (i.e., construct programs). The fact that the children succeeded in solving the puzzles indicates that the puzzles have succeeded and that the children are learning ToonTalk and computer programming. This kind of tutorial puzzle sequence is strictly linear. A less linear game based upon the idea of a treasure hunt or an adventure game should also be considered. The player would explore and find puzzles to solve. The game designer could still maintain some control by making certain areas open only to those who have succeeded in solving some prerequisite puzzles. Pictorial instructions Children can often be seen building a toy or a kit by following instructions that consist of a series of pictures. Many children, for example, enjoy building LEGO® constructions by following pictorial instructions. They learn design and construction techniques in the process, as evidenced by their own subsequent creations. Might not this technique work for children's software as well?
266
13. ToonTalk
To explore this question, a sequence of about 60 screen snapshots was generated for building an exploding object in ToonTalk. This involves using collision detectors, sound effects, and a change in an object's appearance. While children were able to follow the instructions, we learned that generating good instructions requires good graphic design, lots of testing and revision, and a good understanding of the required prerequisite knowledge and experience. In particular, we found the following: 1. Pictures should show what is necessary and nothing more. Screen snapshots are poor substitute for a good drawing because there are too many irrelevant details in each snapshot that make it hard to find the important parts of a picture. 2. The step size should be just right. Too big or too small a transition between successive pictures confuses the children. 3. Instructions should be appropriate for the level of experience of the child. The exploding-object instructions were too hard for children with just an hour or so of experience with ToonTalk. You might question the focus on pictorial instructions — what about textual instructions? Textual instructions for building things in ToonTalk tend to be awkward and hard to understand. The world of ToonTalk is so visual that text without accompanying illustrations or animations is not very effective. Consider how hard it is to explain to someone how to tie a knot over the phone. Nonetheless, a few children have been observed to repeatedly get hints from Marty to solve a puzzle until they receive full textual instructions from Marty, and only then, do they try to solve the puzzle. Viewing of demos Instructional films and educational TV are generally accepted as effective for some kinds of learning. Why not apply them to the task of learning to program inside of ToonTalk? ToonTalk includes eight different automated demonstrations. They are simply a replay of someone using ToonTalk, accompanied by narration and subtitles. Most of the demos are not different from watching someone give a demo to an audience. They tend to highlight different features or techniques. As with TV, there is no opportunity for the student to ask questions. Two of the demos are unusual. One is scripted like an introductory tour. The viewer imagines she is on a guided tour of the ToonTalk world. The tour guide welcomes the visitor and greets characters in the ToonTalk world and proceeds to show how the basic objects and tools in ToonTalk work. The other demo has a soundtrack of two children trying to build a Ping Pong game in ToonTalk. One of the children, Nicky, is a novice; the other, Sally, has a fair amount of experience but still finds building a Ping Pong game challenging. Nicky frequently asks questions, and consequently explanations of what is happening are given in a natural context. Most importantly, the children frequently make mistakes. This demo shows the process of building something in ToonTalk — including how to deal with
13. ToonTalk
267
bugs and mistakes. This demo illustrates the process of building small pieces, testing them, tracking down and fixing bugs, and then integrating the pieces. It is important to pay attention to production values when making software demos. Children will, quite naturally, compare them to TV shows. If the narration, script, or voice acting is amateurish, for example, the demos will not be as appealing. Another important thing that these demos attempt to communicate is good programming style in ToonTalk. By watching a demo of an expert building something, an observant student will notice not just the necessary actions, but all the other actions that constitute the style of the expert.
5 Social learning Absent from this discussion of ways of learning are the traditional ones like listening to lectures by teachers, asking questions, or doing homework assignments. These techniques can work quite well, especially when the teacher is knowledgeable. In such cases, the techniques described above can augment the activities of the teacher. Unfortunately, not all teachers are good at teaching complex subjects like computer programming [Yoder 94]. And computer programming is something interested children may wish to learn on their own. My hope is that a child can learn on her own with software that supports exploratory learning, problem solving, detailed instructions, and demonstrations. Also absent from this discussion is learning in a social context. Children frequently play or study in pairs or teams and they help and teach each other in the process. How can we design software to facilitate this kind of group activity? The software should do the following: 1. Work with a long viewing distance. Most software is designed to work for a user who is 12 to 18 inches from the display. When 2 or 3 children work together, the distance usually increases. ToonTalk, like most video games, was designed to work in a typical living room, where the display may be 4 to 10 feet from the player: text and objects are large. 2. Support multiple players. Nearly all children's software is designed to work with a single child using the mouse, keyboard, and possibly a joystick. In contrast, many video games today support 2 to 4 simultaneous players, each with their own joystick or game pad. ToonTalk could be enhanced to support multiple users, each with their own controller and screen persona. 3. Support networked collaboration. For software to support children playing or learning together over a network, it must deal with many technical issues such as voice communication, latency, and reliability, as well as social issues such as privacy, inappropriate behaviour, and trust. ToonTalk supports networked collaboration by the use of "long-distance" birds that can fly to nests on other computers running ToonTalk.
268
13. ToonTalk
4. Support an on-line community. Web sites, email, chat, and discussion groups all can contribute to a support network to help children master complex subjects like computer programming.
6 Back to the Future I have argued that computer-based learning environments should be universal in the broad sense of the word. Not only should they be capable of expressing any computation but the expression should be well-suited to the cognitive abilities of most students. Conventional programming languages fail here by being too abstract for most students. Furthermore, programs should be capable of using more than the purely computational capabilities of computers but their input and output devices as well. A Turing Machine cannot show nice graphics or play sound effects. While I believe that ToonTalk meets these goals better than other environments, there remain many areas of improvement. While ToonTalk provides support for 2D graphics, sound effects, force feedback effects, multiple input devices, networked computers, and web page access, it provides no support for 3D graphics, audio input, and video input and display. Support is limited for generating music and intercommunicating with other programs. ToonTalk supports long-distance synchronous communication and collaboration via birds that can fly to nests on other computers, but has no support for asynchronous communication. There are plans to address these shortcomings in the near future (except for 3D graphics). Like many other learning environments, ToonTalk shares the goals of supporting communities of learners, collaborative design and problem solving, and enabling learners to explore and create. Like many other environments, ToonTalk attempts to accommodate different learning styles. It attempts to do all this in a playful manner that appeals to children. Like few other learning environments, however, it also strives to make the full general power of computers available to not just the authors of learning materials but also to the learners themselves. The children can be not just consumers of educational software but producers as well. Unlike others that share this goal, ToonTalk is significantly easier for students to master. By making computation concrete, while keeping its generality, ToonTalk gives learners creative access to the full power of those magical devices we call computers.
References [Boxer 02] www.soe.berkeley.edu/~boxer/bibliography.html [Kahn 96] Ken Kahn, "ToonTalk - An Animated Programming Environment for Children", Journal of Visual Languages and Computing, June 1996. [Kahn 02] Ken Kahn, ToonTalk Web Site, www.toontalk.com [Kay 81] Alan Kay et. al., Byte Magazine, Vol. 6, No. 8, Smalltalk issue, August 1981. [Papert 80] Seymour Papert, Mindstorms: Children, Computers, and Powerful Ideas, New York, Basic Books. 1980.
13. ToonTalk
269
[Papert 93] Seymour Papert, The Children's Machine: Rethinking School in the Age of the Computer. New York. Basic Books. 1993. [Playground 01] Playground Web Site, www.ioe.ac.uk/playground [Resnick 96] Mitchel Resnick and Brian Silverman, "Active Essays" http://el.www.media.mit.edu/groups/el/projects/emergence/active-essay.html [Resnick 1997] Mitchel Resnick, Turtles, Termites, and Traffic Jams: Explorations in Massively Parallel Microworlds, MIT Press, Cambridge, MA, 1997. [Squeak 02] Squeak Web Site, www.squeak.org [Yoder 1994] Sharon Yoder, "Discouraged? ... Don't dispair! [sic]", Logo Exchange, ISTE 1994.
KEN KAHN has over 30 years of engineering and computer programming experience. In addition to his duties as President of Animated Programs, he has research positions at the University of London's Institute of Education and Royal Institute of Technology in Sweden. He was an associate professor at Uppsala University and has lectured at MIT, the University of Stockholm, and the Royal Institute of Technology in Sweden. Dr. Kahn was a visiting scholar at Stanford University from 1992-1997 and has served on the editorial boards of The Journal of Logic Programming, Lisp and Symbolic Computing, New Generation Computing, and IEEE Parallel and Distributed Technology: Systems and Applications. Kahn is the author of over 50 published articles and has earned a patent for animated user interfaces for computer program creation, control and execution. He has served as researcher/consultant for Xerox PARC, IBM, Rand Corporation, Citicorp, Atari Research, the Swedish Institute of Computer Science, the Institute for New Generation Computer Technology in Tokyo, the University of London, and Syracuse University. He has a Bachelor of Arts in Economics from the University of Pennsylvania, a Master of Science in Electrical Engineering from MIT and a Doctorate in Electrical Engineering and Computer Science from MIT. In his spare time, he travelled around the world in 80 days.
This page intentionally left blank
271
List of Authors Edith K. Ackermann MIT School of Architecture Design Technologies 10-491 M 77, Massachusetts Avenue Cambridge MA 02139 USA Antonello Delle Fave Dipartimento di Scienze Precliniche "Lita Vialba" Via G.B. Grassi, 74 20157 Milano ITALY Patrick de Muynck ModeNatie, Flanders Fashion Institute Nationalestraat, 28 2000 Antwerpen BELGIUM H. Ulrich Hoppe Institute of Computer Science University of Duisburg Lotharstr.65 47048 Duisburg GERMANY Celia Hoyles Schools of Mathematics, Science and Technology Institute of Education, University of London 20 Bedford Way London WC1H OAL UK Francois Pachet SONY Computer Science Laboratories, Paris Rue Amyot 6 75005 Paris FRANCE Sara Price University of Sussex Palmer, Brighton BN19RH UK
272
Yvonn Rogers 37A Brunswick St West HoveBNS 1EL UK Takahiro Sasaki Sony Computer Science Laboratories, Inc. Takanawa Muse Bldg 3-14-13, Higashigotanda - Shinagawa-ku, Tokyo, 141-0022 JAPAN Daniel K. Schneider TECFA-FPSE Universite de Geneve 40, bid du Pont d'Arve 1205 Geneve SWITZERLAND John Sivell Department of Applied Language Studies Brock University St. Catharines Ontario L2S 3A1 CANADA Luc Steels Vrije Universiteit Brussel, AI Lab and SONY Computer Science Laboratories, Paris Rue Amyot 6 75005 Paris FRANCE Dr Mario Tokoro Sony Computer Science Laboratories, Inc. Takanawa Muse Bldg 3-14-13, Higashigotanda - Shinagawa-ku, Tokyo, 141-0022 JAPAN Colwyn Trevarthen The University of Edinburgh 7 George Square Edinburgh EH8 9JZ UK Marleen Wynants Nux Publica 91, avenue Leopold Wiener 1170 Brussels BELGIUM
273
Subject Index affective culture 53 Ambient Wood 183 anxiety 103 autotelic 139 autotelic agents 144 behaviourism 138 Bruner, Jerome 60 C3MS 215 Cabri Geometry 246 child-sized printing press 158 Chromarium 178 classroom 229 collaborative game design 260 Comenius 59 Computer-Supported Collaborative Learning (CSCL) 223 constructionism 20 constructivism 18,261 Continuator 114 creativity 200 cultivation 101 cultural learning 48 culture 100 detachment 79 digital augmentation 175 digital mimicry 232 dwelling 26 educational technology 158 embodied learning interactions 173 emotion 57 emotional support 211 emotions 4 Experience Fluctuation Model (EFM) 103 Experience Sampling Method (ESM) 102 flow 105,120,122 Flow Questionnaire 103 flow theory 139 Freinet, Celestin 155 grounding 2 Hunting of the Snark 181 imitation 56 initiatic experience 125 instruction model 61 intent participation 61 interaction protocols 129
intimacy 58 intrinsic motivation 102,105,106 intuitive parenting 40, 51 joint awareness 48 learner-centered learning 7 learner-centered teaching 162 learner-friendly technologies 163 Logo 255 Markov systems 117 mechanism 240 motivation 42 musical style 115 musicality 52 narrative 47 newborn 39 optimal experience 101 Papert 20,23,249 pedagogical scenarios 196 physicality 174,177 Piaget, Jean 19,23 play routines 46 portals 212 pretend play 73 project based learning 198 protoconversational 43 psychological selection 100 quality of experience 145 reflection 128,229 reinforcement learning 138 representation 225 representation 76 representations 24 rhythm 51 self-consciousness 49 simulacre 31 situated 21 social learning 267 social virtual environments (SVE) 29 socio-constructivism 22 symbolic cognitive models 143 symbolic function 86 sympathy neurons 41 teacher communities 213 technology-mediated learning 174 Tomasello, Michael 48
274
ToonTalk 241,256 turn-taking 126 Turtle Geometry 28 ubiquitous computing 173
Vygotsky, Lev 22,23 well-being 99 zone of proximal development 22,49