The Emergence of Mind in a Physical World

459

Philosophy Series

La Deducción trascendental y sus inéditos, 1772-1788 Immanuel Kant Philosophy Series

Conocer sin representar. El realismo epistemológico de Donald Davidson William Duica Philosophy Series

proposal that does not lose sight of the special particularity and alleged irreducibility of some of the most obvious phenomena of our existence, such as the human mind. The author provides a compelling argument against the most widely accepted interpretation of Physicalism, Microphysicalism, highlighting its deep empirical and conceptual problems, as well as an insightful response to the reiterated critiques leveled by some

Convertir la vida en arte: una introducción histórica a la filosofía como forma de vida Andrea Lozano-Vásquez Germán A. Meléndez Philosophy Series

CO L E CC I Ó N   G E N E R A L

The Emergence of Mind in a Physical World

biblioteca abier ta

reductionist philosophers, especially Jaegwon Kim, at the non-reductive physicalist explanation of the causal relevance of mental and higher level properties. Morales argues in

Forthcoming Publications

ISBN 978-958-783-326-3

· Philosophy ·

Francisco Posada Díaz. Textos reunidos Carlos Rincón (editor)

The Emergence of Mind in a Physical World presents an ontologically and methodologically well founded Physicalist

favor of Emergentism as a non-reductive physicalist proposal that explains the causal reality of higher-level properties as metaphysically dependent but not supervenient on their microphysical bases. This novel interpretation will be of great interest to scholars working in the field of philosophy of mind.

The Emergence of Mind in a Physical World

Biblioteca Abierta Collection

Juan Diego Morales Otero

CES Collection

Academia Collection

Camilo Torres Restrepo. Sacerdocio y política Hildegard Lüning Traducción de Jorge Aurelio Díaz

Juan Diego Morales Otero

Microhistorias de la transgresión Max Hering y Nelson Rojas

Departamento de Filosofía Facultad de Ciencias Humanas Sede Bogotá

Author Juan Diego Morales holds a Ph.D. in Philosophy from the Universidad Nacional de Colombia and is currently associate professor of Philosophy at the Universidad de Cartagena, Colombia. In 2013, was visiting scholar at Rutgers University, where he worked with Professor Brian McLaughlin on emergence and the mind-body problem. In 2017, he won first place in the Concurso de Ensaios Categoria Sênior sobre o Problema Mente-Cérebro by Núcleo de Pesquisa em Espiritualidade e Saúde da Universidade Federal de Juiz de Fora, Brazil, with the essay «Mental Causation as Emergent Causation». His publications include: «The Myth of Supervenient Microphysicalism» (2016), «Non-reductive Physicalism: Understanding Our Metaphysical Paradigm» (2015), Imágenes de la mente en el mundo natural (with Meléndez, et al, 2014), «Fodor y Kim en torno a la posibilidad de las ciencias especiales, la realizabilidad múltiple y el reduccionismo» (2013), and «La relación entre acción y pensamiento en la filosofía tardía de Wittgenstein» (2012).

biblioteca abier ta colección general filosofía

The Emergence of Mind in a Physical World

The Emergence of Mind in a Physical World

Juan Diego Morales Otero

2018

catalogación en la publicación universidad nacional de colombia Morales Otero, Juan Diego 1980The emergence of mind in a physical world / Juan Diego Morales Otero. – Primera edición. – Bogotá : Universidad Nacional de Colombia. Facultad de Ciencias Humanas, Departamento de Filosofía, 2018. 232 páginas: ilustraciones en blanco y negro, diagramas. – (Biblioteca abierta. Serie Filosofía ; 459) Incluye referencias bibliográficas e índice de materias. ISBN 978-958-783-326-3 (rústica). – ISBN 78-958-783-327-0 (e-book). 1. Filosofía de la mente 2. Superveniencia 3. Emergencia o surgimiento (Filosofía) 4. Positivismo lógico 5. Filosofía de la ciencia 6. Teleología 7. Causalidad II. Título III. Serie CDD-23

128.2 / 2018

The Emergence of Mind in a Physical World © Biblioteca Abierta Colección General, serie Filosofía

© 2018, Universidad Nacional de Colombia, sede Bogotá, Facultad de Ciencias Humanas, Departamento de Filosofía © 2018, Autor, Juan Diego Morales Otero Primera edición, 2018 ISBN impreso: 978-958-783-326-3 ISBN digital: 978-958-783-327-0

Comité editorial Facultad de Ciencias Humanas Luz Amparo Fajardo Uribe, Decana Nohra León Rodríguez, Vicedecana Académica Constanza Moya Pardo, Vicedecana de Investigación y Extensión Carlos Tognato, Director del Centro de Estudios Sociales -CESJorge Aurelio Díaz, Director Revista Ideas y Valores Rodolfo Suárez Ortega, Representante de la Unidades Académicas Básicas

Preparación editorial Centro Editorial de la Facultad de Ciencias Humanas [email protected] www.humanas.unal.edu.co Camilo Umaña, Diseño original de la colección Camilo Baquero Castellanos, dirección Angélica M. Olaya M., coordinación editorial Juan C. Villamil N., coordinación gráfica - maquetación Rosario Casas, corrección de estilo Impreso en Colombia Prohibida la reproducción total o parcial por cualquier medio, sin la autorización escrita del titular de los derechos patrimoniales.

Contents Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Chapter 1. Causation: A Non-Reductive Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 The Nature of Causal Entities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Causation and Laws of Nature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Counterfactual Theory of Causation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Causation and Locality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Probability and Intervention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Causal Context and the Possibility of Causation in the Special Sciences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Chapter 2. The Concept of the Physical and the Overcoming of the Supervenience Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 What Physicalism Is

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

56

Physicalism as a Contingent Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 What Is It To Be Physical?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Microphysical Supervenience as the Criterion for the Physical. . . . . . . . . . . . . . . . 62 The Empirical Failure of Microphysicalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Consequences for Determining the Concept of the Physical . . . . . . . . . . . . . . . . . . . 72 Finding a Solution to our Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Chapter 3. Emergentism as Type Macrophysicalism. . . . . . . . . . . . . . . . . . . . . . . . . 81 The Idea of Ontological Emergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Emergentism as Type Macrophysicalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Reductive and Non-Reductive Microphysicalism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Property Dualism and Contingent Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Emergence and Metaphysical Dependence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Two Types of Ontological Emergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Chapter 4. Non-Reductive Physicalism and the Problem of Causation in the Special Sciences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Supervenience and Multiple Realizability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 The Causal Exclusion Argument. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

The Supervenience Argument. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 The Principle of Physical Causal Closure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 The Exclusion Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 The Interventionist Account of Exclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 The Argument of Causal Individuation of Natural Kinds . . . . . . . . . . . . . . . . . . . . . . . 140

Higher Causal Powers as Identical to Lower Causal Powers. . . . . . . . . . . . . . . . . . . . 142 The Subset Account of Realization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Chapter 5. Emergent, Downward, and Mental Causation. . . . . . . . . . . . . 155 The Idea of Emergent Causation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

The Principle of Physical Causal Closure of Emergentism . . . . . . . . . . . . . . . . . . . . . . 163 Emergent and Lower Level Causal Powers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Testing the Causal Relevance of Higher Level Properties . . . . . . . . . . . . . . . . . . . . . . . 170 Emergent Causation at the Quantum Level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Mental Causation as Emergent Causation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

Appendix: Functional Reduction of the Special Properties. . . . . . . . . . . 199 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229

For Sri Kṛṣṇa, the Taitas of Putumayo, and my family, without whom this would not have been possible

So my aim is only to establish conditional claims of the form: even if certain facts are emergent vis-à-vis the microphysical realm, Physicalism can still be true.

Papineau 2008 132 Thus, for emergent evolution, conscious events at level c (mind) involve specific physiological events at level b (life), and these involve specific physico-chemical events at level a (matter). No c without b, and no b without a. No mind without life; and no life without «a physical basis.»

Morgan 1923 15 I am not, for example, denying that there occur mental processes. Doing long division is a mental process and so is making a joke. But I am saying that the phrase «there occur mental processes» does not mean the same sort of thing as «there occur physical processes», and, therefore, that it makes no sense to conjoin or disjoin the two.

Ryle 2009 11-2 No one has contributed as much to our understanding of the problems of mental causation in recent years as Jaegwon Kim. We non-reductive materialists must face up to the serious difficulties he has raised for our position.

Block 2003 133 Whereas, if it isn’t literally true that my wanting is causally responsible for my reaching, and my itching is causally responsible for my scratching, and my believing is causally responsible for my saying…, if none of that is literally true, then practically everything I believe about anything is false and it’s the end of the world.

Fodor 1990 156

Acknowledgments

I would like to thank Professor Alejandro Rosas, without whose invaluable help and support I would not be presenting this dissertation. I would also like to thank many people who, in one way or another, have positively influenced the development of the ideas set forth here: Professors Brian McLaughlin, Larry Shapiro, Reinaldo Bernal, Louis Sass, Jaegwon Kim, Jaime Ramos, William Duica, Juan Botero, Adrian Cussins, Ignacio Ávila, Gonzalo Serrano, Barbara Montero, and Alicia Juarrero; my philosophical and non-philosophical friends Omar Contreras, Noelly Guerrero, Angélica Santacruz, Ximena Velosa, Alejandro Gonzales, José Tovar, Luis Yagé, David Cortés, Juan Camilo Espejo, Paula Gutiérrez, Danilo Rodríguez, Mábel Reyes, Aimer Díaz, Carlos Garzón, Alejandro Murillo, Louis Cohen, Leonardo Gómez, and Teyrun Torres; the different students that have endured me; and, finally, my family: my parents Martha Elcy and Pedro Ignacio; my sisters Natalia, Puspanjali, and Sandra; my brothers Achara and Moisés; and, especially, my girlfriend Margoth. During the course of my research, I was awarded the Outstanding Graduate Student Scholarship by the Universidad Nacional de Colombia.

13

Introduction

One of the oldest and most important philosophical problems is the question about human uniqueness. We know that Descartes introduced his dualist proposal with the aim of explaining it, and we also know that his position entails seemingly intractable problems. Since the beginning of Modernity, philosophers like Spinoza and Leibniz have noted that Descartes’ proposal cannot be correct because mind and matter would constitute completely different substances, with such different attributes that they could not even be causally related to each other. It is precisely because of the existence of a prima facie metaphysical difference between minds and bodies that the question about their connection has occupied a central place throughout the history of Western thought, a crucial issue that has been called the mind-body problem. Following the anti-Cartesian spirit, the physicalist proposal argues that our world, and therefore the human mind as one of its most important components, should be understood as fundamentally physical. It is now clear, as Gillett and Loewer (2001 ix) remind us, that the Weltanschauung of much contemporary philosophy is the doctrine of physicalism and, in consequence, as Kim comments,

15

Juan Diego Morales Otero

«[a] strong physicalist outlook has shaped contemporary discussions of the mind-body problem» (2005 1). One of my first and primary purposes in this book is to argue that, in the words of Papineau, «even if certain facts are emergent visà-vis the microphysical realm, Physicalism can still be true» (Papineau 2008 132). This means that emergentism, properly understood, should clearly be seen as a physicalist theory affirming that both mind and the phenomena of the special sciences (from chemistry and biology to psychology, sociology, and economics) are macrophysical entities that metaphysically depend on but cannot be reduced to the properties and relations of microphysics. The idea of emergence was formulated in ancient times and has reappeared in human thought at different points in history. Perhaps we can find the greatest conceptual development of this perspective from the late 19th to the early 20th century in the work of the so-called classic British Emergentists, thinkers like John Stuart Mill, Henry Lewes, Samuel Alexander, C. Lloyd Morgan, and C. D. Broad. It is essential to understand and keep in mind that the concept of emergence is introduced in order to overcome both dualism and microphysicalist reductionism. On the one hand, the dualist perspective asserts that there are two classes of individuals or substances, corporeal and immaterial, which are completely and ontologically independent, but that they can be causally and contingently related. On the other hand, microphysicalist reductionism (called ‘mechanism’ by British Emergentists) argues that all complex phenomena which in principle seem to be located outside or beyond the physical realm, such as biological, mental, and social phenomena, are actually metaphysically determined by, and in fact are «nothing over and above,» basic physical phenomena, i.e., microphysical entities. Emergentism seeks to overcome these positions in a subtle and somehow synthetic way. Against the dualist, the emergentist argues that the world previously conceived as bifurcated, as divided into two separate ontological realms, must be understood as a unitary world wherein we only find entities belonging to a fundamental ontological category, that of material beings. And, against the reductionist, the emergentist claims that the material world is not a simple, linear, and flat domain, whose constituents are located in a single hierarchical

16

Introduction

and ontological level; rather, it is a world consisting of a series of layers that complexly organize its purely material elements. The basic idea of​​ emergentism is that there are macrophysical systems with distinctive characteristics and dynamics arising from the properties, relations, and interactions of purely physical components which, however, cannot be reduced to, explained in terms of, or identified with the latter. According to this view, paradigmatic examples of emergent physical systems are biological and mental. If my goal is to show that emergentism should be articulated as a completely physicalist theory, we can understand the bulk of the monograph as a direct response to Jaegwon Kim’s reiterated criticisms of any non-reductive physicalist theory, according to which the latter’s understanding of the causal responsibility of the higher level properties is finally inconsistent and, therefore, must adopt either causal reductionism or epiphenomenalism. To put it in other words, we can say that much of this book can be seen as a direct response to the challenge that Kim imposes on non-reductive physicalism to come face to face with the problem of downward causation. […to] devise an intelligible and consistent account of how emergent [that is, higher level and non-reducible] properties can have distinctive causal powers of their own —in particular, powers to influence events and processes at the basal level. (Kim 2006a 559)

To that effect, the monograph is structured as follows. Chapter 1 is devoted to examining different philosophical approaches that have tried to define the concept of causation in terms of nomological regularities, counterfactual dependence, and transference of a physical quantity. I stress the importance of not reducing this concept to any one of these factors and claim that a fundamental characteristic of causation is what I call the internal context of the cause, namely, the internal properties or conditions that an event as a cause must have to be nomologically sufficient for its effect. I give reasons to affirm that the possibility of real causation in the special sciences can be explained from the existence and interaction of these internal conditions.

17

Juan Diego Morales Otero

In chapter 2, I focus on the analysis of the concept of the physical. Through the examination of the contemporary characterization of the physicalist theory, I attempt to make explicit the necessary conditions to be fulfilled by any object, event, or entity that should be counted and addressed as physical. I develop a detailed argument to show why the most accepted formulation of physicalism, the theory of the metaphysical supervenience or complete determination of empirical phenomena by the microphysical characteristics, is insufficient both empirically and conceptually. On the one hand, it contradicts results from physical science; and, on the other hand, it contradicts both the scientific and the daily use of the notion of the physical. I argue that the philosophical tradition -at least since Descartes- can provide the criteria for understanding the meaning we give to this notion, so that physicalism should be understood as affirming that all the entities of our world are essentially objective, necessarily spatiotemporal, and, in principle, explainable according to the mathematical methodology of the natural sciences. In this sense, we can say that physicalism posits a world some of whose fundamental phenomena can be essentially macrophysical, i.e., physical phenomena which cannot be reduced to, nor understood purely in terms of the properties and relations of their microphysical components. The purpose of chapter 3 is to develop a careful articulation of the concept of ontological emergence. In the first section, I develop a general characterization of ontological emergence whereby this phenomenon should be understood as a special organization or relational structure that the constituents of a system can acquire, and which introduces a causal and dynamical difference that is not completely determined by the causal factors of these constituents. The second section is devoted to the examination of the relation between the emergentist thesis and the ontological approaches of reductive physicalism, non-reductive physicalism, token physicalism, and property dualism, showing that, despite the numerous readings suggested in recent years, emergentism should be treated as a clear case of non-reductive physicalism. Finally, in the third section, I distinguish two general types of ontologically emergent entities, which will allow us to comprehend the internal

18

Introduction

­ iversity of the phenomenon and to refine the boundaries of the concept d in order to understand its philosophical and scientific consequences. In chapter 4, I carry out an analysis of the problem that nonreductive physicalism in its functionalist account (nrp), the most important philosophical naturalistic position in recent decades, has to face when defending its claim about the reality and irreducibility of the causal power (influence, responsibility) of mental properties and those of the special sciences. This position affirms that mental and higher level properties are legitimate constituents of a fundamentally physical world because they metaphysically supervene on and are realized by basic or microscopic physical properties. In addition to its physicalist commitment, this perspective argues that higher level or special properties cannot be reduced to lower level physical properties, just because they are multiply realizable by them. Some philosophers, especially Jaegwon Kim, have leveled very important critiques at this non-reductive physicalist (nrpist) proposal. Kim develops the well-known causal exclusion argument which purports to prove that for the physicalist there are only two options with respect to the status of special properties: either reductionism or epiphenomenalism. The interventionist approach to causation responds that the exclusion argument turns out to be incorrect when we consider it according to its empirical implications. Although I agree with the interventionist approach that the exclusion argument is not conclusive, I believe there is another important argument set forth by Kim to show that nrp cannot account for the reality of the causal status of higher level properties: what I call the argument of causal individuation of natural kinds, which affirms that on the nrpist proposal, higher level properties cannot maintain a necessary unity through their different physical realizations and, for this reason, cannot be considered as real natural properties or kinds. These criticisms support the functional reductive approach (see Appendix) which argues that to avoid eliminativism about higher level properties, we must reduce them ‘locally’ to each of their physical realizers. This strategy takes the predicates of the special sciences as expressions which contingently refer to different physical properties

19

Juan Diego Morales Otero

that perform the causal role that these expressions define. But the paradoxical consequence is that, on this view, special properties should be finally eliminated. This follows because the categories of the special sciences are not truly referring to real special properties but only to sets of multiple and dissimilar basic physical properties in a somehow contingent and indirect way. The overall conclusion is that neither nrp nor the functional reductive proposal seems to have a satisfactory account of mental and special properties. On the one hand, nrp cannot vindicate the reality of special causal powers because it cannot explain the necessary unity they must maintain through their different physical realizers. On the other hand, although more metaphysically robust, the functional reductive proposal affirms that insofar as the special properties arise from different physical conditions, they finally cannot be real, a consequence that I think is very difficult to swallow. Since this is an empirical claim, what I find most problematic about the functional reductive proposal is that it seems to assume that, in a physical world, the reduction of the higher level properties and the denial of their mr is a conceptual or metaphysical fact. But surely a physical world with mr higher level properties is an empirical possibility, not a metaphysically excluded possibility! And this entire complicated picture arises from the single idea of the metaphysical supervenience of the macro-properties on their microphysical realizers or conditions. This is an empirical question and it could be that the microphysicalist thesis is correct. If this were the case, we should say that the functional reductive proposal is the more plausible approach for understanding our empirical world. But I have said that microphysicalism is a contingent thesis with deep conceptual and empirical deficiencies. On the one hand, it does not allow us to fix or to understand the use we make of the notion of the physical, and, therefore, of the notion of the physicalist theory; this only means that there are alternative ways for understanding and formulating physicalism which are not based on the microphysical supervenience theory. And, on the other hand, it seems to be incompatible with results coming both from physical science itself, as when we talk about holistic or systemic

20

Introduction

physical properties not reducible to their constituent conditions, and from the special sciences’ greatly successful theories and experiments that provide explanations and predictions which, as far as we know, are not reducible to the microphysical laws and explanations from which they must arise. I think emergentism can do better. Finally, in chapter 5, I focus on a careful articulation of the concept of emergent causation and its application to the phenomenon of mental causation. While the notion of emergent causation is the idea that some irreducible macrophysical properties (properties only instantiated in composed physical systems) are causally relevant for the subsequent instantiation of both other macro or higher level properties and microphysical properties, the concept of mental causation refers to the causal relevance or responsibility that mental or psychological properties, such as being in pain, believing that snow is white, and desiring ice cream, have on the subsequent instantiation of other properties, whether psychological, social, biological, or properties of any other organizational level. The prominent concepts of downward causation, lower level causal under-determination, and higher level causal constraint and selection are clarified and interconnected in this final chapter. Although some theorists interested in the nature of dynamical systems and the appearance of emergent properties in a physical world have suggested some possible avenues for understanding the downward causal interaction entailed by the occurrence of any higher level and non-reducible (i.e., emergent) causal process, there has not been any effort to systematically articulate its basic structure. I develop this articulation through the elucidation of three examples: in the first place, I analyze and reconstruct in detail an abstract example that helps us understand the general mechanism and structure of downward causation which is based on the constrictive and selective action of the higher level laws on the underdetermined possibilities of the lower levels. This articulation serves to clarify the kind of nonreductive principle of physical causal closure that emergentism should maintain, the necessary relationship between the emergent and the lower level causal powers that is implicit in every case of emergence,

21

Juan Diego Morales Otero

and the very special and complex method to empirically test the causal relevance of the higher level properties. Secondly, I carry out a conceptual examination of the causal structure of one the most recurrent examples that different theorists have used to analyze the apparent failure of the microphysical supervenience theory, that is, the phenomenon of the quantum states of entanglement. I argue that if our current scientific understanding of the quantum world is correct, then we have a primary empirical example that allows us both to claim the failure of microphysicalism, the most predominant philosophical-scientific theory of recent decades, and, what is more important for our conceptual understanding of the relationship between the different levels of organization of our physical world, to clarify and articulate in a very concrete way the nature and structure of the phenomenon of emergent and downward causation. Thirdly, I examine the neurobiological basis of pain and its connection with the appearance of different levels of personal and experiential psychological phenomena. I describe the two different neurological nociceptive (relative to pain) subsystems, namely, the discriminative and the affective nociceptive neural structures, from which two different and corresponding nociceptive experiences arise: the discriminative and affective nociceptive experiences. I examine the different levels of composition and organization that are implicated here to focus on the conceptual articulation of the causal dynamics that should structure the interaction between the two experiential levels involved in this phenomenon, that is, the level of discriminative and affective nociceptive experiences and the level of our normal and unitary experience of pain. On the basis of this type of examples, the emergentist theory then argues not only that the mind can emerge from atoms, molecules, cells, and neural informational processing, but that the very mental states can become organized in a hierarchical, emergent, and irreducible way. Finally, it is possible to say that the arguments developed throughout this monograph show that macrophysicalism or emergentism is not only a coherent and well suited conceptual proposal about the causal functioning of the different levels of composition and organization of our physical world, but that, as far as we know, it can be its most plausible empirical articulation.

22

CHAPTER 1 Causation: A Non-Reductive Approach

This initial chapter is devoted to examining the concept of causation. Different philosophical approaches have attempted to define it reductively in terms of nomological regularities, counterfactual dependence, and transference of a physical quantity, among others. In the analysis of the different approaches, I find important reasons to affirm that causation should be understood as incorporating these different elements, even without being reduced to them. I claim that one of the most important features for understanding the nature of causation lies in what I call the internal context, namely, the internal properties or conditions that the event as a cause must have to be nomologically sufficient for its effect. I argue that the possibility of emergent causation or causation in the special sciences can be articulated from the existence and interaction of these internal conditions. In the spirit of Wittgenstein’s later work, I think it is very useful to start with the analysis Copi & Cohen (1990) have developed of the concept of causation. They say that there are several different senses of the word ‘cause’, so we must make a distinction that allows us to clarify their differences. They continue: The word ‘cause’ is sometimes used in the sense of necessary and sometimes in the sense of sufficient condition. It is most often used

23

Juan Diego Morales Otero

in the sense of necessary condition when the problem at hand is the elimination of some undesirable phenomenon. To eliminate it, one need only find some condition that is necessary to its existence and then eliminate that condition. […] The word ‘cause’ is used in the sense of sufficient condition when we are interested not in the elimination of something undesirable but rather in the production of something desirable. (1990 378)

In addition to its use as a sufficient or necessary condition, Copi & Cohen have found that the word ‘cause’ sometimes refers to necessary and sufficient conditions, and to remote and proximate determinants, where the temporal relationship between the event considered as the necessary and/or sufficient condition and the effect is emphasized. Some theorists talk about a difference between token and property causation,1 and between ‘thin’ (involving a mere dependency) and ‘thick’ or productive causation.2 The emergentist, committed to the ontology of the physical reality hierarchized through different levels of organization, suggests that the word ‘cause’ takes a related meaning when we speak about the inter-determination of the different levels and their events: as a nomologically necessary condition for the sub3 sequent instantiation of both higher level and lower level -ultimately microphysical- properties.4 We can say that there are different kinds of dependency and determination relations; for example, conceptual (such as that between the concept of a triangle and the concept of a three sided geometric figure), logical (such as the relationship between the propositions of syllogisms), mereological (between the parts and the whole), metaSee, for example, Bunge 1959, Rupert 2008, and Woodward 2013. Hall 2004, Kim 2007b. 3 A higher level property is a property that is instantiated in virtue of (because it depends on) the instantiation of a lower level property. A paradigmatic example of a higher level property is a macrophysical property (a property that is instantiated by a physical system which is mereologically composed of other physical systems) that is instantiated in virtue of some microphysical properties (physical properties that characterize the most basic and simple physical entities that may exist). 4 We will discuss this sense in detail in chapter 5. 1

2

24

Causation: A Non-Reductive Approach

physical (between the facts of being human and being a particular animal, and between the facts of being a water molecule and being an h2o molecule),5 normative (between the rules of a game and a particular move within the game), non-causal nomological (such as the conservation law of energy, and the mass-energy equivalence), as well as causal nomological (e.g., smoking causes lung cancer and the action of an external force causes change in motion). Although many philosophers have equated the concepts of causation and determination, some have acknowledged their difference. In a first approach, and following Bunge (1959) and Kim (1973b, 1974), we can say that causation is a special kind of dependency and determination. In order to understand the concept of causation, we must realize that determination and determinism are not the same. Although causation is a kind of determination, it does not imply determinism. While in principle the determination relation applies to any type of entities (objects, events, properties, propositions, numbers, quantities, etc.), determinism is the idea or doctrine more associated with the determination of the history or development of the world’s events, in which such events are completely fixed, usually by initial conditions and the laws of nature.6 Now, given the most widely accepted interpretation of the quantum theory of matter, most contemporary theorists maintain at least the possibility of a causally and nomologically determined, although non-deterministic, world; a world in which the events are not fully determined by antecedent events and the laws governing their appearance and, therefore, where causation

A metaphysical relation between two facts or properties is the kind of relation that is necessarily maintained within all the metaphysical or logical possible worlds. The fact of being a human is metaphysically dependent on the fact of being an animal because there cannot be (in any metaphysically possible world) a human that is not an animal. And the fact of being a water molecule is metaphysically determined by the fact of being an h2o molecule because once the property of being an h2o molecule is fixed, then, in every possible world, the property of being a water molecule will be fixed. 6 For example, Hoefer (2010 1) says: «Causal determinism is, roughly speaking, the idea that every event is necessitated by antecedent events and conditions together with the laws of nature.» 5

25

Juan Diego Morales Otero

is basically probabilistic (and thus non-deterministic); where causes act by increasing the probabilities of their effects. The Nature of Causal Entities

To examine the special kind of determination and dependency that causation constitutes, let us initially address the question about the kind of entities involved. As Jonathan Schaffer comments, The standard view of the causal relata is that they are of the category of event, and that their number is two, in the roles of cause and effect. So on the standard view, when the cue ball knocks the nine ball into the corner pocket, there is said to be an (actual) event e1 of the cue ball striking the nine ball, and an (actual, distinct) event e2 of the nine ball sinking into the corner pocket, such that e1 is cause and e2 effect. The standard view, in short, holds that e1 causes e2. (2008 §1)

It is fair to say that most theorists hold this perspective.7 In this case, when we say that a cause is a previous condition for the appearance of its effect, we say that this condition is an event and, so, that without the occurrence of such event the effect would not have happened (in the case of being a necessary condition), or that the appearance of the event will necessarily produce the effect (being a sufficient condition). At the same time, this usually means that an event is the instantiation of one or more properties at a spatiotemporal zone8. It is important to underline a point about the relationship between cause and event; causes are events because they fulfill the main conditions of the events: (i) they are instantiations of properties, and (ii) as concrete entities, they occur at a spatiotemporal location. In this sense, then, we must say that, insofar as they are causes, the previous conditions of effects are entities that instantiate properties and occur at a specific spatiotemporal zone. Now, just as circumstances, states, conditions, and processes instantiate properties and occur spatiotemporally, so these 7 8

26

See, for example, Kim 1973a, Fodor 1974, Davidson 1980, and Lewis 1986. See, for example, Fodor 1974 398, J. Bennett 1988 88, Kim 1990 641. Although Davidson does not talk about properties as constituents of events, he agrees with Quine about the idea that «events, like physical objects, are identical if they occupy the same places at the same times» (1980 248).

Causation: A Non-Reductive Approach

types of entities count as causes too. In principle and in a general sense -in the sense of instantiation of properties in a spatiotemporal zone- I will use the concept of event to refer to all of these phenomena. Authors like Kim and Shoemaker developed a similar approach; Kim comments: Besides events, we also speak of ‘states’. If ‘events’ signal changes, ‘states’ seem to be static things, ‘unchanges’, to use a term of C. J. Ducasse’s; […] There are, however, good reasons for not taking this dichotomy of changes and unchanges, or of events and states, too seriously at the initial stage of developing a theory of events. For one thing, there are cases that are hard to classify; […] Moreover, it is a philosophical common place to use the term ‘event’ in a broad sense, not only to refer to changes but also to refer to states, conditions, and the like. When universal determinism is formulated as «Every event has a cause» or «the aim of science» is said to be the explanation and prediction of events in nature, it surely is not intended that states, narrowly so-called, escape the net of causal relations or that it is not the business of science to explain why certain states obtain, e.g., why the sky looks blue or why the earth is pear-shaped. (Kim 1976 33-4)

Meanwhile, in a similar vein, Shoemaker claims: [T]he occurrence of events, states and processes can be held to consist in the occurrence of property instances. And that is what I believe. One sort of event is just a property instance -something’s having a property at a time. Other events are series of property instances, presumably ones whose successive members are causally related. A state can be thought of a monotonous event of this second kind -one in which the successive property instances are all instances of the same property, or of similar ones. Processes will also be events of the second kind, typically instances of kinds of series of events that have a certain characteristic outcome. (Shoemaker 2007 116)

Besides the fact that circumstances, states, conditions, and processes are entities that meet the two general conditions of events, there is another compelling reason to consider them within the general category of cause. On the basis of a passage by the emergentist philosopher Samuel Alexander (1920 8), where the existence and reality

27

Juan Diego Morales Otero

of an entity is intrinsically related to its causal power, Kim articulates what he calls «Alexander’s Dictum». In the analysis of the emergentist proposal about mental reality he says: But what does it mean to say that an emergent property is «real», or to say that anything is «real»? Alexander’s Dictum gives an answer that is apt and irresistible: To be real is to have causal powers. Mental realism entails the claim that mental properties are causally potent. (1992a 134-5)

It seems that the principle that all concrete existence9 must have some capacity or power to intervene on the world, to cause some differences, is a quasi-conceptual or even metaphysical principle. For it seems that we could not think of an entity that exists in a physical and causal world like ours without a power to affect that world, a capacity to determine certain changes and states of the objects that constitute this world10. We can see that this principle is accepted by virtually all contemporary thinkers11. In fact, it is a principle at least as old as Plato’s Sophist (247d-e). This is why D. M. Armstrong, following Graham Oddie and alluding to Plato’s passage, affirms what he calls the Eleatic Principle: «Everything that exists makes a difference to the causal powers of something.» (1997 41) Kistler calls it «the Causal Criterion of Reality (ccr). The ccr says: for an entity to be real it is necessary and sufficient that it is capable to make a difference to causal interactions» (2002 461). Although in the quoted passage Kim talks about the causal powers of mental properties, not the causal powers of mental states or events, which is already of common use in the literature on the matter, this way of putting things is only a shorthand way to talk about the causal powers mental events have qua mental (as falling under mental kinds or having mental properties). Strictly speaking, causal powers pertain to events, not to their properties or to the kinds into which they fall. But, of course, they have causal powers thanks to their properties. We shall see that this causal relevance of properties is essential for understanding the operation of causation. 10 Even Cartesian dualism is clear about this idea: Descartes’ problem is not denying the causal relationship between mind and body, but explaining how such a relation is possible. 11 See, for example, Gillett 2003 593, Kim 2007a 409, Fodor 2003 136, Shoemaker 2007 142. 9

28

Causation: A Non-Reductive Approach

If indeed all existence, at least all concrete existence, inherently requires a causal power, then we must say that circumstances, states, conditions, and processes, as concrete and real entities, must have a causal influence on the world. Despite this, some contemporary philosophers concerned with the phenomenal or qualitative features of mental states, have tried to deny the so-called Alexander’s Dictum and support some kind of epiphenomenalism with respect to such mental characteristics.12 These authors base their claims on the alleged possibility of the absent or inverted qualia cases, i.e. the alleged possibility of beings that are physically and functionally identical to humans, but that differ in their phenomenal states. However, I would claim that there is no such possibility; as Shoemaker affirms in his latest book, «what qualia an experience has makes a causal difference.» (2007 122)13 Moreover, I agree with McLaughlin’s more general statement: «I regard epiphenomenalism as absurd» (2012 3). But let us implausibly suppose that qualia are epiphenomenal. Given that qualia are the only knowable properties that could change without a change in the causal power of their holders, we have to say that except for qualitative properties, Alexander’s Dictum applies for all instantiations of properties, and, therefore, that all instantiations of properties must have their own causal powers. If we talk about conditions, events, states, and such, instantiated in a spatiotemporal zone, they must be causal nodes within a larger network that constitutes our physical and causal world. Causation and Laws of Nature

Let us recall that Hume identified four prima facie constituents in the relation of causation: constant conjunction, contiguity in space and time, temporal priority, and necessary connection. It is almost obvious that causation implies a constant conjunction of the related events. As Hume says, «There must be a constant union betwixt the cause and effect. ’Tis chiefly this quality, that constitutes the relation.» (thn Book i, Part iii, Section xv) Despite this, we could examine 12 13

This is the case of Block 1980, Chalmers 1996, and Kim 2007a. See also Shoemaker 1981, Lycan 1987, Tye 1995, and Van Gulick 2009.

29

Juan Diego Morales Otero

this point more carefully and accept the criticism of authors like C. J. Ducasse (1926), Heathcote & Armstrong (1991), and Max Kistler (2006a) against Hume’s idea, according to which, in the words of the first, «[t]he supposition of recurrence is […] wholly irrelevant to the meaning of cause; that supposition is relevant only to the meaning of law» (Ducasse 1926 129). The idea of these authors is that in order to have a causal relationship, its recurrence or even its instantiation on more than one occasion is not necessary. The following example illustrates the point. In the beginning of the universe, there were neither molecules of water nor sand particles, which constitute the everyday causal process that occurs when sand gets wet on contact with water. Surely, this kind of causal process only occurred after the appearance of planets similar to ours. But it is at least logically possible that this kind of causal process had been instantiated only once, say at time t1, without its subsequent recurrence, because of the possible disappearance of our planet at time t2, as a consequence, for example, of the explosion of a nearby star. If this is correct, there is the possibility of non-recurrent causal relationships and, therefore, we should say that the concept of causation does not imply the idea of recurrence of the causally related events. That is, we should say that the idea that a is the cause of b does not imply that there must be a recurrence of type-a events followed by type-b events, even though it implies the existence of a causal law relating these kind of events -a question that will be considered below. Now, what about the modal relation between cause and effect? Since Hume, philosophers have been hesitant about considering it either a contingent or a necessary connection. It is clear that the Scottish philosopher successfully argues for its contingency; for example, he says: The mind can never possibly find the effect in the supposed cause, by the most accurate scrutiny and examination. For the effect is totally different from the cause, and consequently can never be discovered in it. Motion in the second billiard-ball is a quite distinct event from motion in the first; nor is there anything in the one to suggest the smallest hint of the other. […] When I see, for instance, a billiard-ball moving in a straight line towards another; even suppose motion in the second ball should by accident be suggested to me, as the result of

30

Causation: A Non-Reductive Approach

their contact or impulse; may I not conceive, that a hundred different events might as well follow from the cause? […] All these suppositions are consistent and conceivable. (ehu 4.9-4.10)

Hume shows that the idea or concept of the movement of a billiard ball bumping against another (and generally an event qua cause) does not imply the idea of the ​​ movement of the second billiard ball (or in general the idea of a​​ second particular event qua effect of the first one). Despite the distinction and the contingency that must exist between cause and effect, it seems that the notion or concept of causation implies the idea of some kind of necessitation and determination going from the cause to effect; for this reason, we often say that the cause is sufficient for the effect. As Hume himself affirms, Shall we then rest contented with these two relations of contiguity and succession, as affording a complete idea of causation? By no means. An object may be contiguous and prior to another, without being considered as its cause. There is a necessary connection to be taken into consideration; and that relation is of much greater importance, than any of the other two above-mentioned. (thn Book i, Part iii, Section ii)

But what kind of necessitation is involved in causation? In the analysis of the causal relationship, we have seen that we can think about the cause without thinking about the effect. Therefore, particular causes and their effects maintain neither a logical nor a conceptual, nor a metaphysical relationship; in other words, there are possible worlds wherein the first ball’s movement does not cause the movement of the second. However, we need some kind of necessitation from the cause to the effect that helps us distinguish between causal sequences of events and accidental or non-casual sequences. Such a necessity has been interpreted as a kind of nomological necessity, grounded in the natural laws of our world. This interpretation has led contemporary theorists to develop the more accepted approach today, namely, what has been called both nomological and regularity theory of causation14. 14

See its characterization in, for example, Pap 1962 271, Lewis 1973a 556, Sosa 1993 234, Kim 1973a 5.

31

Juan Diego Morales Otero

I think it is beneficial to follow the tradition at this point: the causal relation between two events, say a and b, involves a kind of nomological necessity according to which, given the existence of a natural law (or set of natural laws) l, the occurrence of a and some conditions c make b occur (where the law(s) l is (are) essential, given that a and c alone do not necessitate b). Moreover, most, if not all, theorists concerned with the problems of the mind-body connection, emergence and mental causation (the issues that centrally concern us here), explicitly or implicitly assume a minimal characterization of causation in nomological terms15. This nomological model supports the analysis of causation developed by J. L. Mackie in terms of necessary and sufficient conditions. For this philosopher, a cause is at least «an insufficient but necessary part of a condition which is itself unnecessary but sufficient for the result.» (Mackie 1993 34) This is what Mackie calls an inus condition, given the initials of ‘insufficient’, ‘necessary’, ‘unnecessary’, and ‘sufficient’. Mackie explains the concept through an example: experts find that a short circuit caused the fire in a house. However, this short circuit is not necessary for the fire, because the fire could have occurred due to a different short circuit, perhaps at a different place, or to the overturning of a lighted oil stove, or to a different event. But the short circuit is not a sufficient condition for the fire, because if there were no flammable materials nearby, or if there was an efficient automatic sprinkler at just the right spot, the fire would not have occurred. In what sense, then, did the short circuit cause the fire? Mackie tells us, At least part of the answer is that there is a set of conditions (of which some are positive and some are negative), including the presence of inflammable material, the absence of a suitably placed sprinkler, and no doubt quite a number of others, which combined with the short-circuit constituted a complex condition that was sufficient for the house’s catching fire – sufficient, but not necessary, for the fire could have started in other ways. Also, of this complex condition, the 15

32

See, for example, Fodor 1990, McLaughlin 1992, Kim 1993c and 2010a, Block 2003, Shoemaker 2007, Papineau 2008. As we shall see, many of the criticisms of the covering law model are based on the affirmation of its insufficiency, not of its emptiness or irrelevance.

Causation: A Non-Reductive Approach

s­ hort-circuit was an indispensable part: the other parts of this condition, conjoined with one another in the absence of the short-circuit, would not have produced the fire. The short-circuit which is said to have caused the fire is thus an indispensable part of a complex sufficient (but not necessary) condition of the fire. (1993 34)

But, what is the relationship between Mackie’s analysis and the covering law model of causation? According to Mackie, when we say that the short circuit in this house was a necessary condition for the fire, what we are saying is that there are certain universals or laws from which, together with the conditions of the house and the assumption that the short circuit did not happen, it can be derived that the house did not catch fire (Mackie 1993 49). Similarly, when we say that the short circuit was a sufficient condition, we say that from statements about the short circuit, laws or universal propositions, and the conditions of the house, it follows that the house caught fire. So, we can say that in this interpretation a cause a is a sufficient condition, and therefore determines its effect, b, only together with conditions c and thanks to a (some) law(s) of nature l. Despite the wide acceptance of the nomological approach even today, many theorists have considered it incorrect or at least insufficient. The work carried out for at least the last century about the concept of causation has produced the idea that there seems to be no reductive definition of cause and, thus, the approach attempting to provide necessary and sufficient conditions of causation is at least implausible16. Given that the initial nomological theories developed as approaches aimed at reducing the concept of causation to the concepts of regularity and natural law17, philosophers of recent decades have built critical, alternative, and complementary approaches that have enabled the development of different theories of causation: counterfactual, probabilistic, processual, of physical transference, intervention, manipulability, among others. Now, while most of these theories are still reductionist approaches that attempt to analyze or identify 16 17

See, for example, Tooley 1990, Sosa & Tooley 1993. For example, Ayer 1936 and Reichenbach 1951.

33

Juan Diego Morales Otero

causation with other relations, it is more beneficial to examine their appearance and development, insofar as they are theories that find traits and conditions, which, although not sufficient, are part of the phenomenon of causation. One of the main problems of the theories that aim to explain causation in terms of nomological regularities is the existence of nomological regular processes that are not causal. For example, an event b may be contiguous, previous and nomologically correlated with an event c, because both b and c are effects of the same cause, the event a (Salmon 1984). There are many medical cases where a pathological condition (a) causes various symptoms occurring one after the other (b and c). Although we find a nomological and projectible regularity between b and c, this occurs only because of the underlying causal relationship between a and b, on the one hand, and between a and c, on the other, and not because of an actual causal link between b and c18. A similar case appears in the scope of nomological regularities of the special sciences in general, from chemistry to sociology. Primarily on the basis of the principles of the causal closure of the physical domain («If a physical event has a cause that occurs at t, it has a physical cause that occurs at t.» Kim 2005 43) and of causal exclusion («no effect has more than one sufficient cause unless it is overdetermined» Karen Bennett 2008 281)19, authors with reductionist tendencies like Kim (virtually in all his works since the 90’s) and Papineau (2008, 2010) have recently argued that higher or emergent regularities can only be causal insofar as they are identifiable with physical regularities on which they supervene/emerge; otherwise, they will end up being epiphenomenal. The idea is that if we accept that higher regularities are derived from, because they supervene on and are determined by physical regularities, and we still maintain that they are multiply realizable, and therefore irreducible to the latter, then, we have to say that they must be epiphenomenal. This is because, given the first point, they will be causal and projectible only because of the causal power of Shapiro & Sober 2007 consider the case of the relationship between phenotypic and genotypic causal determinations with the same result. 19 These principles and the arguments interwoven around them will be discussed in detail in chapters 4 and 5. 18

34

Causation: A Non-Reductive Approach

their physical bases, without the capacity to add any causal element; and, given the second, they could not be identified with these bases. In this way, they will not be strictly causal regularities. The motto of the argument is this: either reduction or epiphenomenalism. Assuming irreducibility, Kim describes the situation in these terms: Things don’t improve much if this suggestion is strengthened by adding that these regularities must be nomological, with a modal force appropriate for laws of nature, and be able to support counterfactuals, be projectible, and so on. […] There surely are interesting and useful laws and regularities in the special sciences. According to the exclusion/ supervenience arguments, these regularities are no causal regularities, although they may arise from genuine underlying causal regularities; mental-mental [in general, special-special] regularities are no more causal than the regular connections seen in a series of shadows cast by a moving car. (2009 45)

If this is correct, higher regularities, just as successive symptoms caused by a common cause, will be projectible, nomological and yet non-causal.20 In this situation, a nomological character does not provide a guarantee of causal power. Counterfactual Theory of Causation

An approach to the concept of causation which apparently does not need to appeal to laws of nature, and has gained great importance and support in recent years is the counterfactual theory. The basic idea of c​​ ounterfactual theories of causation is that statements of causality can be explained and understood in terms of counterfactuals such as «if a (the cause) had not occurred, b (the effect) would not have happened» and «if a had occurred, b would have happened.»21 There are different types of problems a philosophical analysis of causation in terms of counterfactuals must face22, especially regarding I agree with Shapiro & Sober 2007 that the case of higher level regularities is different from that of the effects of common causes (see chapter 4). 21 See, for example, Lewis 1973a 557, Sosa &Tooley 1993 26, Menzies 2009 1. 22 For a sketch of these problems see, for example, Collins, Hall & Paul 2004, Menzies 2009. 20

35

Juan Diego Morales Otero

its semantics, that is, the conditions under which such statements can be evaluated as true or false. We find two major approaches to understanding this semantics: (i) the derivational-nomological model, and (ii) the possible worlds approach. In the derivational-nomological position, a conditional counterfactual that makes explicit a cause as a sufficient condition for its effect of the kind «if a had occurred, b would have occurred» is true in the case a description of b can be derived from a description of a, with certain statements about laws and the conditions standing at the moment. To give a traditional example, we can say that the counterfactual «if this match had been struck, it would have lighted» is true only when the consequent of the conditional can be derived from the antecedent, together with the statement of the law «whenever a dry match is struck in the presence of oxygen it ignites», and the auxiliary premises «the match was dry» and «oxygen was present». It should be obvious that under the derivational-nomological theory of the counterfactuals’ semantics, the problems we have seen about regular and nomological non-causal processes reappear. However, most theorists advocating a counterfactual approach of causation use the semantics developed by Robert Stalnaker (1968) and David Lewis (1973b), which is based on the comparative similarity among possible worlds. Although many authors who favor this semantics claim that this approach is independent of the nomological theory, there are good arguments which affirm otherwise, namely, that the approach of possible worlds’ semantics ends up resorting to a nomological analysis and its associated problems23. Kim, for example, develops the following argument (2011 212-3). Suppose that Brian’s desire to check out the noise causes him to go downstairs, and that we explain this situation appealing to the true counterfactual «if Brian had not wanted to check out the noise, he wouldn’t have gone downstairs». Now let us consider the following two worlds: w1: Brian did not want to check out the noise; he did not go downstairs.

23

36

See, for example, K. Bennett 2003, Kim 2007b, 2009, Robb & Heil 2013.

Causation: A Non-Reductive Approach

w2: Brian did not want to check out the noise; he went downstairs anyway. Clearly, if w2 is closer to the actual world than w1 is, then our counterfactual would not be true. But if it is true that Brian’s desire caused him to go downstairs, then the counterfactual must be true; and, if it is true, w1 must be closer to the actual world than w2. However, it seems that considering only the actual world’s facts, whereby Brian wanted to check out the noise and he went downstairs, we have to say that w2 is closer to the actual world than w1 —inasmuch as w1 does not share any relevant fact with the actual world, whereas w2 does, namely, the fact that Brian went downstairs. Why, then, do we consider that w1 is closer to the actual world than w2? Kim affirms,

The only plausible answer, again, seems to be something like this: We know, or believe, that there are certain lawful regularities and propensities governing Brian’s wants, beliefs, and so on, on the one hand, and his behavior patterns, on the other, and that, given the absence of something like a desire to check out a suspicious noise, along with other conditions prevailing at the time, his not going downstairs at that particular time fits these regularities and propensities better than the supposition that he would have gone downstairs at that time. (2011 212)

Returning to an earlier example, we can say that according to the counterfactual theory the statement «if this match had not been struck, it would not have lighted» explains the causal relation between striking a match and lighting it. Following the approach of Lewis and Stalnaker, this counterfactual is true because a possible world1 in which the match is not struck and does not light is closer to the actual world than a possible world2 wherein the match is not struck and does light, and a possible world3 wherein the match is struck and does not light. But, again, we must ask why the first possible world is closer to the actual world than the second and the third. Let us consider each world’s relevant facts: Actual world : {the match is struck; it does light} w1: {the match is not struck; it does not light} w2: {the match is not struck; it does light} w3: {the match is struck; it does not light}

37

Juan Diego Morales Otero

It is clear that regarding their mere facts, w2 and w3 are closer to wa than w1, since they share with wa at least one of their relevant facts, while w1 does not share any of these. So, why do we consider w1 to be closer to wa than w2 and w3? Kim appears to be right when he argues that it should be that certain nomological regularities maintained in the different worlds make w1 closer to wa. In particular, this is because the law «whenever a dry match is struck (in the presence of oxygen) it ignites» is instantiated in both wa and w1, while it is instantiated neither in w2 nor w3. If this is right, the counterfactual approximation to causation returns to the nomological characterization and, therefore, to its problems. Furthermore, the nomological perspective can happily receive the counterfactual analysis’ advances without assuming a reduction to it. As K. Bennett tells us, «[p]roponents of nomological accounts of causation should not object to the reliance upon counterfactuals; it might be the case that the counterfactuals hold precisely because certain strict laws do.» (2003, note 10) This is the conclusion suggested by the problem of the causal counterfactuals’ semantic determination: we must appeal to laws of nature in order to account for the dependency at issue. But yet other authors argue that the counterfactual approach to causation, regardless of whether or not it depends on nomological factors, is insufficient to account for causal processes and interactions. This statement can be based on two further points: (i) the relation of counterfactual dependency is maintained not only between causally related events, but also between any entities holding some sort (as logical, metaphysical, mereological, etc.) of determination, so that, as Kim says, «counterfactual dependency is too broad to pin down causal dependency.» (1973b 206) (ii) both the problem of causal forks in medicine and that of causation in the special sciences exemplify regular successions of events grounded in underlying causal processes that ensure the truth of counterfactuals about such regularities, even though they invalidate their causal character. In view of these cases, Papineau argues that, «notwithstanding all the recent enthusiasm for counterfactual theories of causation, it is by no means clear that the

38

Causation: A Non-Reductive Approach

mere truth of such a counterfactual is sufficient to vindicate a special fact as a cause». (2010 193) Causation and Locality

I have claimed that the counterfactual approach seems to return to the problems of the nomological characterization, and that, for this reason, it is insufficient to account for causation. In this section, I will argue that the counterfactual approach can neither explain nor get rid of an important constituent of causation, namely, the principle of locality or spatiotemporal continuity between causally related events. In line with the idea that a counterfactual dependence, or at least a chain of counterfactual dependence relations, between completely different events is sufficient for causation, some authors have claimed that only a counterfactual approach which does not resort to the principle of locality can constitute a non-reductive theory of mental and higher level causation.24 In the final section of the chapter, I will argue that this is not the case and that even recognizing that causation implies the principle of locality, we still have very important reasons to acknowledge a very important and viable non-reductive approach to mental and higher level causation. Especially in recent decades, some theorists have argued that in order to account for causation in a real physical world, where a clear distinction between causal and pseudo-causal processes and interactions must be made, we must develop a theory of causation that starts from what science tells us about the world, its structure and dynamics. The so-called conserved quantity approach to causation argues that the only way to distinguish between causal processes from non-causal sequences of events is appealing to the existence of transference of energy or conserved physical quantities between the events or processes causally related.25 This kind of theory aims to be an empirical and usually reductive approach to the nature of causation, in which it is important to consider two basic concepts: causal process and causal interaction. As Dowe 24 25

See, for example, LePore & Loewer 1987, Horgan 1989, Loewer 2007. See, for example, Salmon 1994, Dowe 1995, Kistler 2006a.

39

Juan Diego Morales Otero

says, a «causal process is a world line of an object which possesses a conserved quantity. […] A causal interaction is an intersection of world lines which involves exchange of a conserved quantity.» (1995 323) The basic idea of the conserved quantity theory is that a causal interaction between two processes or events transmits a local modification between them. Such a local modification consists of an exchange or transfer of a physical quantity. We have seen that since Hume, spatiotemporal contiguity or locality constitutes a defining part of causation. We can characterize the locality principle in the following way: causes are connected to their effects via spatiotemporally continuous sequences of causal intermediates. As the characterization of the conserved quantity theory makes clear, the idea about a​​ connection between processes and events that exchange a physical quantity such as energy presupposes the principle of locality and, therefore, preserves this traditional tenet as one of the principal components of causation. A further problem of a counterfactual approach of causation is its inadequacy to account for the principle of locality. In fact, the proponents of this approach typically assume that this is one of the advantages of the counterfactual dependence, since it would allow us to explain alleged cases of causation not involving any local physical interaction: omissions and the so-called cases of ‘double prevention’. If we recognize cases of causation not involving local interactions, we should affirm that, as Ned Hall claims, Causation, understood as a relation between events, comes in at least two basic and fundamentally different varieties. One of these, which I call ‘dependence’, is simply that: counterfactual dependence between wholly distinct events. In this sense, event c is a cause of (distinct) event e just in case e depends on c; that is, just in case, had c not occurred, e would not have occurred. The second variety is rather more difficult to characterize, but we evoke it when we say of an event c that it helps to generate or bring about or produce another event e, and for that reason I call it ‘production’. (2004 225)

But, don’t such cases as omissions and double prevention involve local interactions? This is Hall’s main basis for affirming the supposed

40

Causation: A Non-Reductive Approach

distinction between these types of causation. Nevertheless, as we will see, there are good reasons to affirm that these cases do not violate the principle of locality, and, therefore, that there is no compelling argument to reject this traditional tenet as one of the principal components of causation. If this is correct, it seems that we do not need to claim that causation comes in fundamentally different varieties, and that we can consider it as constituting a singular and unitary phenomenon. Let us then begin with the cases of double prevention. We can use a similar example to Hall’s to describe cases of double prevention which apparently do not respect locality. Let us say John is a fighter pilot who is going to bomb an enemy base, while Mary is another pilot flying over an area remote from John’s position and who intersects an enemy plane that, if not stopped, will destroy John’s plane and thus prevent his bombing. Mary destroys the enemy plane and prevents the destruction of John’s airplane and, thus, allows the occurrence of his bombing of the enemy’s base. As in the story the fact of destroying the enemy plane is a necessary, although insufficient, condition for the occurrence of John’s bombing, we can say that the destruction of the enemy by Mary is a cause of the latter. According to the counterfactual theory of causation, this is confirmed by the fact that the conditional «if Mary had not shot down the enemy plane, John would not have thrown the bomb» is true. Now, Hall asks us to imagine that the destruction of enemy aircraft by Mary has happened several hours before and several kilometers away from the release of the bomb by John. Based on the spatiotemporal distance, Hall argues that although there is a dependency relation between these two events that enables us to state a causal connection, there exists neither a physical nor a local connection and, therefore, we cannot sustain the existence of a causal productive link, in which an exchange of some physical conserved quantity is involved. In this case, according to Hall, we do not need spatiotemporal contiguity to have causation. But the alluded spatiotemporal distance between the destruction of enemy aircraft by Mary and the release of the bomb by John does not imply a spatiotemporal disconnection. Moreover, it is not at

41

Juan Diego Morales Otero

all clear that there is not a chain of physical and local connections between these two events. Recall that if Mary had not shot down the enemy, he would have destroyed John’s plane; so there is a clear spatiotemporal connection which could have been used by the enemy for his purposes. We can say that this spatiotemporal connection had been constituted by at least the following events: (a) the enemy flying over an area remote from John’s position at the spatiotemporal zone s1t1, (b) his flying to a zone near John’s plane at s2t2, and finally (c) his intersection and destruction of John’s plane at s3t3. His problem was that Mary found and destroyed his plane at s1t1. Now, although the events brought about by Mary were different, they are still spatiotemporally connected: (a’) Mary destroys the enemy plane at s1t1, (b’) the enemy did not approach John’s airplane at s2t2, and (c’) the enemy did not destroy John’s airplane at s3t3. And it is precisely because (a’) occurred that the event (d’), John’s bombing the enemy base at s4t4, occurred. As we said before, it is for this reason that (a’) is a necessary condition for and a cause of (d’). Moreover, we can reconstruct this spatiotemporal chain of connected events in terms of counterfactual and causal relations: it is not just that (a’) is necessary for (d’) but, given the circumstances, it is necessary for and a cause of (b’) and (c’). (b’) is a cause of (c’) and (d’), and (c’) a cause of (d’). Here there is no problem finding a local connection between these different events and, therefore, considering them as causally related in a productive way. The problem, rather, is that some of these events are not positive happenings but failures of events or omissions: the enemy did not approach John’s airplane, did not destroy it, and did not stop the bombing. Hall recognizes at least the possibility that we can «grant that these omissions exist and are located where the events omitted would have occurred» (2004 243). In this case, some omissions can be considered spatiotemporally continuous events which can be causally and productively related. But Kim considers that there are other kinds of omissions that should be understood as locally disconnected with their supposed effects. Following the principle of locality, Kim considers that these situations cannot be really understood as constituting causal interactions. Although I agree with Kim about the necessary

42

Causation: A Non-Reductive Approach

spatiotemporal connection that must exist between the causally related events, I think that there are reasons to claim that the kind of omissions that initially appear to violate the principle of locality, in fact do not. So I will claim with Kim and the conserved quantity theory that causation should preserve the principle of locality, but with Hall that omissions should be considered as real causal factors; this is because we can make a case for the idea that omissions are really spatiotemporally connected to their effects. Let us follow Kim’s example. We usually accept that the truth of the counterfactual «if Mary had watered my plants, the plants would not have died» shows that the fact that Mary had not watered the plants, Mary’s omission, is the cause of the plants’ death. But Kim says: «obviously there was no flow of energy from Mary to the plants during my absence (that exactly was the problem!); nor was there any other physical connection, or any spatio-temporally contiguous chain of causally connected events» (2007b 258-9). Suppose that Mary had promised to water the plants in my yard on Monday morning. And it was because Mary did not water them on Monday morning that they died on Monday afternoon. Let us analyze the event that caused the plants to die. This event occurred in my yard on Monday morning, and consists of certain physical, chemical, and biological processes that happened to my plants. But we can also say that this is the event which my cat witnessed on Monday morning, which produced a special smell in the house, and Mary promised that she would intervene and ultimately didn’t, that is, its omission. But if this is correct, if the event which occurred in my yard on Monday morning is the same as Mary let happen and so constitutes her omission, then there is actually a spatiotemporal connection between this omission and the death of my plants and, thus, these events can maintain a productive causal relationship. If this is correct, the problem of omissions as causal factors is not about their spatiotemporal disconnection with their alleged effects. It seems that omissions must have a location in a specific place and time, and it seems that the most plausible candidate for this location is the place and time at which the omitted action should have been done. The real problem is rather that there seems no clear way to discard an

43

Juan Diego Morales Otero

indefinite number of irrelevant and inappropriate omissions which could be cited as causal factors of the same event. Let us follow the previous example, about the causes of the death of my plants and see the following counterfactuals: «If George W. Bush had watered my plants, the plants would not have died. If Laura Bush had watered my plants, the plants would not have died. If Hillary Clinton had watered my plants, the plants would not have died.» (Kim 2007b 259) Obviously, it was not the failure of George W. Bush, or Laura Bush, or Hillary Clinton that killed my plants. My plants died because Mary, who promised to water them on Monday morning, did not. But why do we attribute the omission to Mary and not to Bush or Clinton? That is, why does the event that caused my plants die is, or has the property of being, the omission of Mary and not that of Bush or Clinton? The answer is that this event cannot be the omission of Bush or Clinton because they were not the ones who promised to water my plants. This event can only be understood and considered as an omission within an intentional context, a context of promises, statements, beliefs, desires, decisions, etc. It is clear that if we analyze this context we will easily realize that it is Mary’s omission. That is, we will say that part of the cause of my plants’ death is her omission. A further problem is whether the intentional property of being an omission (the property of being an event that someone promised she would not let happen but did) can count as an additional factor to the physical, chemical, and biological properties of the event at issue. This question, however, is much more complex and constitutes precisely the main object of this monograph as a whole, whose answer will begin to be articulated in the final section of this chapter. On the basis of the idea that omissions and double prevention cases cannot be properly understood as locally connected with their effects, Hall believes that we have to hold that there are two different kinds of causation, one of mere counterfactual dependence and another of production. And Kim concludes that we should exclude them as causes; although there is a counterfactual dependency between wholly distinct events that in principle we consider causally related, Kim thinks that their connection is not really physical and, therefore, that

44

Causation: A Non-Reductive Approach

their dependence cannot be causal. What is the kind of determination that explains this alleged non-causal dependency? Kim does not say anything about it. What is clear to Kim is that causation implies a kind of dependency that can be expressed counterfactually. But, unlike Hall, Kim thinks that causation is not what must be understood in terms of counterfactual dependence, but rather that, in his words, «it may well be that the dependency involved […] has an explanation in terms of the productive/generative relations» (2007b 258). Although I do not agree with Kim’s treatment of omissions, I agree with his overall consideration: counterfactual dependency, or at least a chain of counterfactual dependency among different events, is a necessary, though not sufficient, condition for causation. In order to reach causation, we need much more than dependency; we need a real, local, and physical connection, ensured by the transference of an amount of energy or any conserved quantity between the cause and the effect. This implies a local connection which, in fact, would explain the dependency in question, and would allow us to distinguish between real causal processes and pseudo-causal interactions. Although I have not provided a definitive argument to sustain the necessity of the principle of locality, I have given reasons for holding that intuitive cases of causation which initially appear to violate this principle in fact do not. Now, according to this account of a strong, real, productive causation involving some transfer of a physical quantity between the events or processes causally related, what is the place of the nomological approach? What is the role of the laws of nature in this framework of explanation? This theory holds that the conservation laws of physical quantities such as energy, charge and momentum are necessary for any causal transaction. But they are not enough. The rationale is that conservation laws are symmetric whereas causal laws are asymmetric. If a causes b, then there is a transfer of a physical quantity that is conserved in virtue of a natural law, like the energy conservation law according to which the total energy of an isolated system cannot change. But this kind of law does not determine the order of the events, so it is compatible with the fact that a causes b, as

45

Juan Diego Morales Otero

well as with the fact that b causes a.26 To account for the asymmetry between a and b, we must resort to causal laws, which give a central role to the properties of the causally related events. This is the proposal set forth by Kistler: We shall say that the fact that the cause c has property f is causally responsible for the fact that the effect e has property g if and only if, first, c and e are related as cause and effect in terms of transference, and second, there is a law linking f and g that is instantiated in the situation. (Kistler 2006a 232-3 note 83)

In this sense, we can say that the nomological approach to causation remains largely correct: a causal transaction involves and requires the instantiation of laws, both causal and of conservation, according to which a cause is a necessary element within some conditions or causal context, which together form a sufficient condition for the effect. Before analyzing in more detail the relevance of properties and causal laws for causation, let us see how this proposal can be refined by introducing some conditions that seem to be included in the causal transactions of our world, namely, indeterminacy and intervention. Probability and Intervention

According to its most widely accepted interpretation, quantum theory, perhaps the physical theory that provides the most powerful and accurate explanations and predictions in science, argues that the dynamics and processes of the microscopic level of the world are essentially indeterminate and probabilistic. This fundamental fact, together with the now widespread use of probability in different sciences and theories, has led researchers interested in the nature of causation to develop theories that consider probability an essential factor. The central idea of this approach is that causes are events that increase the probability of their effects. As Hitchcock comments, [t]he central idea behind probabilistic theories of causation is that causes change the probability of their effects […] Thus, smoking is a cause of lung cancer, not because all smokers develop lung cancer, 26

46

See Cartwright 1979 429, Aronson 1982 301-2, Kistler 2006a 68.

Causation: A Non-Reductive Approach

but because smokers are more likely to develop lung cancer than nonsmokers. (2012 §1.1)

Now, if processes within our world, among them causal processes, are essentially indeterminate, then causes cannot be sufficient conditions for their effects. We should then reformulate the idea of ​​a cause as an inus condition; we can say a cause is at least an insufficient but necessary part of a condition which is itself unnecessary but sufficient to increase the probability of its effects. Thus, the sufficient condition for the effect in an indeterminate world will be sufficient only for increasing the probability of the latter; and, therefore, the existence of the cause, or the instantiation of properties that given certain natural laws increase the probability of the effects, will be compatible with some sporadic failures of the effect’s happening. So, if smoking causes lung cancer, that is, it causes its probability to increase, then the fact that Charles is a frequent smoker, along with certain conditions or causal context, increase the probability of his developing lung cancer, for example, from a probability of 0.2 to 0.9. We have seen that the characterization of causal relations leads us to speak of two kinds of natural laws: conservation and strictly causal laws, where the latter explain the asymmetric character of causation. Probabilistic theorists argue that this asymmetric nature can be explained in terms of probability: it is because, by definition, causes increase the probability of their effects, while the effects do not increase the probability of causes, that we can distinguish the order of dependency and determination of causation. Thus, we can understand the direction of the causal link between the fact of frequent smoking and the development of lung cancer, because the first state of affairs increases the probability of the second, while this does not occur in the opposite direction. Another theory of causation that has reached wide acceptance in recent years is the interventionist theory, according to which we can only say that a is a cause of b if doing an intervention on a is a way of intervening on b.27 This approach is based on the central idea that 27

See Woodward & Hitchcock 2003, and Woodward 2013.

47

Juan Diego Morales Otero

causes should make differences. If a causes b then an intervention on a will make a difference to b. In turn, the idea of making differences can be understood in the terms we have discussed about the sufficiency and/ or necessity of a cause with respect to its effect. If a, along with certain conditions, is sufficient and/or necessary to increase the probability of b, thus, an intervention on a will necessarily produce a change and an intervention on the probability of b. As we will return to the analysis of this theory in relation to the problem of mental and higher level causation,28 it only remains to state the way it would be compatible with the characteristics of causation we have found so far. In this case, we can say that a causal relation of the type a causes b implies: 1. A counterfactual dependence, or at least a chain of counterfactual dependences, between a and b. 2. A transfer from a to b of a physical quantity that is conserved in virtue of a law of nature. 3. That a is at least an inus condition for b, in the sense that a along with a causal context is sufficient to increase the probability of b. 4. That an intervention on a makes a difference in (or entails an intervention on) b. 5. Finally, that a causes b only because a has (or is the instantiation of) certain properties a, b and c related to the properties d, e and f of b in virtue of natural causal laws, in the sense that a’s properties increase the probability of the instantiation of b’s properties as a nomological matter. Causal Context and the Possibility of Causation in the Special Sciences

We have seen that a cause produces its effect only within a causal context or conditions. However, as necessary parts for sufficiency, all these conditions can be considered causes of the nomologically necessitated effect,29 so our choice of ‘the’ cause in many, if not all cases, can be seen as an arbitrary or pragmatic choice. In this sense, Lewis comments: 28 29

48

See chapter 4. See Bunge 1959 48-9, Hall 2004 228, and Strevens 2007 96.

Causation: A Non-Reductive Approach

We sometimes single out one among all the causes of some event and call it ‘the’ cause, as if there were no others. Or we single out a few as the ‘causes’, calling the rest mere ‘causal factors’ or ‘causal conditions’. Or we speak of the ‘decisive’ or ‘real’ or ‘principal’ cause. We may select the abnormal or extraordinary causes, or those under human control, or those we deem good or bad, or just those we want to talk about. (1973a 558-9)

Given this framework of explanation, I want to emphasize two points: on the one hand, if there is actually a causal relationship between, for example, the lightning bolt and the forest fire, then, there must be a causal law relating the first as a sufficient condition for the second only together with a necessary causal context, the fact that there is enough oxygen and enough flammable material in the environment, among other conditions, that may be required. Putting things this way, we can say that the natural law which relates to the striking of the lightning bolt and the forest fire is what we might call, at least in a weak sense, a ceteris paribus law, since the first fact is sufficient for the second only if certain conditions are maintained. In this sense, Robb and Heil tell us, for example, that «the properties of being a match-striking and being a match-lighting [are related]. If there is a law connecting such properties, it is evidently non-strict: striking causes lighting only ceteris paribus» (2013 §5.3). And, on the other hand, as we see, the context usually used for the sufficiency of any of the events we are considering as causes is an external context; in the case of the forest fire, the existence of enough oxygen and flammable material in the environment. We can call these conditions, which are extensionally located outside the event considered as a cause, the external context. However, as we will see, we must pay special attention and emphasis to what might be called the internal context, namely, the different conditions or properties that the event as a cause must have and that become relevant to its sufficiency as a cause of another event. Returning to the first point, it is fair to say that most theorists argue that the laws of physical science, such as Newton’s second law (f = ma) or the equivalence law of mass and energy (e = mc2), are

49

Juan Diego Morales Otero

universal and strict, in the sense of having no exceptions.30 In contrast with this idea, the laws of the special sciences, from chemistry to sociology, have been conceived as non-universal, non-strict, with exceptions, as laws maintained only thanks to ceteris paribus clauses, that is, as laws that apply only when certain conditions are met. In recent decades, there has been much controversy about the validity and even the very interpretation of the ceteris paribus (cp) laws. One of the fundamental problems of this kind of laws is their risk of vacuity; if it is stated, for example, that type a events are nomologically sufficient for type b events, ceteris paribus, then there may be (recurring) a instantiations that are not followed by b instantiations, which would be explained by a simple appeal to the absence of the required circumstances. If such circumstances are not specified, the law would simply say «as cause bs only if this is the case». As Reutlinger, Schurz and Hüttemann comment, If we instead suppose that an indefinite exclusive ceteris paribus clause is attached to the law so that it means «All as are bs, if nothing interferes», then the cp-law in question is in danger of lacking empirical content. It lacks empirical content because it seems to say nothing more than «All as are bs or not-(All as are bs)». If this is true, then exclusive cp-laws are analytically true sentences and, therefore, trivially true. (2011 §4)

In the same sense, Kim claims31: But how can a ceteris paribus law, a law whose antecedent, say f-events, does not necessitate its consequent, g-events, ground a causal relation between f-events and g-events? Given the law, it is amply possible for an f-event to occur without being followed by a g-event. Being qualified by a ceteris paribus clause, the law is immune to falsification by such counter-instances; that is exactly the point of ceteris paribus hedges. (Kim 2007b 247) Remarkably, Cartwright 1980 and Lange 1993 deny this claim. Nevertheless, their arguments could be framed in what we will call definite ceteris paribus laws or ceteris paribus laws with completions. 31 See, also, Papineau 2010 181. 30

50

Causation: A Non-Reductive Approach

From this problem we could develop a characterization of the ceteris paribus (cp) laws in terms of what some authors have denominated definite cp laws (Reutlinger, Schurz & Hüttemann 2011) and others call lazy cp laws (Earman, Roberts & Smith 2002), namely, a cp law that «specifies the disturbing factors which are excluded (or the validity conditions which are required) in the antecedent of the law.» (Reutlinger, Schurz & Hüttemann 2011 §3.2) It seems, for example, that Fodor’s proposal is developed in these terms. He argues that a causal law of the type cp a → b is true non vacuously if (a) there is a completer c, such that a & c are strictly sufficient for b; (b) it is not the case that a is sufficient for b, and (c) it is not the case that c is sufficient for b (Fodor 1991 23). But if this is the case, what initially was articulated as a ceteris paribus law becomes just another strict law, whose distinguishing feature is the reduction in the scope of its application. At this point, we can even question the very existence of cp laws, simply because, in the words of Reutlinger, Schurz & Hüttemann, «[i]t is a widely agreed in philosophy that the real significance of exclusive cp-laws lies in situations where strict completion is impossible» (2011 §3.2). Ceteris paribus laws have been introduced mainly with the purpose of grounding a realist perspective of higher level causation or causation of the special sciences. But I think this is a mistake. Let us then agree with Fodor and Kim (2007b 247-8) that, no matter the label we give them, laws cannot be real laws, unless, given a causal context or conditions, they make their causal instantiations strictly sufficient for their effects (or for the increase of the objective probability of their effects). In this case, we could say with Kim that all laws end up being strict (some with larger extension -those being true without conditions- and others with less extension -those true only in specific conditions), and, as a result, may support a determination relation between the events instantiating them. Although Fodor and Kim would finally agree about the nature of cp laws in relation to the conditions in which they instantiate (which would make them strict laws), these authors would not agree about the possibility of reduction of the special sciences and laws. But I think the point about the possibility of higher or emerging laws, not reducible to

51

Juan Diego Morales Otero

physical sciences, does not reside in the nature of ceteris paribus laws, at least not in the examined sense, simply because the latter depends on the presence or absence of external causal conditions, overlooking the possibility and relevance of internal causal context or conditions which, as I will argue, is the key element that allows us to articulate a non-reductive causation in special sciences. The basic point from where we must start and that is now widely accepted is the fact that, as Papineau affirms, «causes are in some sense property-involving. This will be true if causes are facts, or “Kim-events”, or even if they are Davidsonian-events, that enter into causal relations in virtue of some of their properties» (2010 192). We should then give a fundamental role to properties in our understanding of causation. As we have seen, an event is the instantiation of different properties at a spatiotemporal zone, properties that determine the different types of causal and nomological dynamics in which the event is involved. We can cite an example that Kistler develops on this point: For our purposes, we may conceive of the particular copper wire as of the complex of all properties instantiated at the space it occupies. What acts as a cause is the complex of all properties instantiated at the spatio-temporal region the copper wire occupies at the moment of the causal relation at issue. […] One of these properties, which this particular copper wire shares with all other samples of copper, is its electron structure, more precisely the distribution of its electrons in energy bands in a way characteristic for copper. It is this property that determines many aspects of its behaviour, for example its capacity to conduct electric current. Other properties of the wire determine other aspects of its behaviour: the properties of the crystal lattice made up by its atoms contribute, together with its electron structure, to the wire’s thermal conductivity. (2006a 167-8)

In this way, different and varied properties can converge in the same object and event; together they determine the behavior of the object and the causal contribution that the event may develop. In this case, it is not necessary to go beyond the internal conditions of the event we consider as a cause to account for a determination that is

52

Causation: A Non-Reductive Approach

found at the higher levels, a determination of the properties and laws studied by the special sciences.32 If it is true that different types of properties of the same event are responsible for its different causal dynamics, and if it is true that there are higher level properties, non-reducible to the lower level properties on which they supervene or emerge, just as the non-reductive physicalist argues and which we will have space to examine, then the emergentist or the anti-reductionist theorist can argue that an event a, together with an external context, is causally sufficient for an event b in virtue of its properties a, b, c… (i.e., its internal context, both its microphysical and macrophysical or higher properties), which nomologically interact (in virtue of certain natural causal laws) to bring about b with all its properties f, g, h… It follows that a natural causal law that relates a specific property, say a, to another property f, by making the first’s instantiation increase the probability of the second’s instantiation, will have to interact with other laws of the same event in order to produce the overall increase, decrease or conservation of the probability of f’s instantiation. Following the concept of multiple determination developed by Roy Bhaskar, Elder-Vass claims a similar point, arguing that we must conceive causes and effects as ‘laminated events’ with all their causal and nomological properties going from the micro to the macro levels. This would allow us, as this author says, to «construct causal accounts of multi-leveled single instance causation in which all the levels of the prior situation can have an appropriate influence on the various levels of the outcome» (Elder-Vass 2010 51). Clearly, this argumentative path is favorable to the anti-reductionist theorist only to the extent that it is philosophically and scientifically plausible to postulate t​​ he existence of higher level properties which are dependent but irreducible to their lower bases, properties deploying dynamics which are ruled by laws that special sciences must discover 32

Although the example of the copper wire talks only about lower level physical properties, in chapter 5 we will articulate the idea about the connection of lower level and higher level causal properties and illustrate it with some particular examples.

53

Juan Diego Morales Otero

and articulate. The examination of this possibility must be left, however, for subsequent chapters. Summary

In this chapter, I have examined different philosophical approaches which have tried to define the concept of causation in terms of nomological regularities, counterfactual dependence, and transference of a physical quantity. In the analysis of the different approaches, I have found important reasons to affirm that causation should be understood as incorporating these different elements, without being reduced to them. I claimed that one of the most important features for understanding the nature of causation lies in what I have called the internal context, namely, the internal properties or conditions that the event as a cause must have to be nomologically sufficient for its effect. I have provided different reasons to affirm that the possibility of real causation in the special sciences can be explained from the existence and interaction of these internal conditions.

54

CHAPTER 2 The Concept of the Physical and the Overcoming of the Supervenience Theory Every era has its Weltanschauung and in much contemporary philosophy the doctrine of ‘physicalism’ plays this role. gillett and loewer 2001 ix

This chapter is devoted to a careful analysis of the concept of the physical. I examine the contemporary characterization of the physicalist theory, and attempt to make explicit the necessary conditions for any object, event, or entity to be physical. I argue that the contemporary formulation of physicalism, the theory of the metaphysical supervenience or complete determination of empirical phenomena by the microphysical characteristics, is insufficient both empirically and conceptually. On the one hand, it contradicts results from physical science; and, on the other hand, it contradicts both the scientific and the everyday use of the notion of the physical. We will see that the philosophical tradition can provide the criteria for understanding the concept of the physical, so that physicalism should be understood as the theory that claims that all the entities of our world are essentially objective, necessarily spatiotemporal, and, in principle, explainable according to the mathematical methodology of the natural sciences.

55

Juan Diego Morales Otero

What Physicalism Is

Physicalism is the ontological theory which claims that the entities that constitute our world are physical entities, phenomena that the physical sciences must discover and articulate in their theories.1 Following the naturalistic trend of philosophy in recent times, contemporary philosophers have considered that physicalism is both an a posteriori and contingent thesis. It is a posteriori in the sense that it is a position that tries to overcome the problems of its direct predecessor, materialism. The latter was established as a metaphysical doctrine that attempted to specify the entities of our world in an a priori way, in terms of a specific set of features that supposedly defined the material, such as conservation, deterministic and on contact interaction, impenetrability, inertia, and solidity.2 But this a priori specification proved to be wrong. It is now clear that if any of these conditions is necessary for something to count as material, then physics speaks of immaterial entities.3 Despite the distinction between materialism as an a priori doctrine and physicalism as an a posteriori approach, it is a fact that contemporary philosophers have assimilated the terms ‘materialism’ and ‘physicalism’ in the a posteriori meaning, and, in consequence, they are used interchangeably.4 But a posteriori physicalism faces a not minor problem, the so-called Hempel’s dilemma, which is based on an intuitive distinction between current physical science and complete or ideal future physical science. Hellman puts it in these terms: [C]urrent physics is surely incomplete (even in its ontology) as well as inaccurate (in its laws). This poses a dilemma: either physicalist principles are based on current physics, in which case there is every reason to think they are false; or else they are not, in which case it is, at best, difficult to interpret them, since they are based on a ‘physics’ that does not exist —yet we lack any general criterion of ‘physical Here I will use the general sense of ‘entity’ and ‘phenomenon’ to include both particulars (objects, events, processes, and so on) and what many theorists take as universals (properties, relations, and laws). 2 See, for example, Cavendish 1668, Mettrie 1748, and d’Holbach 1770. 3 See Crane & Mellor 1990 186, and Stoljar 2010 10-11. 4 See, for example, Kim 1992a 122, Block 2003 133-134, and Stoljar 2010 11. 1

56

The Concept of the Physical and the Overcoming of the Supervenience Theory

object, property, or law’ framed independently of physical theory. (Hellman 1985 609)

But physicalists respond to this problem in a very interesting and, I think, successful way, by arguing that their doctrine can be understood in terms of the complete physical science we can find and clearly recognize as a descendant of the current physical science.5 This is Papineau’s formulation: The idea here is to appeal to the categories represented by current Physics Departments, but to allow some wiggle room for future developments. So we might think of ‘physical’ as referring to all those categories that bear some resemblance to the categories recognized in contemporary Physics Departments. For example, ‘physical’ might be understood as equivalent to something like ‘displaying mathematically simple and precise behaviour’. (Papineau 2008 130)

In this case, the physicalist affirms that the entities that constitute our world are those that physical science postulates for its understanding and explanation. It is then clear that physicalism must satisfy some empirical and a posteriori conditions and, therefore, will be true only if the claims of the physical science upon which it rests come to account for our empirical world in a proper and all-sufficient way. Despite this, as Jessica Wilson says, «physicalists have not handed over all authority to physics to determine, a posteriori, what is physical.» (2005 428, ff.) The idea is that physicalism must also respond to some conceptual conditions imposed by the very notions of physical science and physical entity, which, as such, cannot be resolved in an a posteriori way by physical science itself. These are conditions that apply to both the concept we now have of the physical and the one we can ever come to have. It is in this sense that if we recognize, as Papineau suggests and it is intuitive to affirm, that physical science is essentially mathematical and quantitative, then a complete future science intending to be recognized as physical will have to satisfy this condition. If the 5

See, for example, Lewis 1983 361, Jackson 1998a 8, and McLaughlin 2007a note 3.

57

Juan Diego Morales Otero

final science reaches a complete explanation of at least some facts of our world without having to resort to mathematical language, then physicalism will end up being false. It is then clear that physicalism must also satisfy some a priori criteria. In fact, part of my argument is that many of the philosophical problems of the theorists working on the development of a correct characterization of what counts as physicalism arise precisely because of the lack of sufficient clarity about its conceptual and a priori conditions. In order to have an understanding of the concept of the physical and of the physicalist doctrine which should be built on this understanding, I will examine the way in which the contingent aspect of physicalism has been formulated, and the role that the concept of supervenience has played in such formulation. Physicalism as a Contingent Theory

The notion of supervenience6 has been introduced and developed with the primary aim of accounting for two very relevant ideas which are understood as basically compatible: the priority of natural and physical phenomena within our world, and the irreducibility and difference of properties and phenomena that, in principle, we would not call physical, such as mental, moral, political, and economic phenomena. It is precisely this idea that philosophers of morality, such as G. E. Moore and R. M. Hare, and philosophers of mind, like Davidson, Fodor, and Putnam, have in mind when they affirm their naturalistic commitments. We find the classic formulation of the supervenience of mind on the physical in Davidson: [M]ental characteristics are in some sense dependent, or supervenient, on physical characteristics. Such supervenience might be taken to mean that there cannot be two events alike in all physical respects but differing in some mental respect, or that an object cannot alter in some mental respect without altering in some physical respect. (Davidson 1970 214) 6

58

In general terms, we can say that a set of properties a supervenes on another set of properties b, just in case there cannot be a difference in a without a difference in b. For an analysis of the concept, see, for example, Kim 1994, and McLaughlin & Bennett 2014.

The Concept of the Physical and the Overcoming of the Supervenience Theory

But the concept of physical supervenience, which was introduced in order to articulate a non-reductive naturalistic approach, has in recent times served to account for the contingent aspect of physicalism. In order to achieve this goal, in one of the most quoted formulations of the physicalist doctrine, Jackson argues that the physicalist commitment should be interpreted as stating that «[a]ny world which is a minimal physical duplicate of our world is a duplicate simpliciter of our world» (1998a 12), where a minimal physical duplicate is what results from duplicating all and only all the physical facts; nothing more. Most analytical philosophers, whether physicalists or dualists, agree that this formulation captures the contingent aspect of physicalism, that is, the possibility of being true or false given only the facts of our world. And the reason that this formulation captures the contingency of physicalism is that, as Karen Bennett says, it «allows the possibility of worlds physically like ours, but with ghostly ‘extras’, and thus does not require that physicalism be necessarily true» (2008 284).7 The idea then is this: physicalism is the thesis that our world is completely physical, that physical properties determine all its properties. But this statement does not say anything about other possible worlds different from ours; in particular, this statement is consistent with the fact that other possible worlds are, completely or partially, made up of immaterial entities, Cartesian substances, or incorporeal souls. In fact, this is the statement that gives support to the idea that the physicalist theory is contingent about our world: according to this view, it is simply false, but completely coherent, to affirm that our world is constituted at least in part by immaterial entities, Cartesian substances, or incorporeal souls. But here lies a deep problem that most contemporary theorists have not even suspected. The problem, in a nutshell, is that philosophers like Ryle, Wittgenstein, Davidson, Putnam, and Burge, among others, have built very important criticisms of the very intelligibility of Cartesian dualism, and not merely of its empirical possibility.8 See also Chalmers 1996 38-9, Jackson 1994 28, 2007 192-4, Lewis 1983 362, Shoemaker 2010 125-6, and Stalnaker 1996 231, 2004 395. 8 Part of the overall purpose of this dissertation is to take some steps in this direction, through the articulation of an emergentist approach about the 7

59

Juan Diego Morales Otero

But if the contingency of physicalism is based on the possibility of our world being partially constituted by immaterial entities, Cartesian substances, or incorporeal souls, and this possibility is defeated in a conceptual manner, then it seems that physicalism should be accepted as a conceptually necessary theory, which could not be grounded in the empirical findings of science. But this is completely contrary to the fundamental principles of any theory which purports to be physicalist. Against the current unanimous conviction, I think that the contingency of physicalism need not depend on the feasibility of the Cartesian approach. There are different ways in which the physicalist theory could be empirically false. To take an extreme example, perhaps we could consider the possibility of an entity that has no causal or nomological dynamics, which could not qualify as physical. If this is so, then we must say that there are at least possible entities that do not fall within the concept of the physical. We might also remember that it is almost obvious that a panpsychist world, in which electrons, quarks, and superstrings are endowed with intelligence and/or consciousness, cannot qualify as a physical world; it seems that the very concept of physical entity excludes this possibility (we will return to this point). Also, as we have suggested, a world with some nonmathematically expressible phenomena does not seem to qualify as a physicalist world. If so, the idea that physicalism is contingent does not require the controversial possibility of a Cartesian mind. What Is It To Be Physical?

We have seen that physicalism is both an a posteriori and a contingent theory, in the sense that it should respond to the empirical results of the sciences. But we have also seen that this theory must respond to some conceptual conditions presupposed by the very notions of physical science and physical entity, conditions which would allow us to assess whether some eventual final and complete science mind and its relation to the world–particularly in the direction of showing that the irreducibility of the causal powers of mind should be understood as the idea of a non-reductive higher level organization and constraint of the same physical elements that compose other things, and not as a mysterious and plausibly incoherent action of a non-spatial and non-material entity.

60

The Concept of the Physical and the Overcoming of the Supervenience Theory

constitutes or not a physical science. In this section, I introduce the problem of articulating these conditions. We know that most philosophers consider that physicalism should be understood as the idea that the physical characteristics of our world constitutively determine all its empirical characteristics. According to Jackson, for instance, physicalism about phenomenal consciousness (mental states that have qualitative properties or qualia) affirms that the properties required for the constitution and characterization of the non-sentient or non-conscious items of our world, such as water molecules, force fields, cells, and nerve fibers, «are enough for all of our world including, for example, the humans and the cats» (Jackson 2007 193). According to this way of understanding physicalism, physical, chemical, biological, and neurological properties are both ontologically and epistemologically sufficient for the constitution and explanation of animal and human consciousness. In this case, physical simply means ‘non-sentient.’ But here lies one major issue: why and in what sense do chemical, biological, and neurological entities, such as water molecules, cells, and nerve fibers, count as physical? Accepting that chemical and biological elements qualify as physical in their own right is a ubiquitous fact within both the philosophical and the scientific domains. This is why most reductionist theories regarding not only consciousness but mind in general, not to mention moral, economic, political, and related phenomena, are articulated in terms of brain and nerve activities that persons or bodies have. Some philosophers address the issue without giving it real importance. In another text, Jackson states the following about the characterization of what counts as a physical property: Roughly, I will mean what is typically meant: the kinds of properties that figure in, or are explicitly definable in terms of, those that figure in physics, chemistry, biology, and neuroscience. This rough characterisation leaves it open why those sciences, rather than say psychology or politics, are chosen to settle the favoured class, and it says nothing about how committed this approach is to those sciences being roughly right in the kinds of properties they need for their own internal purposes. Nevertheless, I think that the rough characterisation will do for our purposes here. As far as I can see, nothing in what

61

Juan Diego Morales Otero

f­ ollows turns on the answers to these two controversial matters. What is important here is that there is a favoured list, not how a property or relation gets to be on that favoured list. (Jackson 1994 26)

This philosopher seems to suggest that it is not relevant to know the criterion we use to determine that certain elements, such as chemical and biological elements, count as essentially physical. It seems that what is relevant to physicalism is to build a list of items and categories that must be understood as constituting an undeniable part of the physical. On the basis of that list, physicalists must build their version of reality.9 But this approach is unsatisfactory. Let us recall that physicalism is an a posteriori thesis affirming that its empirical foundation must be given by a recognizable descendant of the current physical science. But in order to assess whether some eventual science constitutes or not a physical science, it is not enough to have a list of items currently considered as paradigmatically physical, a list of the form: electrons, atoms, water molecules, cells, nerve fibers, etc. In the same sense, it is not enough to argue that such a final science will rely solely on categories of microphysics, chemistry, biology, and neurology. We do not know if neurology will require appealing to psychological, sociological, or aesthetic properties to account for what happens in the brains of people. In this case, would psychological, sociological, or aesthetic properties count as physical in their own right? Or rather would they falsify physicalism? What we need is a more or less stable criterion about what we understand under the concept of the physical. Microphysical Supervenience as the Criterion for the Physical

Theorists of recent decades have tried to solve this problem by appealing to the understanding that we can acquire from notions such as dependence, determination, and supervenience. The idea is that a theory that deserves the name of physicalism should assert that there are some privileged items on the basis of which everything 9

62

Kim (1993c 340) seems to affirm the same claim.

The Concept of the Physical and the Overcoming of the Supervenience Theory

else is constituted, understood, and explained. If we are physicalists about the mind, we cannot simply say that the mind is one of these physical entities, but we have to show how it is constituted and understandable on the basis of those entities we have always considered to be physical, or those we can identify as physical, irrespective of their mental character. If we are physicalists with respect to biology, we cannot simply assert that biological entities are part of the physical entities, but we have to show how they are constituted, determined, and become understandable on the basis of those entities which have always been considered physical, or those that can be identified as physical regardless of their biological nature. The criterion that has been predominant in recent decades for what we understand by the concept of the physical is based on this idea: chemical, biological, and neurophysiological entities qualify as physical because they supervene on -are metaphysically (in every metaphysical or logical possible world) determined by- their basic physical constituents, that is, their ultimate microphysical elements. Like Pettit (1994), we can call this position microphysicalism.10 My purpose is to argue that microphysicalism (i.e., the thesis that the properties of our world are constitutively determined by the microphysical properties) is neither necessary nor sufficient for understanding the concept of the physical and, therefore, that it cannot provide us with an adequate characterization of what should count as physicalism. For this purpose, it is important to have a clear picture of the microphysicalist thesis and its consequences. This can be achieved by examining the recent controversy about whether physicalism should be understood as affirming that the facts of our world are determined either in an a priori or an a posteriori form by the physical facts, that is, the controversy between what has been called a priori physicalism and a posteriori physicalism.11 Some of the theorists who understand physicalism in this way are Pettit 1994, Chalmers 1996, Kim 2005, McLaughlin 2007, and Shoemaker 2007. 11 About this discussion, see, for example, Block & Stalnaker 1999, Chalmers & Jackson 2001, Jackson 2003, 2007, Kim 2005, and McLaughlin 2007. This differs from the controversy we have examined between materialism and physicalism about the possible definition of the physical appealing or not to 10

63

Juan Diego Morales Otero

The discussion is based on the important difference between a priori necessary truths (for example, that a triangle is a three-sided figure) and a posteriori necessary truths (for example, that water = h2o). Although both the a priori and the a posteriori physicalists claim that all the properties of our world are metaphysically or logically supervenient on the microphysical properties, the former states that all the truths of this world can be in principle deduced in an a priori way from the microphysical truths, while the latter holds that there are epistemic constraints that could override the possibility of such deduction.12 In fact, as both kinds of physicalists argue that the base on which the facts of our world supervene is the set of the microphysical facts, then, it is not only fair but clearer to consider these positions as two kinds of microphysicalism. The a priori microphysicalist would be committed to the idea that the properties, laws, and phenomena that only appear at macrophysical, chemical, and other higher or special sciences levels can, in principle and a priori, be deduced from the basic properties and laws of microphysics, properties and laws of the simplest level that can exist in our world. The basic reason behind this statement is the idea that we can only sustain a physicalist metaphysics if we find that the connection between, on the one hand, the microphysical level and, on the other, the macrophysical levels and the phenomena which in principle could be considered as non-physical (e.g. mental and social items), is not simply a connection of metaphysical supervenience and determination, but a relation -to put it in the terms of various theorists that, it should be clarified, do not necessarily support this idea- of commensurability (O’Regan 2011), intelligibility (Levin 2002), what the physical sciences come to say. As noted by McLaughlin, the labels ‘a priori physicalism’ and ‘a posteriori physicalism’ «are misleading since they suggest that what is at issue is whether physicalism is a priori, and that it is not what is at issue. But they have caught on in the literature, and so are used here» (2007 nota 2). 12 The truths in question are reasonably restricted to those found in the empirical sciences, from (macro) physics to economics, and sociology (see McLaughlin 2007a 204-6). These actually are the relevant domains for our discussion.

64

The Concept of the Physical and the Overcoming of the Supervenience Theory

understanding (Nagel 1974, Kim 2011), and explanation (Levine 1983, Chalmers 1996, Jackson 2007). This theory argues that in order to be physicalists we have to not only maintain the determination of the empirical world by the microphysical facts -the supervenience thesis-, but also give a sufficient reason why such determination is established. The reason is that the truths of the empirical world follow a priori from the truths of microphysics. Thus, this thesis claims that physicalism can only be correct if it is possible to, at least in principle (that is, if we had all the relevant information and the necessary computational power), deduce a priori all the empirical truths of our natural, mental, and social world from the microphysical truths. This would imply that, for example, given the supervenience of a mental property m on a microphysical structure p1, p2… pn, it should be possible to deduce in an a priori form the truths about m from the truths about p1, p2… pn. Consider one of the examples that Jackson uses to illustrate this approach: Consider the very complex sentence that gives the location of every molecule in the universe at some given point in time (we can suppose that there is a finite number for our purposes here). The shape of each and every molecular array in the universe at that time follows a priori from this sentence. One could in principle deduce the shape of every array, but only in principle. In practice the deduction of shape from locations is only possible in cases involving a relatively small number of accessible molecules. (2007 187)

One of the main virtues of this approach is that it would allow us to understand and affirm in a very clear way how and why different domains and properties of our world are completely determined by the basic level of physics. The answer to our question about why the chemical, biological, and neurological levels count as physical, would then be structured as follows: these higher levels are explained, or are explainable, in principle, in a deductive and a priori form only from the microphysical properties on which they supervene. Meanwhile, the a posteriori microphysicalist denies that in order to defend physicalism we require the a priori thesis, the idea that the facts of the world must be deducible a priori from microphysical facts.

65

Juan Diego Morales Otero

This theory affirms that we can account for the microphysical supervenience thesis even if the a priori approach ends up being wrong, namely, appealing to the necessary identities which are knowable in an a posteriori way -just as water = h2o and heat = molecular kinetic energy. Then the idea is that microphysicalism, as a contingent thesis about our world, claims that all facts are determined by the microphysical facts because the former are either identifiable with these in a transparent manner, that is, they are deducible a priori, or end up being identical to these a posteriori- as in the case of water and heat. In this case, the a posteriori theoretician will assert: even if it is not possible to derive, for example, all the truths about water from the properties of hydrogen and oxygen atoms and their relationships, since we have empirically identified water with h2o, we can say that this is a clearly physical substance. She also will affirm: even if it is impossible to derive, for example, the phenomenal truths about pain from the truths about p1, p2… pn (e.g. certain neuronal activities), since we can empirically identify pain with p1, p2… pn, we can say that this is a clearly physical phenomenon. Let us return to the example of the identification of water with h2o. Both parties to the dispute start from fact that, as Polger tells us, «[n]obody has ever actually provided a complete [deductive] reduction of water to h2o, much less to physics proper» (2008 111). They also know that it is very likely that this reduction cannot be accomplished. The a priori microphysicalist argues that we need at least the possibility of such deduction; and that if it ends up being infeasible, it will be because of purely epistemic factors: lack of exhaustive microphysical knowledge (e.g., about the quantum behavior of the subatomic components of the water molecules), lack of a comprehensive conceptual characterization of the macroscopic phenomenon to explain, or lack of the computational power to produce the deduction. On the other hand, the a posteriori microphysicalist says that we do not need a reduction in terms of possible deduction. She argues that we have the right to be microphysicalists because we can only account for everyday phenomena through their identification with microphysical facts. She says, for example, that it is only through

66

The Concept of the Physical and the Overcoming of the Supervenience Theory

empirical a posteriori identities as those between water and h2o, heat and molecular kinetic energy, pressure and molecular momentum transfer, and boiling and a certain kind of molecular motion, that we can explain everyday phenomena as the fact that heat causes water to boil. As Block and Stalnaker tell us: «Assuming that heat = mke, that pressure = molecular momentum transfer, etc. allows us to explain facts that we could not otherwise explain. Thus, we are justified by the principle of inference to the best explanation in inferring that these identities are true» (Block & Stalnaker 1999 24, emphasis added). My purpose is not to give arguments to resolve the dispute between these two readings of physicalism. Rather, as I said, it is to give some characterizations that allow us to examine their implications. In doing this, I want to suggest that both approaches have plausibly correct intuitions (one about the necessary intelligible connection that all the levels of reality should maintain, and the other, that this connection cannot be articulated in a priori deductive terms), but that they are also wrong about central points (specifically, issues concerning their empirical application, as we will see in the next section). On the one hand, I consider that microphysicalism, if it ends up being empirically correct, should take the aprioristic option, since this would allow us to understand in a transparent form the unity, dependency, and relation of intelligibility that should exist between the truths of the different domains of our world (as well as between the different sciences which attempt to explain these truths). In contrast, it is far from clear to me that a posteriori physicalism can account for this type of unity as opposed to merely affirming it (or assuming it, in the words of Block and Stalnaker).13 As a microphysicalist, the a priori physicalist accepts the identities we can know in an a posteriori way, such as water = h2o. But she adds that for physicalism this is not enough. We must be able to understand, show, and explain why water is transparent, boils under certain conditions, has certain solvent power, etc., from the knowledge of its 13

Kim (2005 146) puts the point in this way: «identities [as such] do not seem capable of generating explanations on their own; the best they can do is “transfer” explanations that have already been completed» in the lower level, e.g. in the level of the molecules of h2o.

67

Juan Diego Morales Otero

microphysical behavior. In this regard, the properties of water are not only identical to the properties of h2o molecules, but the explanation of the former properties is achieved in terms of the latter properties. In fact, it is quite plausible that part of the work of the scientist is to provide us with such explanations.14 Although I agree with the a priori physicalist on the fundamental idea that the different levels of composition and organization of our world must maintain an internal articulation of intelligibility, I also consider that the a posteriori proposal defends a rather important intuition which, as we will see below, is strongly supported by the empirical findings of the different sciences: it is not necessary that, for example, water molecules, heat, cells, and nerve fibers be deducible from their microphysical constituents to consider them physical phenomena, in fact, physical phenomena in a paradigmatic form. More broadly, I consider entirely correct the idea that for physicalism it is not necessary for the facts of our world to be in principle deducible from the microphysical facts. The reason, however, puts pressure on both the a priori and the a posteriori approaches. The Empirical Failure of Microphysicalism

Although the microphysicalist approach (in either of its forms) is the most accepted by philosophers and scientists, it entails very deep empirical and conceptual difficulties. One of the central problems of this approach is that there is strong empirical evidence that seems to show its falsity: there are examples of systems within the scope of physics itself which seem to have holistic properties that do not supervene on the properties of their most basic constituents. This means that from an empirical standpoint, as claimed by the emergentist, there would be at least some physical properties that are not completely determined 14

68

It is for this reason that I tend to agree with Morris’ formulation of the criticism of a posteriori physicalism: «an account of the necessity of a posteriori identities that does not end up implying a priori connections is not much better than a “supernatural” account of how the Necessitation [supervenience] Thesis could be true» (Morris 2011 12). For a similar criticism see, for example, Chalmers 1996 166-7, Kim 2005 chapter 5, and Jackson 2007 196-7.

The Concept of the Physical and the Overcoming of the Supervenience Theory

by their physical constituents; properties which, in this sense, we can call essentially macrophysical, holistic, or emergent.15 In this case, strict microphysicalism would become a falsified theory. One of the most recurrent examples that seem to show the falsity of microphysical supervenience is the phenomenon of the quantum states of entanglement.16 An entangled quantum state is a state of a system of subatomic entities (such as electrons) characterized by having a state vector (or wave function) that is not factorizable in terms of the state vectors of its components; that is, a quantum state which contains additional information, not reducible to its components’ information.17 There is strong empirical evidence that supports the objective existence of this kind of physical phenomenon that violates the microphysical supervenience thesis,18 and which also appears to be a ubiquitous fact of our world: it is not only present in physical systems composed of elementary particles but, as Papineau says, also «present in systems comprising basic physical persisting objects, like atoms and molecules» (2008 146). But it seems that this case is not isolated; it seems that within the scope of physical science, we find numerous examples of holistic properties which do not metaphysically supervene on the properties of their most basic constituents, properties that are not applicable or even make no sense to apply to the most basic levels. It is in this sense that physicists such as Paul Davies argue that emergence is an ubiquitous and rather obvious fact within the field of physics: Chapter 3 is devoted to the articulation of the concept of emergence; for present purposes, the idea of an emergent phenomenon will be understood as equivalent to the idea of a phenomenon that is not metaphysically supervenient on the properties and relations of its constituents. 16 See, for example, Teller 1986, Healy 1991, Maudlin 1998, Esfeld 1999, Silberstein & McGeever 1999, Hüttemann 2005, Papineau 2008, Schaffer 2010, and Bernal 2012. 17 In chapter 5 we will analyze the causal dynamics that should be happening in a case of quantum entanglement, which will allow us to appreciate more clearly the functioning of causation among different levels of organization and composition that must occur in the case that our world contains ontologically emergent phenomena. 18 See, for instance, Aspect, Grangier & Roger 1981. 15

69

Juan Diego Morales Otero

All physicists concede that at each level of complexity new physical qualities, and laws that govern them, emerge. These qualities and laws are either absent at the level below, or are simply meaningless at that level. Thus the concept of wetness makes sense for a droplet of water, but not for a single molecule of h2o. The entrainment of a collection of harmonic oscillators such as in an electrical network makes no sense for a single oscillator. The Pauli exclusion principle severely restricts the behaviour of a collection of electrons, but not of a single electron. Ohm’s law finds no application to just one atom. Such examples are legion. (Davies 2006 36)

This is the thought of two of the important scientific voices who reject the widespread microphysicalist assumption, A. J. Leggett, Nobel laureate in Physics, and Roald Hoffmann, Nobel laureate in Chemistry:19 [T]here are also a number of cases, particularly in the traditional areas of the physics of gases and crystalline solids, in which a model which treats the behavior of the whole as essentially just the sum of that of its parts (atoms or electrons) has been quite successful. […] But these cases, despite the fact that they totally dominate the presentation of the subject in most elementary textbooks, are actually the exception rather than the rule. In virtually all the frontier areas of modern condensed-matter physics, the relationship between our understanding of the behavior of matter at the microscopic level of single atoms and electrons, and at the macroscopic level of (say) liquids and solids, is actually a good deal more complicated than this. […] I would claim that the most important advances in this area come about by the emergence of qualitatively new concepts at the intermediate or macroscopic levels -concepts which, one hopes, will be compatible with one’s information about the microscopic constituents, but which are in no sense logically dependent on it. (Leggett 1987 113-5)

19

70

In this non-reductive physicalist direction, see also the work of the Nobel laureates in Physics, P. W. Anderson (1972) and Murray Gell-Mann (1994, especially pp. 111-3).

The Concept of the Physical and the Overcoming of the Supervenience Theory

Meanwhile, on the relationship between physics and chemistry, Hoffmann claims: My stronger claim is that most concepts in chemistry that have proven productive in the design of new experiments are not reducible to physics. By this I mean that they cannot be redefined adequately using only the language of physics. To be specific, I’m thinking of aromaticity, the acid-base concept, the idea of a functional group, a substituent effect, and that of a chromophore. And let us not forget the chemical bond. (Hoffmann 2007 329)

This leads us to consider that from a scientific point of view, it is highly plausible that within the physical and chemical scopes of matter we come to find phenomena that cannot be understood under the criteria of microphysical supervenience.20 But there is also evidence for the idea that the failure of metaphysical supervenience goes beyond the physical and chemical scopes. For instance, there is an empirically based growing agreement on the idea that biological properties cannot be properly and completely understood and explained on the basis of their underlying chemical processes.21In this sense, and focusing on the particular analysis of the behavioral genetics of the nematode Caenorhabditis elegans, a simple worm, Schaffner concludes that «there is no simple [reductive] explanatory model for behavior even in simple organisms. What C. elegans presents us with is a tangled network of influences at genetic, biochemical, intracellular, neuronal, muscle cell, and environmental levels.» (1998 237) And, as we could expect, the empirical results available regarding the interaction between mental and neural properties support a failure of reductive explanation that, therefore, at least suggests a failure of metaphysical supervenience of mental properties on their neuronal bases.22 On the empirically plausible failure of microphysical supervenience within the (non-quantum) physical and chemical scopes, see also, for example, Primas 1983 and 1998, Cartwright 1997, Sklar 1999, Hendry 2006, and Kistler 2006b. 21 See, for example, Campbell 1974, Rothschild 2006, Wimsatt 2007, Dupré 2012, and Davies 2012. 22 See, for example, Sperry 1975, Van Gulick 1993, Hardcastle 2001, Velmans 2002, Thompson 2007, Scott 2007, Brown 2007, Newsome 2009, Juarrero 2009, and Frith 2009. 20

71

Juan Diego Morales Otero

If at least some of these phenomena end up being emergent, in the sense of having systemic dynamics not completely determined (underdetermined23) by the properties and relationships of the elements that constitute them, then emergence could become an everyday and practically ubiquitous fact of our world. It is on this explanatory background that Wimsatt, one of the philosophers of science who has defended an anti-reductive physicalist approach over several decades, states that «[e]mergence in this sense is thus extremely common – much more so than normally supposed» (Wimsatt 2007 175). Consequences for Determining the Concept of the Physical

Beyond the philosophical and scientific implications of these empirical results with respect to the relationship of the different ontological and explanatory levels -which we will examine in subsequent chapters-, these involve a direct and crucial result for our current research: they lead us to the conclusion that it is not the relation of microphysical supervenience which plays the role of criterion for considering the holistic physical systems, much less the chemical and biological systems, as physical phenomena in their own right. In this sense, if the microphysicalist theory were the only solution to our problem, we would have no way of asserting, as we obviously do, that the physical holistic levels, along with the chemical and biological, count as paradigmatic physical realms. With regard to the philosophical implications of the empirical failure of the microphysical theories, authors like Barbara Montero and David Papineau draw similar conclusions to those outlined here. Montero tells us: [I]magine that duplicating fundamental physics could give us a world with, say, just quarks, leptons, their antiparticles, and such like, but no chemical bonds, no molecules, no cells, no organisms. If our world were like this, chemical properties, among others, would fail to supervene on fundamental physics. Must this be a world in which physicalism is 23

72

In chapter 5, which deals with the nature of higher or emergent causation, we will analyze in detail the operation of this underdetermination.

The Concept of the Physical and the Overcoming of the Supervenience Theory

false? Or to narrow the question down, must we understand the chemical bonds (and the properties thereof) in this world as nonphysical? There seems to be no reason to think that we must. Yet if this is correct, physicalism can be true even if chemical, biological, psychological, and other such properties fail to supervene on the entities, properties, laws, and relations of physics. (Montero 2013 101)

In the same sense Papineau affirms: The properties of atoms and molecules may be irredeemably holist because of quantum non-locality, but these properties are surely still physical. We don’t want to say that the total spin of a number of electrons doesn’t count as ‘physical’ just because it isn’t determined by the local properties of the individual electrons. (Papineau 2008 146)

Before advancing any characterization of the physicalist approach that helps us respond to the deep problems we found in the microphysicalist doctrine, I want to insist on the kind of conceptual problems that the philosopher who tries to understand the physicalist approach based on a supervenience theory should face. We will see that the problem of microphysicalism is based not merely on its assumption of a priority of the microscopic, but on its more general acceptance that we can come to understand the physicalist position through a theory of supervenience. The conclusion will be that the theory of microphysical supervenience is neither necessary, since we have a concept of the physical regardless of its truth or falsity, nor sufficient, since assuming its truth we could derive unacceptable conceptual consequences. Many contemporary philosophers have relied so much on the criterion of microphysical supervenience that they have assumed the conceptual and metaphysical possibility that mental phenomena, as well as shopping centers and wars (Jackson 2003 85), botanical and automotive properties (Shoemaker 2010 125-6), biological properties of reproduction and evolution (Chalmers 1996 39), and even entities such as water and heat (Block &Stalnaker 1999 18, 29), can be instantiated in and realized by ‘immaterial mechanisms.’ But I think that this idea is simply incoherent.

73

Juan Diego Morales Otero

These authors come to accept what I consider an incoherence, because they assume the combination of the following two ideas: (i) that ‘physical’ really means ‘microphysical,’ and (ii) that a physical world is that which is completely determined, that is, which metaphysically supervenes on the microphysical facts. Since shopping centers, wars, automotive and botanical properties, the biological properties of reproduction and evolution, water, heat, cells, neurons, etc., are not microphysical phenomena, then, under their criterion, these are not physical phenomena in their own right, but only in a derivative way, only if they come to supervene on microphysical facts. Two important consequences follow: first, that given that the macroscopic properties are not essentially physical, they could be instantiated in non-material mechanisms; and, second, that the contingency of physicalism should be understood as the possibility that macroscopic features can be realized by immaterial mechanisms. As I pointed out, contemporary physicalists make a mistake when they assume that the contingency of their thesis should depend on the possibility of a Cartesian mind which can be instantiated in immaterial mechanisms, i.e., in structures that are neither physical nor spatial. Moreover, I think the idea of ​​an immaterial mind does not make sense, but much less the idea that reproduction, evolution, dogs, trees, rocks, heat, and water can be instantiated in non-physical stuff! That is, that there can be, in the words of Block and Stalnaker (1999 18, 28), ​​ ghost heat and water, immaterial heat and water. The mistake of these authors derives from the fact that their two main assumptions -(i) that ‘physical’ really means ‘microphysical,’ and (ii) that a physical world is one which metaphysically supervenes on the microphysical- are deeply problematic, as the latter is plausibly false and the former, conceptually incorrect. On the one hand, we have seen that current physics, not to mention other sciences, considers that there is strong evidence of holistic physical phenomena which do not supervene on their microphysical constituents. On the other hand, plausibly emergent or holistic phenomena, such as entangled quantum systems, the Pauli exclusion principle, and relativistic gravitational fields, among others, are considered clearly physical entities by the branches of the physical sciences that study them, which makes

74

The Concept of the Physical and the Overcoming of the Supervenience Theory

it clear that physical cannot mean microphysical. Precisely because physical does not mean microphysical, we can say that water, rocks, heat, cells, and nerve fibers can be, and indeed, are considered paradigmatic cases of physical entities. Finding a Solution to our Problem

Why then do these types of entities which are not instantiated in the basic microphysical level, and which may not supervene on the elements of such a minimal tier, qualify necessarily as physical? We can begin to articulate the answer to our question from the fact that these entities possess certain traits which are traditionally held as necessary criteria for physicality, namely: (1) they are ontologically objective phenomena, that is, they exist and have some properties regardless of the existence of other entities and phenomena,24and, following Papineau’s above-mentioned suggestion, (2) they display general mathematically expressible behavior. In this sense, Carnap claims: By ‘physics’ we wish to mean, not the system of currently known physical laws, but rather the science characterized by a mode of concept formation which traces every concept back to state-coordinates, that is, to systematic assignments of numbers to space-time points. Understanding ‘physics’ in this way, we can rephrase our thesis -a particular thesis of physicalism- as follows: psychology is a branch of physics. (Carnap 1959 197)

Beyond this, I think it is beneficial to maintain Descartes’ almost tautological assertion, namely, that material entities are those that must be instantiated in a spatiotemporal structure. In this sense, our criterion (3) is that the physical implies spatiotemporality. Now, we can say that although General Relativity asserts that every physical entity in time should also be in space, this is a nomological principle which could be falsified in possible worlds very different to ours. In this sense, it does not give us the conceptual strength to say that the 24

Save for the entities that constitute them, such as water depends on hydrogen and oxygen atoms, and some peculiar relations between them, but not on the existence of gold.

75

Juan Diego Morales Otero

physical implies the spatial. Nonetheless, I think that the following argument that relates the physical with the objective does. Following a Kantian thesis, some authors argue that the notion of spatiality ​​ is a necessary component of the concept of objectivity. The idea is that a phenomenon can only be understood as objective if it can be instantiated in other phenomena without losing its independence. And it seems that the possibility of such instantiation makes sense only within a spatial structure, an extended structure which can situate various objects and events in different locations, even simultaneously.25 For example, if we say that entities a and b have independent existence, we should say that a can change its properties while b remains identical. But this only makes sense within a spatiotemporal structure, a structure that makes possible the occurrence of simultaneous different events. Following this idea, we should affirm that rocks, water, trees, cells, and nerve fibers, insofar as they are objective phenomena, can only be understood as phenomena that must be instantiated in a spatiotemporal structure. We also have to say that if we consider the mind and its attributes to be objective phenomena, existing independently of the existence of other entities with which they can be contingently related (whether other minds or physical objects), we should claim -against Cartesianism- that they must be necessarily instantiated in a spatiotemporal structure.26 In addition to these three necessary but plausibly not sufficient features for something to count as a physical entity, we must add a condition applicable to the most basic levels of matter. Against Lewis (1983 363), I think it is natural to agree with most theorists about the conceptual and metaphysical incompatibility of a physicalist and a panpsychist world.27 I believe the fundamental reason for adopting this view is that physicalism is empirically based on a physical science that should be a descendant of the current physics, which clearly denies that See, for example, Strawson 1959 and Evans 1985. I think this is the idea that underlies Kim’s criticism of the very possibility of the Cartesian mind’s causal interactions. See Kim 2005 chapter 3 and 2011 46-56. 27 See, for example, Wilson 2005 428, Montero 2006 182, and Stoljar 2010 88. 25

26

76

The Concept of the Physical and the Overcoming of the Supervenience Theory

the basic levels of matter are constituted by consciousness or purpose. If this were not the case, physicalism would end up being compatible with any theory that could finally reach a full empirical explanation of the world (e.g., a double-aspect theory, a theory of pre-established harmony, occasionalism, and property or substance dualism), which is not the case. Following these ideas, (4) an entity that qualifies as physical cannot be both microscopic and possess consciousness or mentality. To put it in other words: any entity that is both physical and mental (as is commonly assumed to be the case of many brain states) can only be a macrophysical entity, which metaphysically depends on its microphysical non-mental constituents. There is also a fifth fundamental feature that allows us to understand the unity, the ontological and explanatory dependence, and the intelligibility relation that should exist among the different levels of physical composition and organization: (5) all physical entities, except those microphysical entities that make up the smallest level of reality -if this level exists28- are, and cannot be more than a supervenient or emergent organization of the physical elements that compose them. This idea follows from the fact that the physical domain is essentially an extensional and mereological structure that must maintain an internal order and organization. In this sense, macrophysical entities only exist by virtue of, thanks to, and, therefore, metaphysically depending on the elements from which they emerge or on which they supervene. And a macrophysical entity can only be understood as the organization of the elements on which it either metaphysically supervenes (as when the organization is completely determined by its elements), or emerges (as when the organization is underdetermined29 by these constituents).

Some theorists (for example Crane & Mellor 1990, Block 2003, and Earley 2008) affirm that from an empirical point of view, the existence of a level of non-divisible entities is highly implausible. See also the claim of the Physics Nobel laureate Hans Dehmelt on this point (1990 539), and see also Montero’s (2006) discussion about the problem and possibility of physicalism in a world without such a level. 29 About this underdetermination, see chapter 5. 28

77

Juan Diego Morales Otero

Following this idea, physicalism should affirm that entities such as atoms would be the kind of objects that are made ​​up of electrons, protons, and (save hydrogen) neutrons organized in an especially particular way. In other words, an atom would be nothing more than some specific organization of some lower level physical elements, electrons, neutrons, and protons. A molecule would be no more than a specific organization of some lower level physical elements, atoms, and so forth, with respect to entities ranging from the microscopic to the highest levels: from superstrings to biological, mental, and social entities. To summarize: at this point we can meet Hellman’s demand, cited at the beginning of chapter, about ​​ the «lack [of] any general criterion of ‘physical object, property, or law’ framed independently of physical theory.» We can say that an entity can only be understood as physical if it meets at least the five conditions listed above (regardless of whether it supervenes or not on microphysical elements): (1) it is ontologically objective, (2) it displays at least general mathematically expressible behavior, (3) it can only be instantiated in a spatiotemporal structure, (4) it is not both microscopic and mental, and (5) in case it is macroscopic, it is a (supervenient or emergent) organization of the physical elements that constitute it. As it is clear that holistic physical, chemical, biological, and neurophysiological properties meet these requirements, it is clear that they are physical entities in their own right. Following these criteria, we can then say that physicalism should be understood as the contingent and a posteriori theory which states that our empirical world is completely constituted by entities that meet at least the five listed conditions. It follows that if our world ends up being physicalist, then it will be a world that is essentially objective, necessarily spatiotemporal, and, in principle, explainable by the mathematical methodology of the natural sciences, a world some of whose fundamental dynamics can be essentially macrophysical (that is, physical phenomena which cannot be reduced to or understood purely in terms of the properties and relations of their microphysical components). The important consequences of an understanding of the

78

The Concept of the Physical and the Overcoming of the Supervenience Theory

physical and the physicalist theory along these lines will be worked out inthe next chapter. Summary

In this chapter, I have focused on the clarification of what should count as the criteria for an entity to qualify as physical. The clarification has a direct impact on the way we understand the physicalist theory, which argues that the facts of our empirical world are essentially physical. We have seen that the most accepted formulation of physicalism, the theory of the supervenience or complete determination of empirical phenomena by microphysical characteristics is insufficient, both empirically and conceptually. On the one hand, it seems to contradict the results of physical science, and, on the other hand, it allows us to understand neither the scientific nor the daily use of the notion of physical entity. To overcome these problems, I have argued that following certain traditional views we can come to understand in what sense an entity counts as physical. I conclude with the idea that if our world ends up being physicalist, then it will be a world that is essentially objective, necessarily spatiotemporal, and, in principle, explainable by the mathematical methodology of the natural sciences, a world some of whose fundamental dynamics can be essentially macrophysical.

79

CHAPTER 3 Emergentism as Type Macrophysicalism Thus, for emergent evolution, conscious events at level c (mind) involve specific physiological events at level b (life), and these involve specific physico-chemical events at level A (matter). No c without b, and no b without a. No mind without life; and no life without «a physical basis.» morgan 1923 15

As I have said, the idea of emergence has reappeared throughout the history of ideas, and it is probable that we can find its greatest conceptual formulation in the work of the so-called classic British Emergentists, thinkers like John Stuart Mill, Henry Lewes, Samuel Alexander, C. Lloyd Morgan, and C. D. Broad. These authors wanted to overcome both dualism and microphysical reductionism (called ‘mechanism’ by British Emergentists) in a subtle and synthetic way. Against the dualist perspective, emergentism argues that the world must be understood as a unitary, material world, solely composed of material atoms or more fundamental particles in motion. And, against the reductionist, the emergentist claims that the material world consists of a series of layers that complexly organize its purely material elements. The purpose of this chapter is to develop a careful articulation of the concept of ontological emergence, which subsequently will allow us to assess the extent of its contribution to the solution of central problems, such as the feasibility of a non-reductive physicalism and the problem of causation in the special sciences. In the first section, I develop a general characterization of ontological emergence whereby

81

Juan Diego Morales Otero

this phenomenon should be understood as a special organization or relational structure that the constituents of a system can acquire, and which introduces a causal and dynamical difference that is not completely determined by the causal factors of these constituents. The second section is devoted to the examination of the relation between the emergentist thesis and the ontological approaches of reductive physicalism, non-reductive physicalism, token physicalism, and property dualism, and to show that, despite the numerous readings developed in recent years, emergentism should be treated as a clear kind of non-reductive physicalism. Finally, in the third section, I distinguish two general types of ontologically emergent entities, which will allow us to realize and comprehend the internal diversity of the phenomenon and to refine the boundaries of the concept, in order to understand its philosophical and scientific consequences. The Idea of Ontological Emergence

In this section, I analyze the relation between the notions of ontological and epistemological emergence, and I develop a general characterization of the first in terms of fundamentality, dependence, and in principle or theoretic deducibility. In the first part, I will review Elizabeth Barnes’ very interesting and fruitful characterization of this concept. Then, I will examine the way in which C. D. Broad gets rid of the epistemological reading of his position, in order to understand it as completely ontological and objective. Finally, I make some comments about the very sense in which we must understand the ontological difference or addition that the emergent involves with respect to the basis from which it emerges. As is well known, not few philosophers introduce a distinction between a kind of epistemological emergence, which Bedau calls weak emergence, and a kind of strong, ontological, or metaphysical emergence.1 Epistemological emergence emphasizes a kind of de facto cognitive or epistemic limitation that we, as cognizers, have in order to account for, explain, or understand (expressed in our inability to 1

82

See, for example, Bedau 1997, Silberstein &McGeever 1999, Chalmers 2006, and VanGulick 2007.

Emergentism as Type Macrophysicalism

predict, calculate, compute, or deduce) certain features of holistic phenomena on the basis of the knowledge of their basal lower level conditions, i.e., the pure information of the properties, relations, and laws of their parts. Meanwhile, on the ontological conception, emergence is understood as an objective, independent of mind phenomenon, consisting in the ontological fact that some of the properties of the physical systems in our world are different from, and so neither metaphysically determined by nor reducible to, the properties and relations of their constituent elements. This ontological difference is typically spelled out in terms of new, different, or additional causal powers which are introduced with the appearance of the higher level, emergent phenomena.2 Most theorists seem to have considered that the concept of ontological emergence has a relative unity that creates a more or less clear and stable boundary vis-à-vis the domain of the merely resultant (nonemergent) phenomena. Nonetheless, despite the constant discussions on where to draw the precise limit between the ontologically emergent and the merely resultant, most theorists have resorted to an epistemological criterion of non-deducibility and non-predictability to this end.3 Characterizing the emergentist theory, for example, C. D. Broad claims: Put in abstract terms the emergent theory asserts that there are certain wholes, composed (say) of constituents a, b, and c in a relation r to each other; that all wholes composed of constituents of the same kind as a, b, and c in relations of the same kind as r have certain characteristic properties; that a, b, and c are capable of occurring in other kinds of complex where the relation is not of the same kind as r; and that the characteristic properties of the whole r(a, b, c) cannot, even in theory, be deduced from the most complete knowledge of the properties of a, b, and c in isolation or in other wholes which are not

This is true about the British Emergentist proposals, as well as of the contemporary formulations of ontological emergence. See, for example, Klee 1984, McLaughlin 1992, Silberstein & McGeever 1999, O’Connor & Wong 2005, and Kim 2006a, 2009b. 3 See, for example, Mill 1843, Lewes 1875, Morgan 1923, Chalmers 1996, and Scott 2007. 2

83

Juan Diego Morales Otero

of the form r(a, b, c). The mechanistic theory rejects the last clause of this assertion. (Broad 1925 61)

Now, although most of the authors use an epistemological criterion of non-deducibility or non-predictability to characterize a concept of ontological emergence, it is clear that they intend to examine its objective and ontological kind. On this point, I agree with most interpreters of classical emergentism, particularly with McLaughlin when he clarifies: Predictability is, of course, an epistemological notion, while causal determinism is an ontological one. These authors [the British Emergentists] are not always as careful as they might have been in distinguishing epistemological from ontological theses. But charitable commentators have rightly taken them to be stating the ontological thesis of causal determinism. I should note in this connection that Emergentists often speak of emergent properties and laws as unpredictable from what they emerge from. But, contra what some commentators have thought, the Emergentists do not maintain that something is an emergent because it is unpredictable. Rather, they maintain that something can be unpredictable because it is an emergent. Emergence implies a kind of unpredictability. But it is a mistake to conflate emergence with this consequence of emergence. The British Emergentists do not. (1992 note 31)

The problem is that this purely epistemological characterization of the concept of ontological emergence is, at least in principle and without sufficient clarification, manifestly inadequate. Thus, it allows us neither to understand nor evaluate the specific logical boundaries of the concept, which are necessary for its articulation within central issues such as the feasibility of a non-reductive physicalism and the problem of causation in the special sciences, problems that are of central concern in this monograph. It is precisely for this reason that we must make a careful articulation of the concept of ontological emergence which does not resort to epistemological interpretations, or which, if it does, clearly shows the ontological meaning that should be inferred from such interpretations.

84

Emergentism as Type Macrophysicalism

The idea of ​​ontological emergence has been often considered suspect and enigmatic (Bedau 1997), somewhat mysterious (Klee 1984), and even paradoxical, inconsistent, and incompatible with a physicalist approach (Kim 1992a, 2006a). Barnes claims that «much of the mystery about emergence arises from having no clear statement of what its meta-ontological background should be» (2012 874). That once we reach this kind of clarity we will be able to see that emergence, «whether or not you think there is any reason for endorsing it, is not on this interpretation mysterious» (2012 874). Barnes’ idea is that the basic and essential characterization of the phenomenon of ontological emergence should be articulated in terms of the notions of fundamentality and dependence and, accordingly, that we should understand «emergent entities as those which are fundamental but not independent» (2012 875). The fundamentalist ontology affirms a central ontological distinction between what exists fundamentally and what exists derivatively. We often think that some worldly phenomena are more fundamental than others, as when we say that electrons are more fundamental than molecules, molecules more fundamental than cells, and cells more fundamental than organisms and cognitive agents. That is, under traditional considerations, fundamentality is a matter of degree. But Barnes claims that «on the fundamentalist view things are either fundamental or they are not, in which case they are derivative. Fundamentality does not come in degrees» (Barnes 2012 876). Now, the central idea is that fundamental entities are those that are completely necessary for the existence of some phenomenon or specific ontological sphere. Using the familiar theological metaphor to understand when an entity belongs or not to this category, we can say with Barnes that «the fundamental entities are all and only those entities which God needs to create in order to make the world how it is. [… Whereas] derivative existents are those which God does not need to include in her “ontological shopping basket”» (2012 876-7). For example, if the microphysicalist were correct, the only entities that would be required to and, therefore, would fundamentally exist in our world, would be the microphysical entities. Whereas other phenomena, such as chairs, people, stars, neurons, minds, societies, etc.,

85

Juan Diego Morales Otero

would be metaphysically derived from these. To put it another way: once microphysical entities are fixed, then, in every possible world, all these other entities will be fixed. In the jargon that is now traditional, all entities of our world would be nothing over and above the microphysical entities. Nonetheless, if holistic phenomena as entangled quantum systems end up being real, objective, and non-eliminable phenomena of our world, these entities will have to be fundamental, «entities which God needs to create in order to make the world how it is.» Although our world would be over and above microphysical entities, it is clear that it would be nothing over and above physical phenomena. But there is another basic ontological distinction that is independent of and orthogonal to the fundamental/derivative distinction: that between entities which are ontologically dependent and those which are ontologically independent. As Barnes says, «[t]his potential separation between fundamentality and independence will be what gives rise to the basic characterization of emergence that follows» (2012 879). Let us see the definition of ontological dependence that the author develops: An entity x is dependent iff for all possible worlds w and times t at which a duplicate of x exists, that duplicate is accompanied by other concrete, contingent objects in w at t. That is, in any situation in which there is something exactly like x, you’ve got to have other things existing alongside it. You cannot ever just have x by itself. And in that sense, x depends (at every moment of its existence) on other things. (2012 880)

Clear examples of dependent entities are most concrete objects and phenomena of our world, as they are mereological compounds ontologically dependent on their physical constituent parts. Complementarily, the paradigmatic example of an ontologically independent entity will be a completely simple element, an atom in the strict sense of the word, the class of entity to which the microphysicalist thesis is committed. At this point, then, we can understand the central idea of Barnes that the existence of ontological emergence is equivalent to the fact that some entities which are fundamental are not ontologically

86

Emergentism as Type Macrophysicalism

independent. More formally, «An entity x is ontologically emergent if x is fundamental and dependent» (2012 884). Following this characterization of emergence, we can say that if the empirical evidence that strongly supports the existence of holistic physical phenomena as quantum entangled systems, relativistic gravitational fields, and the entrainment of collections of harmonic oscillators is finally correct, then we have to say that these phenomena should be understood as ontologically emergent. This is so because they would be fundamental but dependent: fundamental, in the sense that they would not be metaphysically fixed by, or supervenient on, their microphysical constituents, and so they would be phenomena which «God needs to create.» In dynamic terms, as we will carefully examine in chapter 5, they would be causal essential ‘additions’ for the development of the history of our world. And they would be dependent in the sense that they would always need their microphysical bases, their microphysical constituents, in order to exist. To put it in other terms, we can say that although the microphysical facts will always (in every possible world) be necessary for the ontological emergence of holistic physical properties and entities, they will be insufficient; fundamentally macro or holistic properties and laws are needed. Let us now turn to Broad’s proposal. I have said that most emergentists, especially British Emergentists, try to give an ontological and objective characterization of emergence, but that in this attempt they only find epistemological criteria to separate the emergent from the merely resultant. But these theorists, particularly Broad, have been well aware of this situation. It is in this context and for this reason that Broad introduces the idea of ​​a mathematical archangel: If the emergent theory of chemical compounds be true, a mathematical archangel, gifted with the further power of perceiving the microscopic structure of atoms as easily as we can perceive hay-stacks, could no more predict the behaviour of silver or of chlorine or the properties of silver-chloride without having observed samples of those substances than we can at present. And he could no more deduce the rest of the properties of a chemical element or compound from a selection of its properties than we can. (Broad 1925 71)

87

Juan Diego Morales Otero

The idea, then, is that an emergent property is one that cannot be known or predicted, because it cannot be deduced either in principle or in theory from the most complete knowledge of the microphysical elements from which it emerges. It is a property that cannot be deduced by a mathematical archangel who can have a perfect knowledge of the microphysical conditions, and who is logically and mathematically omniscient. And it cannot be deduced, either in principle or in theory, from the most complete knowledge of the microphysical elements just because it is not there, because it is something more than these conditions, that is, an emergent. As Kim comments on this passage, If the archangel cannot produce a proof, it’s because there is no proof and the proposition is not deducible, in an absolute sense, from the premises. Thus, if a property is emergent from a set of basal conditions, there is no deduction of it from those conditions, and this has nothing to do with the epistemic powers of any cognizers. This idealization of deduction is Broad’s attempt to purge any epistemic and relativistic aspects from the notion of deduction and thereby objectify deducibility, or nondeducibility. For him, there being no deduction of an emergent property from its basal conditions is not an epistemological fact. It is not because we are not smart enough, or don’t have enough time or inclination […] This surely seems like an objective fact about the relationships between sets of truths, or facts. (Kim 2009b 95-6)

From these ideas we can see that the emergentist and the a priori microphysicalist are following the same rules, agreeing about the conceptual criteria that should be applied in order to examine and decide whether our world is a microphysicalist world or whether it contains holistic and emergent phenomena that in some sense are more than, or different from, the microphysical conditions from which they emerge. If the microphysicalist is correct, and thus the properties of our world are at least in principle or in theory deducible in an a priori way from a complete knowledge of the basic microphysical entities, then we have to say that ontological emergence is not an actual constituent part of our world. Meanwhile, if the emergentist is correct, such an ideal deduction will not be possible.

88

Emergentism as Type Macrophysicalism

Following these ideas, I think we can agree with Broad, Kim, Chalmers, and others that this type of in principle deducibility/nondeducibility can be understood as a criterion of the existence and nonexistence of ontological emergence because is designed to clearly and directly reflect the objective and ontological domain. We could say: if there is no such ideal deducibility (even better, if we find empirical or conceptual reasons to reject the possibility of such deduction), that is because the higher level property does not supervene on, and therefore, is not metaphysically determined by, the microphysical properties from which it emerges. The higher level or macroscopic property would be something additional to the microphysical conditions from which it emerges. In the characterization of Barnes: although that property would be dependent on its microphysical conditions, it would also be fundamental, neither deducible nor reducible to such conditions. Even though it is perhaps obvious from what has been said here, we should bear in mind, in addition to the notion of in principle deducibility, the corresponding idea of in principle or theoretical predictability. To clarify this notion, we have to distinguish between inductive predictability and theoretical predictability. The first refers to the idea that we can predict a phenomenon because we have observed the conditions under which it usually appears, as when we predict that it will rain based on certain apparent features of the clouds, and because we have had an experience of the correlation of these facts. In this sense, even an emergent property could be predicted on the basis of the microphysical conditions on which it appears: if we have continually observed that an alleged emergent property e (say, a quantum state of entanglement) appears whenever certain microphysical conditions m are instantiated (e.g., particular electron states), then we may predict that e will be instantiated whenever m occur. Or, put more generally, according to Kim, «on the basis of such empirical data we may have a well-confirmed “emergence law” to the effect that whenever a system instantiates basal condition m it instantiates an emergent, e.» (1999 13-4) But if e is really an emergent property, it cannot be the subject of a theoretical prediction. Suppose that we can have full theoretical knowledge about the microphysical conditions m, and that we can

89

Juan Diego Morales Otero

predict that m will be instantiated at time t. If e is really an emergent property, and we don’t have knowledge about the empirical correlation between m and e (say, the fact that whenever electrons possess particular properties and maintain certain relations an emergent global property appears), then we could not predict the appearance of e at t. That is, we could not have a theoretical prediction of e on the basis of knowledge about the future instantiation of m.4 We have seen that the idea of in principle non-deducibility should be considered as a criterion for ontological emergence. But here we find an important issue. Emergentism is committed to the idea that we cannot in principle deduce emergent properties of a whole from its microphysical conditions. But a passage from Kim reminds us that neither can we deduce, even in principle or in theory, for example, facts about the Nile River from facts about Jupiter, or geological truths from the truths of macroeconomics, or facts about the surface composition of the moon from facts about neurotransmitters in the human brain; not even the logically all-powerful archangel can do that, and that is because there are no deductions between these sets of truths. (Kim 2009b 96)

But in this case the reason is simple; it is because these different pairs of facts are really and completely distinct. But the case of emergence is the opposite: the emergent whole is constitutively dependent on, and so necessarily connected with, its lower components. The criterion of in principle non-deducibility allows us to affirm that the emergence relationship is maintained between phenomena (sets of properties or facts) that somehow are objectively different. We obtain the same idea from Barnes’ assertion that ontologically emergent entities are those which are fundamental vis-à-vis their lower bases, completely necessary for the existence of some phenomenon or specific ontological sphere. But considering that the relation of emergence is a type of mereological relation, and therefore, as Barnes argues, that the emergent entity necessarily depends on the instantiation of its lower-level basis, we need to raise the question about the specific 4

90

See Broad 1925 70, and Kim 1999 13-4.

Emergentism as Type Macrophysicalism

sense in which the emergent entity goes beyond, is different from, or additional to the microphysical entities from which it emerges. I think that the question about the sense of the difference b ­ etween the emergent and its basis is a crucial point that has often been overlooked and has tended to produce frequent conceptual tangles. It is clear that the concept of emergence entails the idea of something new and different. However, it is important to understand that such a novelty should be understood in a precise way. When we say that an emergent entity is new and appears from certain elements, we do not say that what appears is a different entity of the same type as the elements from which it emerges. Claiming this is to commit a category mistake. Recalling Ryle’s famous example, we can say that claiming such an idea is just like stating that the University of Oxford is an element or entity of the same type of its libraries, playing fields, museums, scientific departments and administrative offices. But this is an obvious conceptual mistake. Rather, what emerges is a certain organization of its constituents.5 Then, the idea that the emergent entity is different from or additional to the microphysical entities from which it emerges, is tantamount to the claim that these elements acquire a special organization or relational structure that is neither metaphysically determined by its constituents’ interactions, nor already in the other wholes that these constituents can compose and, so, introduces a causal and dynamical difference which is not completely determined by the causal and dynamical factors of these constituents.6 Emergentism as Type Macrophysicalism

In chapter 2, we saw that the concept of the physical is logically independent of the microphysicalist proposal, the theory which claims that all the properties of our world supervene on -and so, are metaphysically fixed by- its microphysical properties. The conceptual problem of microphysicalism is that it does not allow us to understand the criteria by which we affirm that holistic physical, 5 6

See the fifth condition of physicality proposed in chapter 2. As I have said, a causal and dynamical difference that is underdetermined by the causal and dynamical factors of its lower elements. See chapter 5.

91

Juan Diego Morales Otero

chemical, biological, and neurophysiological properties are physical in their own right. In response, I argued that an entity can only be understood as physical if it meets at least the following conditions: (1) it is ontologically objective, (2) it displays at least general mathematically expressible behavior, (3) it can only be instantiated in a spatiotemporal structure, (4) it is not both microscopic and mental, and (5) in case it is macroscopic, it is a (supervenient or emergent) organization of the physical elements that constitute it. As it is clear that holistic physical, chemical, biological, and neurophysiological properties meet these requirements, it is clear that they are physical entities in their own right. One of the important consequences of this perspective is that a physical world can be a world some of whose fundamental phenomena are essentially macrophysical, that is, physical phenomena (insofar as they meet the previous conditions) which cannot be reduced or understood purely in terms of the properties and relations of their microphysical components. But on the basis of the formulation of ontological emergence we have developed so far, this is simply to say that this physical world would be an emergent world, a physical world some of whose fundamental physical phenomena are essentially emergent. In terms of properties or kinds, this would be a physical world partially constituted by fundamentally macro or emergent physical properties or kinds. And this is the sense in which emergentism is understood as type macrophysicalism. Just as the traditional identity theory has claimed, emergentism affirms that all the properties of our world are physical; but, unlike this theory, it argues that some of these essentially physical properties are only instantiated in macrophysical systems and cannot be reduced to the microphysical properties from which they emerge. And it is for this reason that emergentism or type macrophysicalism is a kind of non-reductive physicalism. In spite of this, some theorists in recent years have developed an understanding of emergentism whereby this theory cannot be considered a form of physicalism. On the basis of the idea that emergentism denies the metaphysical supervenience of the higher levels on their lower physical bases, these authors believe that this theory should be interpreted as a kind of dualism, specifically, a property

92

Emergentism as Type Macrophysicalism

dualism, which states that the relation among the different levels of organization of our world is merely contingent. The purpose of this section is to analyze the relation between the emergentist thesis and the ontological approaches of reductive physicalism, non-reductive physicalism, token physicalism, and property dualism, and to show that, despite the numerous readings developed in recent years, emergentism should be treated as a clear kind of physicalism, specifically, as the non-reductive physicalism that has been characterized here. The basic line of the argument consists in showing the logical independence of the concepts of metaphysical determination and metaphysical dependence. It is true that both emergentism and property dualism deny the metaphysical supervenience (and therefore the metaphysical fixation or determination) between the microphysical level and the higher levels of our world. However, there exists a crucial difference between these perspectives: emergentism affirms, while property dualism denies a metaphysical dependency connection between the higher levels and the microphysical level. In fact, this emergentist statement is derived from Barnes’ characterization of emergence in terms of dependent fundamentality. Here, however, I explain the reason why this must be so. In the first part of the section, I discuss the relations and differences of the two approaches that affirm a metaphysical determination between the microphysical and the higher levels, that is, between reductive and non-reductive microphysicalism. In the second part, I will clarify the sense in which property dualism denies both metaphysical determination and metaphysical dependence between higher and microphysical properties. Finally, I will show why emergentism is committed to the idea of a metaphysical dependence between higher and microphysical properties and, in consequence, it should be treated as a clear form of physicalism. Reductive and Non-Reductive Microphysicalism

We have seen that most contemporary philosophers have understood the physicalist perspective in terms of the supervenience theory, according to which the properties of our world supervene on and are, therefore, metaphysically determined by its microphysical facts. I said that the notion of supervenience has been introduced and

93

Juan Diego Morales Otero

developed with the primary aim of accounting for a naturalist and physicalist non-reductive proposal, which intends to support both a priority of the natural and physical phenomena of our world and the irreducibility and difference of properties and phenomena that in principle cannot be understood as physical, such as mental, moral, political, and economic properties and phenomena. For example, on the basis of the idea that there cannot be «strict» psychophysical laws, Davidson articulates his non-reductive physicalist proposal, which he calls anomalous monism, claiming that the mental properties supervene on the physical properties even though they cannot be reduced to these: [M]ental characteristics are in some sense dependent, or supervenient, on physical characteristics. Such supervenience might be taken to mean that there cannot be two events alike in all physical respects but differing in some mental respect, or that an object cannot alter in some mental respect without altering in some physical respect. Dependence or supervenience of this kind does not entail reducibility through law or definition. (Davidson 1970 214)

But Davidson’s proposal is not the only theory accepting the conjunction of the priority of the physical that is articulated in terms of supervenience and the irreducibility of the mental or the special7 properties in general. Another very important theory is the non-reductive physicalism that Putnam and Fodor developed in the 1960’s and 70’s, which is based on the powerful argument of the multiple realizability (mr) of the special or higher level properties/kinds. Following the inter-theoretic model of reduction proposed by Nagel (1961), and his idea of «bridge-laws» that can correlate predicates of the special sciences with predicates of physics in a bi-conditional form, Putnam and Fodor argue that special properties can be instantiated by, or realized on, multiple dissimilar physical structures and that, for this reason, only an open, extensive, and artificial disjunction of all the actual and possible realizers of a special property could constitute its physical reducer. 7

94

Henceforth, I will use ‘special property’ to refer to the properties of the special sciences.

Emergentism as Type Macrophysicalism

But the problem is not only that such disjunction could be empirically implausible, but that even if such disjunction could turn out to exist, it could not be logically sufficient to achieve the reduction simply because, to put it in Kim’s terms, «[a] disjunction of heterogeneous kinds is not itself a kind.» (Kim 1992b 9)8 Most authors believe that even if the higher level properties cannot be reduced because of their mr, they are metaphysically determined by their physical realizers. In fact, this idea seems to follow from the very formulation of the physical realizability of the special properties. It was Putnam (1970) who introduced this theory to account for the relation between the logical and functional states of a Turing Machine and their particular physical implementations in terms of what he called a relation between first order and second order properties. A second order property is the property of having one or another property which plays a specific role (of causal and non-causal dependencies). According to this perspective, higher level properties are both second order and mr properties because there are different physical properties which can play the functional role specified by the former. Moreover, because second order properties are fully defined in terms of their functional role, and because this role is played by each of their physical realizers, the kind of non-reductive physicalism that appeals to this notion of realization can be understood as assuming a clear metaphysical determination between the physical realizers and the higher realized properties. Commenting on this kind of nonreductive physicalism, K. Bennett, for example, says: One can picture nonreductive physicalism as occupying [a] middle ground […] After all, there are two choice points here. One is the question of the Distinctness premise —are mental [and special] properties identical to physical ones? The other is the question of physicalism— does the mental, along with everything else, supervene with metaphysical necessity upon the physical? The reductive physicalist says ‘yes’ to both. The nonreductive physicalist says ‘no’ to the former and ‘yes’ to the latter. (Bennett 2008 285-6)

8

See also Fodor 1974 404, and LePore & Lower 1989 179.

95

Juan Diego Morales Otero

Moreover, given that on this perspective each instantiation of a macro or higher level property is metaphysically supervenient on its microphysical conditions, higher properties are causally and ontologically fixed by their physical realizers and, agreeing with a priori physicalism, the instantiation of these properties will be completely explainable on the grounds of their physical bases. As Chalmers says:9 In a certain sense, phenomena that can be realized in many different physical substrates —learning, for example— might not be reducible in that we cannot identify learning with any specific lowerlevel phenomenon. But this multiple realizability does not stand in the way of reductively explaining any instance of learning in terms of lower-level phenomena. (Chalmers 1996 43)

And with respect to biological properties: Reductive explanation requires only that high-level phenomena can be explained wholly in terms of low-level phenomena. This is compatible with the «multiple realizability» of high level phenomena in low-level phenomena. For example, there may be many different ways in which digestion could be realized in a physiological system, but one can nevertheless reductively explain a system’s digestion in terms of underlying physiology. (Chalmers 2003 135, note 4)

Although this form of non-reductive physicalism which affirms the irreducibility of the higher level properties on the basis of their mr is one of the most accepted approaches, plausibly the most accepted theory of the second half of the 20th century, in recent years it has received very strong philosophical criticisms especially regarding its capability to account for the causal irreducibility of the higher properties. Authors such as Kim affirm that «the popular view that psychology constitutes an autonomous [i.e. irreducible] special science, a doctrine heavily promoted in the wake of the mr-inspired antireductionist dialectic, may in fact be inconsistent with the real implications of mr.» (1992b 3-4) One of his underlying ideas is that higher level properties 9

96

See also Fodor 1974, Melnyk 2003, Kim 2005, Shoemaker 2007, K. Bennett 2008, and Stoljar 2010.

Emergentism as Type Macrophysicalism

can be ontologically reduced once we can explain them from their lower level bases (an explanation that is completely accepted by these non-reductive proponents), just because «reductive explanation is not separable from reduction» (2005 162, note 11). Given that I will carefully examine this non-reductive physicalist proposal and its associated philosophical problems in chapter 4, here I want to focus only on the way in which this perspective understands the metaphysical connection between the basic physical level and the higher levels of our world. At this point, it is important to make explicit the idea of metaphysical (or logical)10 supervenience that these philosophers have in mind. We can use Chalmers’ formulation for this purpose: b-properties supervene logically [or metaphysically] on a-properties if no two logically possible situations are identical with respect to their a-properties but distinct with respect to their b-properties. (Chalmers 1996 35)

In other words, we can say that a set of properties b supervenes metaphysically on another set of properties a, in case it is metaphysically impossible for there to be a difference in b without a difference in a. And this is simply to say that once a-properties are fixed or determined, b-properties will also be metaphysically determined.11 As we have seen, the microphysicalist idea is that the properties and facts of our world are metaphysically supervenient on, and so, metaphysically determined by their microphysical properties or facts. Given that the non-reductive physicalism that is based on the mr of the special properties affirms that the physical bases of the supervenience of the higher level properties is not only physical but microphysical,12 As is usually accepted, we will assume that metaphysical supervenience is identical to logical supervenience, because, as McLaughlin says, «the metaphysically possible worlds are the logically possible worlds: they are one and the same space of possibilities.» (McLaughlin 2007 217, note 6) 11 As Chalmers comments, «[t]he notion of supervenience formalizes the intuitive idea that one set of facts can fully determine another set of facts.» (1996 32) 12 See, for example, Chalmers 1996, Jackson 1998, Block 2003, Kim 2005, and Shoemaker 2007. 10

97

Juan Diego Morales Otero

this kind of perspective is taken as a kind of microphysicalism; that is, as the non-reductive microphysicalism that states that because of their mr, special properties cannot be identical to, and so, not reducible to their microphysical bases. We can put the relation between reductive and non-reductive microphysicalism in this way: both are microphysicalist because they agree that all the properties of our world are metaphysically supervenient on their microphysical conditions. Nonetheless, the former claims that all the properties of our world are in fact (identical to) microphysical properties; and that this is the reason why all the facts of our world metaphysically supervene on the microphysical. But the second considers that higher level properties are second order and mr properties that, as such, cannot be identical with or reducible to their microphysical bases. But a second crucial difference between reductive and nonreductive microphysicalism has to do with their commitment to a metaphysical dependence between the higher level properties and their microphysical bases. As I will explain below, while the reductive perspective states that higher level properties are metaphysically dependent on microphysical properties, the non-reductive approach is committed to the idea that higher properties can be realized by nonmicrophysical properties and, therefore, to the idea that the former are not metaphysically dependent on the instantiation of the latter. To see this point clearly, it is useful to begin by introducing Stoljar’s idea of a weak modal difference that can exist between two properties or sets of properties f and g, namely: f is weakly modally distinct from g if and only if it is possible that f is instantiated and g is not or13 it is possible that g is instantiated and f is not. (Stoljar 2008 265)

The idea behind the weak modal distinction between two properties is that they can be both different and necessarily related. For instance, the property of being red is weakly modally distinct from the property of being colored. Something cannot be red and not colored, and in this sense they are necessarily related; but something can be colored 13

98

As it will become clear, this must be understood as an exclusive disjunction.

Emergentism as Type Macrophysicalism

and not red and, so, these properties are numerically distinct. Now we can see that this is the kind of difference that the non-reductive microphysicalists have in mind regarding the connection between the higher level properties and their microphysical bases. As Stoljar himself comments, On the second-order view psychological properties are weakly modally distinct from the physical properties that necessitate them, for, while it is impossible that the relevant physical properties are instantiated without the psychological properties being instantiated, the psychological properties might perfectly well be instantiated without the physical properties. (Stoljar 2008 265)

But for the non-reductive microphysicalist this characterization applies not only to the mind-matter connection, but to all the levels of organization of our world. As we have seen, many contemporary philosophers have relied heavily on the criterion of microphysical supervenience for the formulation of physicalism; they have assumed the metaphysical possibility that mental phenomena, as well as shopping centers and wars (Jackson 2003 85), botanical and automotive properties (Shoemaker 2010 125-6), biological properties of reproduction and evolution (Chalmers 1996 39), and even entities such as water and heat (Block & Stalnaker 1999 18, 29), can be realized by ‘immaterial mechanisms.’ And this is tantamount to the idea that all the macro or higher level properties of our world are both determined by their lower microphysical bases, and metaphysically independent of them. To achieve further clarity about the metaphysical independence of the higher properties that is claimed by this non-reductive proposal, we can recall Barnes’ characterization of the concept of metaphysical dependence: An entity x is dependent iff for all possible worlds w and times t at which a duplicate of x exists, that duplicate is accompanied by other concrete, contingent objects in w at t. That is, in any situation in which there is something exactly like x, you’ve got to have other things existing alongside it [.] You cannot ever just have x by itself. And in that sense, x depends (at every moment of its existence) on other things. (Barnes 2012 880)

99

Juan Diego Morales Otero

Although here Barnes talks about the metaphysical dependence of an entity x, she does not explicitly address the point of an entity b being metaphysically dependent on another entity a. But on the basis of her characterization, we can introduce the idea and say: An entity b is metaphysically dependent on another entity a iff for all possible worlds w and times t at which a duplicate of b exists, that duplicate is accompanied by a duplicate of a in w at t. That is, in any situation in which there is something exactly like b, you’ve got to have something exactly like a existing alongside b.

And applying this template to sets of properties we can say: A set or class of properties b is metaphysically dependent on a set of properties a iff for all possible worlds w and times t at which there is an instantiation of b, that instantiation of b is accompanied by an instantiation of a in w at t. That is, in any situation in which there is an instantiation of b, you’ve got to have an instantiation of a.

Given that according to non-reductive microphysicalism the instantiation of mental and special properties need not be realized nor accompanied by microphysical properties, it is clear that for this perspective mental and special properties are not metaphysically dependent on the second. But reductive microphysicalism argues the contrary: just as mental and higher level properties are simply microphysical properties picked out by different concepts and vocabularies, you must have microphysical properties when you have mental and higher level properties. It is metaphysically impossible to have the second without having the former.14 To summarize, both kinds of microphysicalism affirm the determination of the macro or higher level properties by their microphysical conditions, but only the reductive kind claims the metaphysical dependence of the first on the second. I now turn to evaluate the relation between the different levels under the perspectives of property dualism and emergentism. 14

100

It is important to note that this characterization of metaphysical dependence does not imply an assertion of ontological priority that is excluded by the relation of identity.

Emergentism as Type Macrophysicalism

Property Dualism and Contingent Connections

We know that British Emergentists assumed a physicalist ontology with respect to the concrete realm, that is, the realm of objects, events, states, processes, and every entity as spatiotemporally conceived; that is, they assumed what Kim calls «ontological physicalism, the view that bits of matter and their aggregates in space-time exhaust the contents of the world.» (2005 71) In this sense, for example, Alexander comments: We thus become aware, partly by experience, partly by reflection, that a process with the distinctive quality of mind or consciousness is in the same place and time with a neural process, that is, with a highly differentiated and complex process of our living body. We are forced, therefore, to go beyond the mere correlation of the mental with these neural processes and to identify them. There is but one process which, being of a specific complexity, has the quality of consciousness. […] It has then to be accepted as an empirical fact that a neural process of a certain level of development possesses the quality of consciousness and is thereby a mental process; and, alternately, a mental process is also a vital one of a certain order. (1920 5-6)

This is precisely the ontological thesis which Fodor defends and calls token physicalism, «the claim that all the events that the sciences talk about are physical events.» (1974 397) Although at first glance this seems to be a completely viable way to state a physicalist commitment, many authors15 have developed arguments that show that token physicalism is too weak to be established as an acceptable and sufficient form of physicalism, since it is compatible with property dualism, the theory that claims that the properties of the higher levels of our world are connected with the physical level properties in a merely contingent form. Let us develop a more detailed characterization of this dualist proposal and its assertion about the merely contingent connection. Kim can help us introduce the issue: Ontological dualism positing immaterial minds as entities in their own right has not been a serious option for most philosophers in 15

See, for example, Chalmers 1996 and 2006, Kim 2005, K. Bennett 2008, and Stoljar 2010.

101

Juan Diego Morales Otero

contemporary philosophy of mind. The later-day dualists are ontological physicalists who embrace a dualism of physical and nonphysical properties. They believe that complex physical systems wholly composed of bits of matter can have non-physical properties, properties that are neither properties investigated in physics nor reducible to them. Many property dualists hold that special-science properties in general, such as biological and geological properties, are no more physically than mental properties. (Kim 2009a 38)

Two important characteristics of property dualism are that it assumes that in our world there are no immaterial minds, souls, or other non-physical concrete entities, and that the concrete physical systems can and usually have properties which are not reducible to the properties of the basic physics, that is, the microphysical properties of our world. But this characterization is insufficient; both emergentism and non-reductive microphysicalism state token physicalism and irreducible higher level properties. To solve this problem, most contemporary philosophers have articulated property dualism, as opposed to a physicalist theory of supervenience, stating, as a basic tenet, that higher properties do not metaphysically supervene on the microphysical conditions. This is the main and almost only sense that contemporary philosophy has given to the idea of property dualism about a merely contingent connection between the different domains of our world. Following this thought, philosophers who advocate property dualism can accept the nomological supervenience of the higher levels on the basic physical facts. Articulating his dualist proposal about the connection between the physical and the conscious, for example Chalmers comments: I think that even if consciousness is not deducible from physical facts, states of consciousness are still systematically correlated with physical states. In particular, it remains plausible that in the actual world, the state of a person’s brain determines his or her state of consciousness, in the sense that duplicating the brain state will cause the conscious state to be duplicated too. That is, consciousness still supervenes on the physical domain. But importantly, this supervenience holds only with

102

Emergentism as Type Macrophysicalism

the strength of laws of nature (in the philosophical jargon, it is natural or nomological supervenience). In our world, it seems to be a matter of law that duplicating physical states will duplicate consciousness; but in other worlds with different laws, a system physically identical to me might have no consciousness at all. (Chalmers 2006 247)

But property dualism denies not only the metaphysical supervenience of the higher properties on the microphysical conditions, but also that there is a metaphysical dependence between them. This is precisely the meaning of the statement that the higher properties are connected with the microphysical properties in a completely contingent form: there is neither metaphysical determination nor dependence between them. Just like the non-reductive microphysicalist, the property dualist argues that all higher properties in our world are instantiated in entirely physical systems, but that there are possible worlds in which these properties can be instantiated in or realized by nonphysical systems, perhaps Cartesian mechanisms. We can articulate this idea from Stoljar’s notion of the strong modal distinction that can exist between two properties or sets of properties f and g, namely: f is strongly modally distinct from g if and only if it is possible that f is instantiated and g is not and it is possible that g is instantiated and f is not. (Stoljar 2008 266)

According to this formulation, property dualism holds that higher level properties are strongly modally distinct from the microphysical properties on which they can nomologically supervene. About the mind-matter relation, Stoljar makes explicit the assertion of property dualism as follows: [I]t is possible that psychological properties are instantiated and not physical properties, and it is possible that physical properties are instantiated and not psychological properties. Such a position says in effect that the psychological and the physical are only contingently related, but does not deny that there might be various laws connecting the psychological and the physical, so long as these laws are themselves contingent. (Stoljar 2008 270)

103

Juan Diego Morales Otero

So far, I have made it clear that the concepts of metaphysical supervenience and determination do not imply the concept of metaphysical dependence, and that property dualism denies both the relations of metaphysical determination and dependency between the higher and the microphysical properties. In what follows, I will argue that although both emergentism and property dualism can accept only a nomological supervenience of the higher levels on the microphysical ones, there is a crucial difference between these perspectives: emergentism affirms, while property dualism denies a connection of metaphysical dependency between the higher levels of our world and its basic physical level. In turn, this will make it clear that the concept of metaphysical dependency does not imply the concept of metaphysical supervenience and, therefore, that these concepts are logically independent. Finally, this will show that emergentism is a completely physicalist position that affirms the necessary connection among the different levels of the physical world, but maintaining that the higher levels do not supervene on and cannot be reduced to the lower bases from which they emerge. Emergence and Metaphysical Dependence

We have constantly claimed that emergentism states that emergent properties do not supervene on, and, therefore, are not metaphysically determined by the properties and relations of the constituents from which they emerge. It is in this sense that we say that an emergent is something different from, additional to, and non-derivative from its emergent basis. Although emergentism denies a metaphysical supervenience relation between the microphysical basis and the emergent property, it does not regard a merely haphazard or coincidental connection between them. This theory argues that in our world, if the very same configuration of basal conditions were to recur, the same emergent phenomenon would emerge again. If this were not the case, this proposal would not have a reason to say that the emergent property emerges ‘from’ some lower level conditions. It is in this sense that emergentism is committed to a nomological supervenience of the higher level properties on the basal lower properties and relations from which they emerge.

104

Emergentism as Type Macrophysicalism

It is important, however, to clarify the meaning of this nomological supervenience. To do this, we have to make explicit the components of the lower level conditions that the emergentist is denying as metaphysically sufficient for the appearance of the emergent property. This emergent basis includes not only objects, properties, and relations of the constituents, but also the laws that hold at this lower level, the set of laws governing the functioning of the basal constituents of the emergent system.16 This implies that the statement that emergent properties do not metaphysically or logically supervene on their microphysical (or constituent) basis includes the idea that these properties are emergent even with respect to the laws operating at this basal level. But the metaphysical supervenience that the emergentist accepts includes within its basis not only the laws that hold at the level of the parts, but the laws connecting the properties and events throughout the various levels of organization; the laws which Broad called ‘trans-ordinal’, Kim takes as ‘laws of emergence’, and Chalmers denominates ‘supervenience laws’. In his account of emergence, Broad considers that [T]here would be two fundamentally different types of law, which might be called ‘intra-ordinal’ and ‘trans-ordinal’ respectively. A transordinal law would be one which connects the properties of aggregates of adjacent orders. a and b would be adjacent, and in ascending order, if every aggregate of order b is composed of aggregates of order a […] An intra-ordinal law would be one which connects the properties of aggregates of the same order. (Broad 1925 77-78)

Broad thinks that trans-ordinal laws connecting the properties and entities of different levels of composition and organization can in some cases be emergent and in others merely resultant. According to Broad, as McLaughlin comments, «a trans-ordinal law is emergent if and only if it is not deducible from the laws of lower orders, lowerlevel conditions, and any compositional principles manifested at lower-levels.» (McLaughlin 1992 81) To illustrate the notion of a

16

See McLaughlin 1992 81, Chalmers 1996 33, Jackson 2003 85, Shoemaker 2007 14, and Kim 2009b 98-9.

105

Juan Diego Morales Otero

trans-ordinal emergent law and contrast it with the concept of an intra-ordinal law, Broad says: The law which asserts that all aggregates composed of such and such chemical substances in such and such proportions and relations have the power of reproduction would be an instance of a Trans-ordinal Law. The laws connecting the reproduction of living bodies with other ultimate characteristics of living bodies would be instances of Intraordinal Laws. (Broad 1925 78-9)

Now, given that the notion of a trans-ordinal emergent law implies the failure of compositional principles, it is important to have a characterization of these principles. A compositional principle is a contingent and a posteriori law that connects the values and properties of parts in order to yield the values and properties of wholes in an algorithmic or mathematical way. That is, a compositional principle is the kind of principle that allows us to calculate a global property exclusively from the properties the parts would have in isolation or in other combinations, because the global property does not introduce any organization that is not already in the other wholes that the parts can compose. British Emergentists offered scalar addition (e.g. the composition of mass and charge) and vector addition (e.g. the parallelogram law for force-vectors) as paradigms of compositional principles. In order to illustrate the operation of this kind of principles, Mill says: If a body is propelled in two directions by two forces, one tending to drive it to the north and the other to the east, it is caused to move in a given time exactly as far in both directions as the two forces would separately have carried it; and is left precisely where it would have arrived if it had been acted upon first by one of the two forces, and afterwards by the other. (Mill 1843 428, quoted by McLaughlin 1992 59)

But we need not restrict ourselves to linear mathematical functions or equations because of (at least) the possibility of accounting for the global properties of nonlinear systems through nonlinear mathematical functions or equations that take the value of their constituents’ properties in order to calculate their global behavior. A nonlinear system is a system that is characterized by having a complex behavior that

106

Emergentism as Type Macrophysicalism

is not directly proportional to the change of its properties. Given the positive and negative feedback interaction between the constituents of a nonlinear system that leads to their interference, cooperation, or competition, the causal effects of these systems are normally an exponential function of its changes or perturbations. This implies that, as Luchinsky and Stefanovska affirm, «For nonlinear systems […] a small change in a parameter can lead to sudden and dramatic changes in both the qualitative and quantitative behavior of the system.» (2005 647) This is the so called sensitivity to initial conditions that is characteristic of this kind of systems. In this sense, Silberstein and McGeever comment: An example of a chaotic, non-linear dynamical system is a driven hinged pendulum. The behaviour of this is highly sensitive to the initial speed and angle at which it is swung, and it will behave in a seemingly random fashion. The apparent randomness in the behaviour of these systems arises because of their sensitive dependence on initial conditions. This means that very small perturbations of the state of a chaotic, non-linear dynamical system will have relatively large effects on its behaviour. Generally the effects are an exponential function of the perturbation, while non-chaotic systems may respond to perturbations in a linear or non-exponential way. (1999 193)

Although the sensitivity to initial conditions of nonlinear systems normally leads to their dynamical unpredictability, their long-term trajectories typically converge on certain and recurrent patterns that are called ‘attractors’. These attractors have been taken as higher level and emergent features of nonlinear systems insofar as they cannot normally be deduced from their lower level or component’s conditions.17 But as this incapability of deduction is normally based on our factual inability to perform a precise measurement on the initial conditions of the nonlinear systems, or on our factual inability to compute or construct machines to compute the highly complex nonlinear equations that model their dynamics, then we have to claim, as Wilson says, that «nonlinearity is not sufficient indication of fundamental higher 17

See, for instance, Kauffman 1993, Wolfram 1994, and Scott 1996.

107

Juan Diego Morales Otero

level powers/interactions/laws» (2013 208), but only an indication of some kind of epistemological emergence that could be overcome by the empirical progress.18 It follows that the nonlinear mathematical functions or equations can, at least in principle or in theory, account for the global properties and behaviors of nonlinear systems exclusively on the basis of the properties the parts would have in isolation or in other combinations and, in consequence, should be considered as (nonlinear) compositional principles that relate the properties of the parts to the properties of the wholes in a reductive, non-emergent way. In this sense, we can say that the failure of the operation of compositional principles gives a specific content to the idea that an emergent is more than the sum (or the mathematical linear or non-linear function) of its parts. The emergentist statement, then, is that some properties of some purely physical aggregates (such as the property of having the power of reproduction) are emergent because the trans-ordinal laws that connect them with their physical bases are emergent. In turn, the idea that a trans-ordinal law is emergent is that this law relates different levels of composition in a non-derivative way, because it introduces an organization or relational structure that is not already in the other wholes that the parts can compose. It follows that this law cannot be deduced, calculated, or anticipated from the knowledge of the conditions, laws, and compositional principles operating at the level of the parts. Since trans-ordinal emergent laws cannot be deduced, calculated, or anticipated from the lower laws, conditions, and principles, then we must claim that in this sense trans-ordinal emergent laws are principles that connect the different levels in a contingent form. In fact, this is precisely the idea that follows from the tenet that the emergent is different from or additional to its basal conditions. And it is this type of contingency between the emergent and its basis which has served as the ground for the extended formulation of emergence in terms of nomological supervenience.19 But it is precisely on the basis of these On this epistemological reading of the higher level properties of nonlinear systems, see, for example, Kellert 1993, Newman 1996, Bedau 1997, Silberstein & McGeever 1999, Silberstein 2001, and Wilson 2013. 19 See, for example, van Cleve 1990, McLaughlin 1997, Kim 2006a (but see 2009b for the opposite view), and Bernal 2012. 18

108

Emergentism as Type Macrophysicalism

considerations that many theorists have argued that emergentism should be understood as a property dualism. As I have claimed, most contemporary philosophers have articulated the property dualist proposal as opposed to a theory of supervenience, stating that the fundamental tenet of this kind of dualism is that higher level properties do not metaphysically supervene on, and therefore are not completely determined by the microphysical conditions. It follows that if we consider that physicalism should be understood as a metaphysical supervenience theory, both property dualism and emergentism will be counted as non-physicalist proposals. However, in chapter 2, we articulated sufficient reasons to dismiss this formulation of physicalism. Rather, I think that the crucial concept to understand physicalism is that of metaphysical dependence. We have seen that property dualism asserts a merely contingent connection between physical properties and properties of the special sciences, particularly of psychology. This means that according to this view, there is neither a determination nor a connection of metaphysical dependency between the two sets of properties and that, therefore, both the instantiation of the physical properties without the instantiation of the higher level properties, and the instantiation of the special properties without their physical realization are entirely possible. As we said following Stoljar’s notion, property dualism states that higher level properties are strongly modally distinct from the microphysical properties on which they can nomologically supervene. But here lies the crucial distinction between emergentism and property dualism (even non-reductive microphysicalism). Although emergentism argues that emergent properties are not metaphysically determined by the properties and relations of their constituents (ultimately, of their microphysical constituents), it claims that there is a metaphysical connection between them because the former are completely dependent on the latter. That is, the emergentist ontological picture claims that for all possible worlds w and times t at which there is an instantiation of an emergent property e, you must have an instantiation of some microphysical conditions m in w at t. This metaphysical dependence between the emergent and its basis follows from two facts that we have already elaborated: firstly, from the fact that

109

Juan Diego Morales Otero

the emergence connection is a type of mereological relation which, as such, connects the properties of the whole with the properties of the parts in an essential form. And, secondly, it follows from the fact that the emergent property is not simply different from and additional to the properties of the constituents, but a special organization of these elements which, as such, fully and ontologically depends on them. This point can be made clearer by recalling that even if quantum states of entanglement, relativistic gravitational fields, or neurophysiological entities end up not being metaphysically supervenient on the properties of their microphysical constituents, they must still be considered as paradigmatically physical phenomena, in that case, as macro or emergent physical phenomena. It is a conceptual mistake to consider them as ghostly or non-physical phenomena due to the mere fact that they are not metaphysically supervenient on their constituent bases. And the reason is that these phenomena would be no more than a complex organization or relational structure of their lower microphysical conditions, an organization that, as such, cannot be instantiated or realized by ‘ghostly mechanisms’. To put it in other terms, we can say that if quantum states of entanglement, relativistic gravitational fields, or neurophysiological entities turn out to be emergent physical phenomena, it is possible that their lower physical bases be instantiated in worlds different from ours without their correspondent emergence (because of the absence of the necessary trans-ordinal emergent laws). But simply because they are no more than the complex higher organization that their lower microphysical bases can acquire, it is impossible for them to be instantiated in any world without their microphysical bases. It is on the basis of this idea that we must understand Barnes’ characterization of emergence as dependent fundamentality, whose consequences for the relation between the microphysical and the higher levels are illustrated by this author through the case of the mind-body relation. According to emergentism, Barnes says, Mental properties are part of the fundamental make-up of reality. They do some sort of work (causation, truthmaking, etc.) that cannot be done by anything else. If God wants to create a world like ours, with creatures like us, she cannot just create microphysical particles in certain

110

Emergentism as Type Macrophysicalism

arrangements and get the mental properties for free. She has to create the mental properties as well. But the existence of mental properties is both caused and sustained by the collective activity of certain physical properties. Without those physical properties persisting in very specific arrangements, the mental properties would cease to exist. That is, according to the emergentist, the world of pure Cartesian minds is impossible. Mental properties are dependent: they depend on the existence of physical properties and relations. (Barnes 2012 887)

is s metaphysically supervenient on (determined by) m?

is s metaphysically dependent on m?

is s a physical property?

is s fundamental?

is s derivative?

is s explainable in m terms?

Reductive microphysicalism

is s identical to m?

The results we have reached about the characterization of the different ontological approaches are summarized in the following diagram, where we take s as a macro or special property and m as the microphysical conditions which are or can be the basis of its supervenience or emergence:

yes

yes

yes

yes

yes

no

yes

Non-reductive microphysicalism Emergentism

no

yes

no

no

no

yes

yes

no

no

yes

yes

yes

no

no

Property dualism

no

no

no

no

yes

no

no

The conceptual articulation we have developed points out that the emergentist theory cannot be understood according to the logical model of any kind of dualism; furthermore, that it must be understood as a clear physicalist proposal in which the properties of physical systems are metaphysically connected with the properties of their parts even if they cannot be reduced to them. Although the interpretation of emergentism as a kind of non-reductive physicalism in recent years has been eclipsed by the f­ ormulations

111

Juan Diego Morales Otero

of physicalism in terms of the metaphysical supervenience theory, which we have analyzed and rejected in chapter 2, authors like Kim (1992, 1999), Crane (2001), and Wimsatt (2007) consider that emergentism has not only been the first philosophical formulation of a non-reductive physicalism, but that it can prove to be its better articulation, because, as Crane put it, non-reductive physicalism is committed to mental [and special] properties being emergent properties. [And, as this author says,] it follows that the problems for emergentism —especially the problems of mental [and special] causation— are also problems for non-reductive physicalism, and they are problems for the same reason. (Crane 2001 207)

These problems will be our central concern in chapters 4 and 5. Two Types of Ontological Emergence

Now that we have a rather clear and constrained idea of the meaning of emergence in an ontological sense and that we know about the distinctions between emergentism and other related ontological proposals, it is important to use the criteria developed thus far to distinguish two general types of ontological emergence. This task will allow us to grasp the internal diversity of the phenomenon and to refine the boundaries of the concept, in order to understand its philosophical and scientific consequences. We have seen that there are several approaches to the distinction between epistemological and ontological emergence, but few theorists develop a taxonomy and draw differences between types of ontological emergence.20 Van Gulick considers three types: specific value emergence, modest kind emergence, and radical kind emergence. The author characterizes the first type as follows: Specific Value Emergence. The whole and its parts have features of the same kind, but have different specific subtypes or values of that kind. For example, a bronze statue has a given mass as does each of the molecular parts of which it is composed, but the mass of the whole is different in value from that of any of its proper material parts. (Van Gulick 2007 61) 20

112

See, for example, Wimsatt 2007, Van Gulick 2007, and Wilson 2013.

Emergentism as Type Macrophysicalism

Although in this case we find that the whole has properties that are different from those of its parts, the emergentists regard it as a clear and paradigmatic type of a merely resultant or non-emergent phenomenon. The reason is that in this case, the property of the whole can be precisely and exhaustively calculated, explained, and predicted from the properties of its constituents with the use of compositional mathematical principles which take the whole as a mere aggregate of the parts. While in the strict sense the overall mass of an object cannot be logically deduced from the mass of its constituent molecules, simply because the compositional principle of scalar addition of mass is contingent and only known in an a posteriori way, it is clear that we can calculate and explain exhaustively the mass of the whole from its components, because this principle applies to all the different relational configurations that the components can maintain. We can then say that in cases such as mass, charge, and force-vectors, we do not find emergence because the whole is not more than the sum, or the mathematical function, of its parts. Let us examine the other two types of emergence in the categorization of Van Gulick: Modest Kind Emergence. The whole has features different in kind from those of its parts (or alternatively that could be had by its parts). For example, a piece of cloth might be purple even though none of the molecules that make up its surface could be said to be purple. A mouse might be alive even if none of its parts (or at least none of its subcellular parts) were alive. Radical Kind Emergence. The whole has features that are both (1) different in kind from those had by its parts, and (2) of a kind whose nature and existence is not necessitated by the features of its parts, their mode of combination and the law-like regularities governing the features of its parts. (Van Gulick 2007 62)

As we saw, in specific value ‘emergence’, the property of the whole is of the same kind but has different value from that of its parts, while in the last two cases, the property of the whole is different in kind from that of its parts. According to the second condition introduced by Van Gulick, the distinction between these two types of emergence is that in radical kind emergence, the global property is not necessitated or

113

Juan Diego Morales Otero

determined by the basal conditions from which it emerges, namely, the properties, relations, and laws operating at the level of the parts. But this is precisely the characterization of ontological emergence we have developed. As Van Gulick comments, [a]ccepting radical-kind emergence […] would require abandoning the atomistic conception typically embraced by mainstream physicalism. [… The idea that] the only law-like regularities needed for the determination of macro-features by micro-features are those governing the interactions of those micro-features in all contexts, systemic or otherwise. (2007 62)

But then modest kind emergence cannot be considered as a real type of ontological emergence, simply because it would be a case in which the global property would metaphysically supervene on its basal conditions and, thus, it could be exhaustively explained in their terms. In fact, one of the examples which Van Gulick uses to illustrate modest kind emergence, the property of being alive, is precisely one of the paradigmatic examples that emergentists have used to illustrate their idea of an emergent property, a property which is neither determined nor explainable on the basis of its basal lower conditions. Following this idea, we can call kind emergence the type of phenomenon that implies that the emergent property (i) does not metaphysically supervene on the properties, relations, and laws that hold at the level of the parts of the system, (ii) cannot be explained on the basis of these lower level conditions, even using compositional principles that govern the aggregation of lower level properties (that is, linear or non-linear mathematical principles that deliver the global property directly and exclusively from the properties of the components), and (iii) is different in kind from the properties of these parts. Paradigmatic cases that would constitute examples of this type of emergence are biological, psychological, and social properties. But there is another category of emergence which involves a failure of both metaphysical supervenience and compositionality, and is nevertheless characterized by the fact that both emergent and basal properties are of the same kind. We have seen that there are prima facie strong empirical reasons to affirm that at least some clearly

114

Emergentism as Type Macrophysicalism

physical phenomena are emergent: entangled quantum systems and at least some cases of relativistic gravitational fields of two or more objects are physical phenomena that, as far as we know, are neither metaphysically determined by nor explainable from the pure information of the properties and relations of their physical components. Whether these phenomena end up being emergent or not is essentially an empirical question. But these phenomena do not respond to the characterization of kind emergence, just because the emergent property in these cases is of the same kind as, but with different value from, the properties of its basis. In one of the simplest cases of quantum entanglement, an entangled system of electrons has a spin with a value that does not supervene on or cannot be calculated from the value of its constituent electrons. A similar thing seems to happen within the scope of general relativity: at least some cases of relativistic gravitational fields of two or more objects involve nonlinear interactions among the gravitational fields of the constituent objects, and it seems that the total fields cannot be recovered, even in principle, from the lower level interaction of their components (Demaret, Heller & Lambert 1997 141-2). In these plausible cases of emergence, the emergent property and the properties of the parts are of the same kind; however, the emergent property does not supervene on and cannot be deduced or calculated from the properties of the constituents, even using a posteriori compositional principles. Following Van Gulick’s categorization we can call value emergence the type of phenomenon which implies that the emergent property (i) does not metaphysically supervene on the properties, relations, and laws that hold at the level of the parts of the system, (ii) cannot be explained from these lower level conditions, even using compositional principles that govern the aggregation of lower level properties (that is, linear or non-linear mathematical principles that deliver the value of the global property directly and exclusively from the values of the components), and (iii) is of the same kind as the basal properties of the parts. We can now say that the facts of our world could fall under two basic and general types of ontological emergence. While both phenomena involve a failure of supervenience, explanation, and prediction from

115

Juan Diego Morales Otero

the lower bases, in one of the cases we would have the appearance of a different and irreducible value, while, in the other, we would have the appearance of a completely new kind of existent, one case implying a quantity emergence and the other a quality emergence. After the clarification of the concept of ontological emergence, the distinctions between emergentism and other related ontological proposals, and the difference between two general types of ontological emergence, value and kind emergence, we still have to provide a philosophical explanation (absent in most –if not all– theories about causation, physicalism, emergence, and scientific explanation) about how this kind of phenomenon is possible from a dynamic and causal viewpoint. That is, we still have to provide a conceptual articulation of how the dynamics of emergent systems is possible and of how they work, a task that involves the analysis of the very evasive notions of emergent and downward causation. That is the subject of our fifth and final chapter. However, before doing so, it is important to carefully examine the strong criticism that several authors have leveled in recent years at the non-reductive physicalist proposal based on the mr of the higher level properties, especially with respect to its capability to account for the causal irreducibility of these properties. Given that contemporary emergentism should be understood as a clear response to this criticism, it is important to address this issue in our next chapter. Summary

In this chapter, I developed a general characterization of ontological emergence whereby this phenomenon should be understood as a special organization or relational structure that the constituents of a system acquire, and which introduces a causal and dynamical difference that is not completely determined by the causal factors of these constituents. This understanding is articulated following the idea that an entity is emergent if it is metaphysically fundamental and dependent on its constituent parts, without being in principle or theoretically explainable from them. Then, I argued that emergentism should be treated as a clear kind of non-reductive physicalism. The basic line of the argument consisted in showing the logical independence of the concepts of metaphysical determination and metaphysical dependence.

116

Emergentism as Type Macrophysicalism

Although it is true that both emergentism and property dualism deny the metaphysical supervenience between the microphysical level and the higher levels of our world, there exists a crucial difference between these perspectives: emergentism affirms, while property dualism denies a metaphysical dependency connection between the higher levels and the level of microphysics. I explained the reason why this must be so. Finally, I distinguished two general types of ontologically emergent entities, value and kind emergence. I said that while both phenomena involve a failure of supervenience, explanation, and prediction from the lower bases, in one of the cases we would have the appearance of a different and irreducible value, while in the other we would have the appearance of a completely new kind of existent: one case implying a quantity emergence and the other, a quality emergence.

117

CHAPTER 4 Non-Reductive Physicalism and the Problem of Causation in the Special Sciences No one has contributed as much to our understanding of the problems of mental causation in recent years as Jaegwon Kim. We non-reductive materialists must face up to the serious difficulties he has raised for our position.

block 2003 133

In this chapter, I analyze the problem that non-reductive physicalism (nrp) in its functionalist account, the most important philosophical naturalistic proposal of the last four decades, has to face when defending the reality and irreducibility of the causal power of the properties of the mental and special sciences. nrp maintains its physicalist commitment by affirming that mental and higher level properties supervene on and are realized by basic or microscopic physical properties. In addition, this proposal argues that higher level or special properties are irreducible to the properties on which supervene because are multiply realizable by the latter. Jaegwon Kim is one of the authors that have developed very important criticisms of the non-reductive physicalist (nrpist) proposal. With his well-known causal exclusion argument, he purports to show that for the physicalist there are only two options with respect to the causal status of special properties: either reductionism or epiphenomenalism. Appealing to the interventionist proposal, some philosophers argue that the exclusion argument is incorrect on empirical grounds. I agree with the interventionist approach that the exclusion argument is not conclusive, but I argue that there is another

119

Juan Diego Morales Otero

important argument that Kim has built to show that nrp cannot account for the reality of the causal status of higher level properties: what I call the argument of causal individuation of natural kinds. On the basis of these criticisms of nrp, some philosophers support the functional reductive approach (see Appendix), arguing that to avoid eliminativism about higher level properties, we must reduce them ‘locally’ to each of their physical realizers. The problem is that, on this view, special properties should be finally eliminated because the categories of the special sciences are not truly referring to real special properties but only to sets of multiple and dissimilar basic physical properties in a somehow contingent and indirect way. If this is right, the overall conclusion is that neither nrp nor the functional reductive proposal seems to provide a satisfactory account of mental and special properties as real constituents of our physical world. Furthermore, their different problems arise from their shared assumption of the metaphysical supervenience of the macro-properties on their microphysical realizers or conditions, an assumption that is plausibly an empirically false claim -as we have seen in chapter 2. These consequences should be seen as direct reasons to articulate a macrophysicalist (that is, emergentist) account of causation. Supervenience and Multiple Realizability

In chapter 3, I argued that the philosophical theory that is based on the proposals of the early 20th century British Emergentists should be considered a clear form of non-reductive physicalism. However, this position differs crucially from the most influential physicalist approach of the past five decades of analytic philosophy, namely, the anti-reductionist functionalist theory that emerged in the 1960s. This kind of non-reductive physicalism, considered by philosophers like Jaegwon Kim as the «position that can deservedly be called ‘the received view’ of today,»1 can be understood as the conjunction of three basic tenets: first, its physicalist commitment, that is, its assertion of a supervenience theory, according to which the properties of our world supervene on and are therefore metaphysically determined by their basic or microphysical facts. Second, its 1

120

Kim 1993a 339; see also 2009a 46.

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

anti-reductive principle: the idea of the multiple realizability (mr) of the properties or natural kinds of the special sciences –from chemistry and biology to psychology, sociology, and economics–, which implies the difference and irreducibility of the higher level properties vis-à-vis their lower-level, ultimately microphysical, realizers. And third, the idea of a higher level causal realism, whereby the properties of the special sciences are genuine causal factors for the instantiation of both higher level and basic physical properties. The purpose of this introductory section is to clarify the first two tenets of non-reductive physicalism. The question about its higher level causal realism under these two assumptions will be the central issue throughout the rest of the chapter. The physicalist commitment of this kind of non-reductive physicalism (henceforth nrp),2 that is, its assertion of a supervenience theory, seems to follow from the very formulation of the physical realizability of the special3 properties. It was Putnam (1970) who introduced this theory to account for the relation between the logical and functional states of a Turing Machine and their particular physical implementations in terms of what he called a relation between first order and second order properties. A second order property is the property of having one or another property which plays a specific role (of causal and non-causal dependencies). According to this perspective, higher level properties are both second order and mr properties, because there are different physical properties which can play the functional role specified by the former. Because second order properties are fully defined in terms of their functional role, and because this role is played by each of their physical realizers, nrp assumes a clear metaphysical determination between the physical realizers and the higher level realized properties. Two points should be clarified here. On the one hand, it is important to note that functionalism and its associated understanding of the higher level properties as second order properties is not a Throughout this chapter I will label this kind of non-reductive physicalism as nrp, and in the required case I will highlight its differences with emergentism. 3 Recall that the phrase «special properties, laws, or events» is just an abbreviation for «properties, laws, or events of the special sciences.» 2

121

Juan Diego Morales Otero

physicalist doctrine as such. It is physicalist insofar as it claims that, at least in our world, all the realizers of the higher level properties are basic physical properties that play or occupy the functional roles of the latter and that, to that extent, metaphysically determine them. On the other hand, we have to note that on the nrpist proposal, the metaphysical supervenience of a higher level property on its lower physical basis should be understood as implying not only the physical property which plays the functional role in question, but the laws that hold at this lower level, that is, the set of laws governing the functioning of the realizer of the second order property. Let us suppose that we can give a functional analysis of pain in terms of its causal role: as the state that is normally caused by tissue damage (td) and that in turn causes aversive behavior (ab). Let us say that this is the causal role p of pain. Suppose that the neurological property n1 is one of the CR realizers of CR p, that is, that n1 is one of the physical properties that play the causal role of being caused by td and causing ab. Now, it is clear that n1 can be one of the physical realizers of CR p only because it is nomologically related as an effect of td and a cause of ab. If the physical laws governing n1 were different, if n1were no longer nomologically related to td and ab in the relevant ways, it could neither play the causal role nor be a physical realizer of pain. The result is that n1 is a physical realizer of and metaphysically determines the instantiation of pain, only under the assumption that the physical laws that govern n1 are fixed.4 This clarification of the metaphysical supervenience of the special properties on their physical realizers is relevant because, as we shall see, this claim constitutes one of the basic premises on which the criticisms of the nrpist anti-reductionism are woven. Despite the fact that it follows from the very characterization of a second order property that every instantiation of a higher level property is completely determined by its physical realizer, and that this consequence ultimately allows a clear and customary understanding of the relation of realization as implying a metaphysical supervenience 4

122

It is in this specific sense in which Kim affirms that the physical realizer of a functional or second order property should be understood as nomologically sufficient for it (see, for example, Kim 1998 23); that is, as sufficient under the lower physical laws governing its interactions.

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

claim,5 we have to say that it is plausible that this understanding was not explicit when the functionalist proposal arose. As Kim comments, [F]ew functionalists, especially in the early days of functionalism, made an explicit effort to explain what the realization relation consisted in -what this relation implied in terms of the traditional options on the mind-body problem. I believe that the idea of ‘supervenience’ came to the fore in the ’70s and ’80s, in part to fill this void. The doctrine that mental properties are supervenient on physical properties seemed perfectly to meet the needs of the post-reductionist physicalist in search of a metaphysics of mind; for it promised to give a clear and sturdy sense to the primacy of the physical domain and its laws, thereby vindicating the fundamental physicalist commitments of most functionalists, and do this without implying physical reductionism, thereby protecting the mental as a distinctive and autonomous domain. […] This position, standardly called ‘nonreductive physicalism’, has been the most influential and widely shared view not only about the mindbody relation but, more importantly, about the relationship between ‘higher-level’ properties and their underlying ‘lower-level’ properties in other domains as well. (1997 186-7)

Let us now turn to the anti-reductive principle of nrp. We have to say that it is articulated as a response to the inter-theoretic model of reduction proposed by Nagel (1961). At the core of this model of reduction is the idea of the existence of ‘bridge-laws’ that can correlate predicates of the special sciences with predicates of physics in a bi-conditional form, in such a way that there can be a complete and nomologically necessary co-extension of special and physical properties. Putnam and Fodor argue that special properties can be instantiated by, or realized on, multiple dissimilar physical structures and that, for this reason, only an open, extensive, and artificial disjunction of all the actual and possible realizers of a special property could constitute its physical reducer. 5

For the explicit statement of this idea see, for example, Shoemaker 2007 2, K. Bennett 2008 285, and Papineau 2010 180.

123

Juan Diego Morales Otero

One can take the problem as stating that for the reduction of a special property, it is necessary to find all its physical realizers in order to identify it with a disjunctive property in the basal domain. This is the so-called disjunctive approach; as Kim says, The disjunctive strategy, briefly, is this: if m is multiply realizable in, say, three ways, p1, p2, and p3, then why not take the disjunction, p1 ˅ p2 ˅ p3, as m’s coextension in the base domain? Clearly, given that each of the pis is a realizer of m, it must hold with (at least) nomological necessity that pi → m, and given that the pis are all the nomologically possible realizers of m, it must hold with (at least) nomological necessity that m ↔ (p1 ˅ p2 ˅ p3). (1998 93)

But the problem with this strategy is not merely empirical; it not only has to do with the discovery of all the nomologically possible realizers of a special property. It is that even if we come to know all these nomologically possible realizers, their disjunction could not be logically sufficient to achieve the reduction. This is so because, to put it in Kim’s terms, a «disjunction of heterogeneous kinds is not itself a kind.» (1992b 9)6 In this sense, LePore and Lower affirm:

If there are infinitely many physical (and perhaps nonphysical) properties which can realize [a special property] f then f will not be reducible to a basic physical property. Even if f can only be realized by finitely many basic physical properties it might not be reducible to a basic physical property since the disjunction of these properties might not itself be a basic physical property (i.e., occur in a fundamental physical law). We will understand ‘multiple realizability’ as involving such irreducibility. (LePore & Lower 1989 179)

What seem to be sufficient but also necessary for reduction are property or kind identities, not merely nomological or even metaphysical co-extensions. If special properties are really mr by different basic physical properties this type of identities is precluded. The reason is that the closest connection these properties could have is a metaphysical co-extension, that is, a metaphysically necessary 6

124

See also Fodor 1974 404.

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

co-instantiation of every special property with one of the physical disjuncts that in each case functions as its realizer. The problem with nrp is that it has deep troubles articulating the irreducibility of the special properties that follows from their mr in a satisfactory way. Let us consider the arguments supporting this claim. The Causal Exclusion Argument

Although nrp seems quite clear and conclusive in its attempt to combine in a thoughtful way the priority of the physical domain and the irreducibility of the higher level properties, its viability has been extensively and complexly discussed over several decades, especially its success in accounting for the reality and irreducibility of the causal relevance of both the properties and the laws of the special sciences.7 Over several decades, Jaegwon Kim, who is probably the philosopher who has most criticized the nrpist proposal on the basis of metaphysical arguments, has built the most discussed philosophical argument about mental and special properties causation in a physicalist framework, the so-called causal exclusion argument. In a nutshell, as Kim says, «the problem of causal exclusion is to answer this question: Given that every physical event that has a cause has a [sufficient] physical cause, how is a mental [or special] cause also possible?» (1998 38, original italics) The intended upshot of Kim’s argument is «that, for the physicalist, there are only two options left: reductionism and epiphenomenalism.» (Kim 2005 70) Given that «epiphenomenalism strikes most of us as obviously wrong, if not incoherent» (2005 70), or as simply «absurd» (McLaughlin 2006 40),8 the general physicalist result of the exclusion argument would be that the reduction of both the mind and the properties of the special sciences is an inevitable consequence. Let us examine this stunning argument. See, for example, Block 1980, 1990, 2015, Fodor 1974, 1990, 1997, Kim 1989, 1998, 2005, LePore & Loewer 1989, Lewis 1980, 1994, McLaughlin 2006, Papineau 2010, and Shoemaker 2007, 2010. 8 See chapter 1 and section 4 of this chapter about this assumption. See also Levine 2001 23 and Silberstein 2001 84. 7

125

Juan Diego Morales Otero

The Supervenience Argument

To begin with, we can say that the causal exclusion argument can be understood as the second and crucial step of a more general argument that attempts to show that the properties of the special sciences in a physicalist framework, insofar as they are both different from and irreducible to their physical bases of realization, cannot have any causal influence or relevance: neither a causal influence on other special properties, nor a causal influence on the physical properties which function as their realizers. The first step of this general argument is what Kim calls the supervenience argument, which shows that a higher level property can cause the instantiation of another higher property only if the first can cause the instantiation of the physical realizer of the second. That is to say that causation within the higher levels can only exist if there is a causal influence of these levels on the lower ones, ultimately on the microphysical level, a kind of causation that we can call downward causation. The starting point of the supervenience argument is the nrpist assumption that special properties have causal power or relevance over (are causally responsible for) the instantiation of different properties, in particular, over properties of their same level of organization. In one case, then, we can say that a mental property m is causally responsible for the instantiation of another mental property m*.9 As nrp is committed to the supervenience of the higher level properties on the basic physical properties which are their realizers, then we should state that m* has a physical basis of supervenience, say p*, which is, in the already clarified sense, completely or metaphysically sufficient for m*. Now we can quickly see that from these two basic assumptions it follows that, as Kim insists, «the m-instance [can only cause] m* to be instantiated by causing p*, m*’s physical realization base, to be instantiated.» (1993a 352, see also 2009a 40) That is, the necessary occurrence of mental to physical (more generally, special to physical) 9

126

As commonly done in the pertinent literature, I will, in many cases and in order to avoid unnecessary complications, affirm that a property can cause another property, by which I shall mean that the instantiation of the first can cause the instantiation of the second. I will make it explicit when the context so requires.

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

causation; in other words, the existence of downward causation. The rationale is that nrp, insofar as it is a physicalist proposal, argues that at least in our world all the realizers of higher level properties are basic physical properties that play or occupy the functional roles of the latter; in consequence, special properties cannot be instantiated in a brute or basic form and, in this sense, are dependent on the instantiation of their lower physical bases. Briefly, special-to-special causation implies special-to-physical causation because you cannot have a special instance without having its physical realizer; to the extent that you want to cause an m*-instance, you will have to cause its physical base of realization, p* (or p*2, p*3…). The Principle of Physical Causal Closure

Once the supervenience argument has shown that all higher level causation implies a downward causation -from a higher level property to the physical property which realizes a subsequent higher property-, Kim intends to show, with the causal exclusion argument, that the downward causation that is necessary for any higher level causation is not possible. So, according to Kim, the general picture of the nrpist final consequences is that «neither mental-to-physical nor mental-to-mental causation is possible -that is, [that] the mental is epiphenomenal tout court.» (2009a 41) As it is clear that mental properties are taken here as representative of the special properties in general, the overall picture of the nrpist proposal in favor of which Kim proceeds to argue is that the properties of the special sciences are simply epiphenomenal. The first step of the causal exclusion argument is formulated from the principle that Kim calls Causal closure of the physical domain. If a physical event has a cause at t, it has a sufficient physical cause at t. (2009a 38)

Meanwhile, in honour of Ewald Hering, a German physiologist who in the late 19th century reached the empirical conclusion «that there seem to be no “gaps” in physiological brain processes to be filled by mental events» (McLaughlin 2006 40), McLaughlin called the following principle ‘Hering’s No Causal Gap Thesis’:

127

Juan Diego Morales Otero

For any (caused) physical event, p, there is a chain of entirely physical events leading to p, each link of which causally determines its successor (or determines the objective probability of its successor). (2006 41)

Like Hering, Papineau considers that this principle is an empirical claim that as far as we know has been confirmed by neurobiological results. Moreover, he argues that this principle has been the main basis for «the rise of physicalist doctrines in the second half of [the 20th] century.» (Papineau 2001 32) The principle of physical causal closure10 (henceforth ppcc) is usually understood as an empirical statement about our world that constitutes the principal ground for showing that there are no immaterial agents or events, non-physical causal influences that may affect the course of the physical processes and, therefore, that the Cartesian interactionist dualism, which mixes physical and nonphysical events in a single causal chain, is empirically excluded. Nonetheless, if we consider that Cartesian dualism is not a doctrine that is falsified by merely empirical means but, as we suggested in chapter 2, that it can be a theory with deep conceptual and consistency problems, then the ppcc could come to be understood as a conceptual rather than an empirical claim. Kim suggests this point when he comments: It isn’t just that the principle of physical causal closure is the operative assumption in scientific research -remember that in science success is what counts. It may well be that there is a conceptual incoherence in the idea that there are nonphysical causal forces outside space-time that can causally intervene in what goes on in the spacetime world. (2011 113)

In spite of this, the ppcc attempts to have a scope beyond the mere exclusion of Cartesian dualism and, with this, the assertion of an ontological or token physicalism. The main objective of its constant formulation is evaluating any proposal which affirms that the 10

128

Or the completeness of physics, or the physical causal completeness, as it is usually called.

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

properties of the special sciences are both different from and causally relevant with respect to the physical properties that function as their realizers. More specifically, the nrpist claims that these two sets of properties are related in an essential and metaphysical way. According to this purpose, the ppcc must be interpreted in one of two following ways. If we consider -as we did in chapter 1 since I believe we must do so in order to develop a clearer and more useful understanding of our central issues- that the causally related entities are events, and that these are understood as instantiations of properties at a spatiotemporal zone and, therefore, that a physical event can instantiate mental and, in general, special properties, then we would have something like the following: (i) If a physical event has a cause at t, insofar as this cause is sufficient (for its objective probability), it has a physical cause at t that is sufficient in virtue of its physical properties.11

In this interpretation, the principle affirms that the objective probability of any physical event is causally determined by the instantiation of the physical properties of its causal antecedents. In this case, the causal relevance is clearly located in the physical properties rather than in the purely physical events or physical concrete entities. In contrast to the anti-Cartesian reading of the ppcc, this formulation has a clear empirical status that should be tested on the basis of the evidence that the different sciences, basic physical and special, could provide through the affirmation of their proposed specific natural kinds. Let us now consider the second and most common way of articulating this version of the ppcc. In recent years, there has been a growing tendency to reject the idea that mental events and events in the special sciences can be considered as identical to physical events. This perspective is mainly based on Kim’s proposal12, whereby the events must be individuated in terms of the different properties that constitute them: since physical properties are different from mental Compare with McLaughlin’s formulation: «Physical Causal Comprehensiveness. Whenever two physical events are causally related, they are so related in virtue of being instances of physical types.» (2006 41) 12 See, for example, Kim 1990.

11

129

Juan Diego Morales Otero

and special properties, then physical events should be different from mental and special events. In this case, nrp is understood as the proposal which argues that even though mental and special events are different from physical events, the former are necessarily realized by the latter.13 According to this idea, the formulation of the ppcc becomes simpler and can be understood as follows: (ii) If a physical event has a cause at t, insofar as it has a sufficient cause (for its objective probability), it has a physical event as sufficient cause at t.14

The basic idea behind these readings of the ppcc is the assertion that the causal relevance of the physical properties is sufficient for the causal determination of the (objective probability of the) instantiation of any physical property. We should note that if this principle is true about any domain of our world, it must be true about its basic physical domain, since it is clear that the macrophysical or higher levels cannot maintain this kind of causal closure. As Kim claims, «neither the mental nor the biological domain is causally closed; there are mental and biological events whose causes are not themselves mental or biological events. A trauma to the head can cause the loss of consciousness and exposure to intense radiation can cause cells to mutate.» (2005 16-7) But in this sense, the issue is not only that mental and biological levels are not causally closed; neither mental and biological, nor geological, physiological, and aerodynamic domains are causally closed. As Kim highlights, «[t]he same is true of macro level physics and chemistry. It is only when we reach the fundamental level of microphysics that we are likely to get a causally closed domain.» (2005 65) A key point is whether in this reading the ppcc is empirically correct. For present purposes, however, the examination of this empirical feasibility is not opportune and we will leave it for our next chapter, given that our current task is to assess the causal exclusion argument from which the main criticisms of the nrpist proposal have See, for example, Melnyk 2003, McLaughlin 2006, Shoemaker 2007, Bennett 2008, and Papineau 2010. 14 Compare with the previous formulation by Kim. 13

130

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

been woven. What is required for this purpose is to know whether nrp accepts the ppcc as a correct principle. Most theorists argue that this is the case; there is not, however, a clear justification for this acceptance. I think that the reason why nrp accepts the ppcc is relatively simple. Since this approach assumes that the higher level properties are metaphysically determined by their physical realizers, then this perspective assumes, in particular, that the causal powers of the higher level properties are metaphysically determined by the causal powers of their physical realizers. We can say, then, that on this proposal, all causal powers are either basic physical causal powers or metaphysically derived from basic physical causal powers. It follows that if a physical event p* has a cause at t, there is a physical event p causally determining (the objective probability of) p* at t, simply because p is causally-nomologically sufficient for (the objective probability of) p* and metaphysically sufficient for any nonphysical special property s which could also be causally-nomologically sufficient or relevant for (the objective probability of) p*. On the nrpist perspective, there is no metaphysical possibility of the causal occurrence of p* without another physical property p causally determining its objective probability. The Exclusion Principle

Let us then recall that the first premise of the exclusion argument is the ppcc. The second premise is the anti-reductionist principle: mental and higher level properties are different from and irreducible to the basic physical properties that realize them. We have said that the supervenience argument shows that on the nrpist perspective all higher level causation implies some kind of downward causation. Let us return to the example we have been working on and let us consider that an instance of the mental property m causes an instance of the physical property p*. Due to the ppcc, there is a physical event that is causally sufficient for p*; let us say p. And given the anti-reductionist tenet of nrp, we have to claim that m ≠ p; that is, that the mental cause of p*, the m-instance, is different from its physical cause, the p-instance. At this point the argument applies what is known as the exclusion principle, which is the third and crucial premise. Kim formulates it as follows:

131

Juan Diego Morales Otero

Causal exclusion. No event can have more than one sufficient cause occurring at any given time -unless it is a genuine case of causal overdetermination. (Kim 2009a 39)

A genuine case of overdetermination is the situation in which (the objective probability of) a single effect is causally determined by two completely different events, in such a way that if one of them had not been there, the other would have been completely sufficient for the result. Paradigmatic examples of causal overdetermination are two bullets simultaneously hitting a person in the heart and causing his death, and two children simultaneously throwing rocks at a window and breaking it. As some authors have noted,15 special properties and their physical realizers cannot be understood as causally overdetermining their effects. The reason is that they are not independent properties. They are metaphysically related in the sense that, as second order properties, higher level properties are completely determined by the instantiation of their physical supervening bases. So, for example, if we consider that p is the physical realizer of a mental property m, then we cannot understand p and m as causally overdetermining a single effect, say e. This is because although on the nrpist view p and m are different events, they are not really independent. If one of them had not been there to cause the effect, then the other would also be incapable of being there for the same result. That is, neither p could have been there without m, nor m could have been there without p -or another of its physical realizers. What this shows is that cases of causal overdetermination are cases of causal coincidences, in which two or more independent causal processes converge on a single effect in a merely accidental and non-nomological way. It is clear that this is not the case of the higher level properties and their physical realizers. As a result, if the exclusion principle stated above is correct, we have to apply it to all cases involving higher level causation.

15

132

See, for example, Block 1990 159, Kim 1998 52-3, Shapiro & Sober 2007 12, and Shapiro 2010 595-6.

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

Let us return, then, to our example about m causing p*. We said that given the ppcc there is a physical event that is causally sufficient for p*, namely, p, and that given the anti-reductionist tenet of nrp, m ≠ p. We have added that m and p cannot be understood as two causal overdeterminants of p*. Now, insofar as both m and p are considered as sufficient causal determinants of p*, by the exclusion principle we have to exclude either m or p. But excluding p and trying to retain m will not work because the ppcc will reinsert p as causally sufficient for p*. The conclusion of the argument is that, as Kim says, «[t]he mental cause, therefore, must be let go.» (2009a 39) The Interventionist Account of Exclusion

Given that this consequence crucially depends on the so-called exclusion principle that, applied to our example, affirms that p* has no more than one sufficient cause, either m or p, we should examine it in depth in order to determine whether or not nrp must accept it. Kim is clear about this point and thinks that nrp must accept the principle because he believes that «it is virtually an analytic truth with not much content» (2005 51). Moreover, it seems that most authors, whether they accept or reject the overall argument and its results, agree with Kim about the analytical or at least a priori status of the principle. But in recent years some philosophers have argued that the exclusion principle is neither a necessary nor an a priori true statement. For example, List and Menzies analyze the concept of causation in terms of difference-making in order to «show that when causation is understood in this way, Kim’s formulation of the exclusion principle is false.» (2009 476) In a similar vein, Shapiro for example comments: Jaegwon Kim’s causal exclusion argument has rarely been evaluated from an empirical perspective. This is puzzling because its conclusion seems to be making a testable claim about the world: supervenient properties are causally inefficacious. An empirical perspective, however, reveals Kim’s argument to rest on a mistaken conception about how to test whether a property is causally efficacious. (2010 594)

133

Juan Diego Morales Otero

Agreeing with some authors, Shapiro argues «that Kim’s problem of causal exclusion depends on a tendentious conception of causation» (2012 1).16 His idea is that if we adopt an interventionist account of causation, the exclusion principle turns out to be false and we have to accept that its corresponding stunning argument no longer works.17 As we have said in chapter 1, the interventionist theory of causation is based on the central idea that causes should make differences. If c causes e, then an intervention on c will make a difference on e. However, to test whether c is a cause of e, we must not only make an intervention on c, but also hold fixed the common causes of c and e. For example, we can remember that there are many medical cases in which a pathological condition c causes various symptoms occurring one after the other, say e and f. Although we can find a nomological and projectible regularity between e and f, e can be shown as a non-causally relevant factor for the instantiation of f, because an intervention on e that holds their common cause c will make no difference to f. Let us recall that according to the ppcc, there is a physical event that is causally sufficient for p*, namely, p, and that given the antireductionist tenet of nrp, m ≠ p. Following the exclusion principle, Kim’s central idea is that m is completely dispensable because p is causally sufficient for p*. As he asks, on the basis of the above assumptions, «what causal work is there for the mental state to contribute? […] To resist the reductive move of identification is to recognize the existence of something whose causal work is at best superfluous, and nonexistent at worst.» (2007b 262) In the same direction, Papineau affirms that «if the completeness of physics is right, then there is no room left for anything nonphysical to make a difference to physical effects, so anything that does make such a difference must itself be physical.» (2001 8) Kim and Papineau’s idea is that m could have had a causal relevance only if the physical cause p had left some causal work to do; that is, if the ppcc were false and m had the opportunity See, for example, LePore & Loewer 1987, Loewer 2007, Woodward 2008, and List & Menzies 2009. 17 See Shapiro 2010, 2012. See also Shapiro & Sober 2007, Woodward 2008, and List & Menzies 2009. 16

134

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

of making some causal contribution. This means that Kim and Papineau’s claim is that a mental property could be causally relevant only if it could make some causal contribution additional to the causal difference that its realizer makes. As Shapiro comments, Kim appears to wonder why m should count as a cause given that it contributes nothing to the production of p* that p does not already contribute. […] Given that p screens off m from p* -that m contributes nothing in addition to whatever p contributes to the production of p*Kim concludes that m is not a cause of p*. […] In this passage Kim appears to claim that because mental properties supervene on neural properties, mental properties do not have causal powers. For Kim, a mental property can lay claim to being a cause of behavior only if it does something in addition to whatever its neural supervenience base does. (2010 599-600, original italics)

But Shapiro thinks that «Kim’s reasoning rests on a confusion.» (2010 600) His idea, supporting nrp, is that although mental and higher level causal powers are completely determined by the causal powers of their physical realizers and, therefore, can in no way be additional to the latter, they are still real and irreducible. He believes this can be seen from the interventionist approach to causation. Shapiro’s argument18 constitutes an important challenge to the exclusion claim that Kim and Papineau are advocating, particularly because it is a well-grounded argument that attempts to show some empirical consequences that come to falsify the reasoning of these philosophers. According to Shapiro, the interventionist «strategy reveals the right and wrong ways to test whether a mental property is causally efficacious.» (2010 600) From the characterization we have developed above, we can see that to «test whether m is a cause of p*, it is necessary to hold fixed the common cause of m and p* and then intervene on m to see whether there is any change in p*.» (2010 600) But, as Shapiro argues, «Kim suggests that to test whether m causes p*, one must hold fixed p while changing m.» (2010 600) This seems to follow from the fact that, as we have said before, Kim believes that m could have a causal 18

See also Shapiro & Sober 2007.

135

Juan Diego Morales Otero

relevance only if p were not sufficient for P* and had some causal work left to do; it is given that, by the PPCC, P is completely-causally sufficient for P* that M turns out to be dispensable. It seems that if P were not sufficient for P* and M were an additional causal factor for the sufficiency of P*, we could make an intervention on M holding P and show that P* causally depends on M. But the problem is that because «M supervenes on P, changing M while holding P fixed is impossible [!]» (Shapiro 2010 600) As Shapiro says, The right test to perform will hold fixed, not the supervenience base of M, but the common cause that M and P* share. […] P0, not P, is the common cause of M and P*. This means that it is P0 that should be held fixed when testing whether M is a cause of P*. But what happens when P0 is held fixed and M is wiggled? Because changing M is impossible without simultaneously changing M’s supervenience base P, and because P is a cause of P*, a change in M does result in a change in P*. This is evidence that M is a cause of P*. (Shapiro 600-1, original italics)

This is a very interesting idea that attempts to show that because of the necessary connection between M and P, an interventionist criterion for the causal relevance of M can affirm that P* causally depends on M. If you intervene on M and hold fixed the common causes of M and P*, you will have an intervention on P*. In the light of the interventionist approach to causation, these are the criteria we need to show that M is a cause of P*. According to this reasoning, Shapiro claims that Kim’s mistake is based on his confusion about «M’s causing P* [and] M’s making a causal contribution to P* that is additional to the contribution that P makes.» (2010 600) For M to be a cause of P* it is not necessary for M to make a causal contribution additional to that of its realizers. If the argument is correct, it shows us that the exclusion argument fails, particularly because its crucial exclusion principle is an empirically false statement insofar as there are events having more than one sufficient cause occurring at one specific time which are not genuine cases of causal overdetermination. I think that Shapiro’s reasoning shows that the exclusion argument is inconclusive, that this argument is insufficient to conclude that on the NRPist perspective mental and special properties end up being

136

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

epiphenomenal and, therefore, unreal. This is because according to the interventionist criterion for the evaluation of the properties’ causal relevance, the exclusion principle turns out to be an empirically false statement. Despite this, I also think that the interventionist approach is insufficient for showing the irreducibility of the higher level causal powers. The reason is that the causal criterion that this approach has introduced, although adequate for testing the causal relevance of diachronically instantiated properties,19 is insufficient for the evaluation of the causal relevance of the higher level properties insofar as they synchronously supervene on lower level properties that are at least nomologically sufficient for them. This can be seen from the fact that the empirical result that we reach when we make an intervention on a higher level property is compatible with, and so equally supports, three different and competing positions precisely about the causal status of this higher level property: NRP, reductionism, and epiphenomenalism. The empirical result that remains untouched when we understand the higher level property M according to these three different theories and we make an intervention on it is the invariable intervention on P*. Shapiro has made the case for NRP. Let us now take reductionism. This doctrine affirms that (a special property) M is identical to its physical realizer P; it follows that M is metaphysically supervenient on P. The interventionist approach says that in order to test whether M is causally relevant for the instantiation of P* we have to hold fixed the common cause of M and P*, say P0. From here Shapiro’s words follow: «But what happens when P0 is held fixed and M is wiggled? Because changing M is impossible without simultaneously changing M’s supervenience base P, and because P is a cause of P*, a change in M does result in a change in P*. This is evidence that M is a cause of P*.» (Shapiro 600-601, original italics) To make it explicit: on the reductionist perspective the causal powers of M are identical to the causal powers of P simply because

19

For example, for testing the causal relevance of the different properties involved in medical cases, in which an underlying pathological condition causes various symptoms occurring one after the other.

137

Juan Diego Morales Otero

M and P are identical and, for this reason, an intervention on M will produce an intervention on P*. What about epiphenomenalism? In general terms, this is the view that some properties are effects of other properties, but are themselves causally inert. This theory, which Kim takes as a plausible understanding of NRP, claims that M is epiphenomenal and that it has a physical realizer P that is sufficient for it. Although by hypothesis M has no causal relevance for the instantiation of another property,20 we can make an intervention on M and reach the same empirical results of NRP and reductionism: as we have said, because changing M is impossible without simultaneously changing M’s supervenience base P, and because P is a cause of P*, a change in M does result in a change in P*. Is it coherent to say both that M has no causal influence on P* and that a change in m does result in a change in P*? It is coherent on the epiphenomenalist proposal. According to this theory we can change or intervene M because M is a causal effect of the physical antecedents of P, its supervenience base, so if we introduce a change in these causal antecedents we will introduce a change in M. But this intervention on M only works because it is an intervention on P, the supervenience basis of M. As we have said, changing M is impossible without simultaneously changing its supervenience base P. But we have also said that P is a causal factor for P*. This means that an intervention on P will result in an intervention on P*. It follows that if we make an intervention on P we will have both an intervention on M and an intervention on P*. And because this result follows only from the fact that P is a cause of P* and that P is the supervenience base of M, it follows even if we consider that M is causally irrelevant for P*, that is, if we consider that M is epiphenomenal with respect to P*. What this shows is that irrespective of the causal status we give to M (irreducible, reducible, or epiphenomenal), an intervention on M 20

138

Or at least for the instantiation of P* (following Shapiro and Sobers’ idea that epiphenomenalism should be understood as the claim «that some set of property instantiations fails to affect some other set, not that the first set has no effects at all,» (2007 3) we can say that epiphenomenalism about M is the claim that M fails to causally affect the set of physical properties, in particular, the property P*).

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

will invariably produce an intervention on P*. As the interventionist criterion for the causal relevance of M on P* is precisely based on the idea that an intervention on the first should produce an intervention on the second, we have to say that this criterion is insufficient for testing the causal relevance of the higher level properties.21 On the basis of the above considerations, we could conclude that both NRP and the reductive proposals of Kim and Papineau are empirically identical, that the empirical consequences of affirming that higher level properties are irreducible, epiphenomenal, or reducible to their physical bases turns out to be the same. Even Kim suggests this idea when he criticizes Block and Stalnaker’s approach, which he interprets as stating that «the mind-body problem is a scientific research problem on the order of the discovery that water = H20 and that heat = molecular kinetic energy.» (2005 142) He complements this by saying: «As we saw earlier, not even Smart, perhaps the most sanguine of the contemporary materialists, thought that the choice between type physicalism and dualism was a matter to be decided by science.» (2005 142) The following is Smart’s passage, which Kim quotes with approval: …if the issue is… between some form of materialism on the one hand and epiphenomenalism on the other hand, then the issue is not an empirical one. For there is no conceivable experiment which could decide between materialism and epiphenomenalism. (Smart 1959, quoted by Kim 2005 122)

I agree neither with Kim nor with Smart on this point. What I think is that we have to recognize that the method to test the causal relevance of higher level properties, insofar as they synchronously supervene on lower level properties that are at least nomologically sufficient for them, is different from and much more complex than the method to test this relevance for diachronically instantiated properties. In chapter 5, I will describe the empirical method that 21

See Baumgartner 2009, 2010, and Kistler 2013, MS, for related reasons about the interventionist’s failure to account for the causal relevance of the higher level properties.

139

Juan Diego Morales Otero

emergentism has for testing the causal relevance of special properties vis-à-vis their realizers’. Besides the reasons we have to affirm that the interventionist approach is finally unsuccessful in showing the irreducibility of the special properties’ causal powers, I believe with Kim and Papineau that nrp has to face deep philosophical problems when defending its claim about the reality and irreducibility of these causal powers. Although Shapiro shows us that the causal exclusion argument (which as I said is perhaps the most debated argument in recent decades in the field) ends up being inconclusive, there is another important argument that Kim developed in the early 90’s that I think clearly shows the problem with nrp. The Argument of Causal Individuation of Natural Kinds

The argument that we can call the argument of causal individuation of natural kinds has as one of its basic premises the so-called causal inheritance principle, which is accepted by nrp because of its idea that higher level properties (particularly, their causal powers) are determined by their lower physical realizers (particularly, by the lower causal powers of their physical realizers). Kim states the principle as follows: [The Causal Inheritance Principle] If M is instantiated on a given occasion by being realized by P, then the causal powers of this instance of M are identical with (perhaps, a subset of) the causal powers of p. (Kim 1993a 355, see also 1992b 18)

As Shapiro comments, Kim claims the possibility that the causal powers of m are a subset of the causal powers of p because he recognizes that not all the causal powers of a realizer need attach to the realized property. For instance, neural cells might realize mental states, but presumably a mental state does not inherit all the causal powers of the mass or color of neural cells. (2010 601, note 7)

I highlight this point because the latest interpretations of the causal and ontological commitment of nrp argue for a subset account of realization, that is, the idea that, in very general terms, p is a realizer of m

140

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

just in case m’s causal powers are a subset of P’s causal powers.22 The central point, however, is whether the subset strategy is sufficient to vindicate both the difference and irreducibility of the special causal powers. The argument of causal individuation of natural kinds argues that it is not. We have seen that NRP claims that mental and special properties are different from and irreducible to their physical realizers because they are MR by the latter. Despite this, and focusing on mental properties, Kim argues in an almost paradoxical way that «the popular view that psychology constitutes an autonomous special science, a doctrine heavily promoted in the wake of the MR-inspired antireductionist dialectic, may in fact be inconsistent with the real implications of MR.» (Kim 1992b 3-4) Kim’s idea is relatively simple. He accepts the anti-reductive principle whereby a disjunction of heterogeneous kinds is not itself a kind. This is an anti-reductionist principle because it does not allow the reduction of a special property to a disjunction of physical properties that can become empirically coextensive with the former. From this point, we can identify the three basic premises of his argument: first, he claims with the anti-reductionist theorist that special properties are realized by events that belong to completely heterogeneous physical kinds (the MR of the special properties or kinds). Second, Kim argues that the causal powers of the special properties are inherited from the causal powers of their physical realizers (the causal inheritance principle), so that higher level properties cannot have a causal power or relevance additional to that of their physical realizers. The third crucial premise consists in affirming the principle of causal individuation of natural kinds: [Principle of Causal Individuation of Kinds] in science are individuated on the basis of causal powers; that is, objects and events fall under a kind, or share in a property, insofar as they have similar causal powers. (Kim 1992b 17)23 See, for example, Shoemaker 2007, Bennett 2008, Shapiro 2010, and Wilson 2011. 23 For our purposes, it is sufficient to affirm that the causal individuation of kinds is maintained within our nomological and similar nomological worlds. In this sense, we can follow Gillett and «assume a weakened 22

141

Juan Diego Morales Otero

Higher Causal Powers as Identical to Lower Causal Powers

From these premises we can in turn find two basic scenarios for the application of the argument that correspond to two possible readings of the causal inheritance principle; one claiming that the causal powers of the realized properties are identical with the causal powers of their realizers, and the other claiming that the former are a subset of the latter. As we will see, the argument intends to show that in each of these readings higher level causation is unsustainable. Let us then consider a special property S that is MR by different physical properties P1, P2… Pn. In the first reading, we can suppose that the causal powers of each instance of S are identical with the causal powers of its different physical realizers P1, P2… Pn; that is, that on a given occasion the causal powers of S are identical with the causal powers of P1, that on another occasion the causal powers of S are identical with the causal powers of P2, and so on. The reductive conclusion follows naturally from the principle of causal individuation of kinds. We can synthesize the argument as follows: 1. S is MR by P1, P2… Pn (and so P1, P2… Pn are different properties). 2. S is a real natural property that is causally relevant for the instantiation of other properties. 3. The causal powers of every instance of S are identical with the causal powers of its different physical realizers P1, P2… Pn (this reading of the causal inheritance principle). 4. A property is a real natural property iff its instances have similar causal powers (by the principle of causal individuation of kinds). 5. The causal powers of P1, P2… Pn are different (by premises 1 and 4). 6. The causal powers of the S-instances realized by P1 are different from the causal powers of the S-instances realized by P2, and from the causal powers of the S-instances realized by P3, and so on (by 3 and 5). 7. S has no similar causal powers through its different physical realizations (by 6). version of Sydney Shoemaker’s ‘causal theory’ of properties under which a property is individuated by the causal powers it potentially contributes in this world, under certain conditions, to the individuals in which it is instantiated» (2007 196).

142

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

8. S is not a real natural property that can be instantiated through its different physical realizers (by 4 and 7). 9. Thus: S should be reduced to each of its physical realizers (by 2, 3, and 8).

The Subset Account of Realization

Nonetheless, the subset account of realization assumes a different understanding of the causal inheritance principle. It argues that the causal powers of every instantiation of a special property S are a subset of the causal powers of its physical realizers P1, P2… Pn; that is, that on a given occasion the causal powers of S are a subset of the causal powers of P1, that on another occasion the causal powers of S are a subset of the causal powers of P2, and so on. In turn, according to this idea we can have different possible scenarios. Let us analyze the first of them following Audi’s formulation of the subset perspective,24 according to which, A given property, F, is multiply realizable provided there is more than one property that confers all the powers associated with F. Suppose F is individuated by the set of powers {1, 2, 3}, G by the set of powers {1, 2, 3, 4}, and H by the set of powers {1, 2, 3, 5}. F’s individuative powers are a subset of those of G and of H, and so both G and H count as realizers of F. But G confers power 4 and H does not, H confers power 5 and G does not, and so G ≠ H. Thus F is multiply realizable. (Audi 2012a 658)

Using our example and simplifying the situation, let us say that S has the causal power CPA, that P1 has the causal powers CPA and CPB, that P2 has the causal powers CPA and CPC, and so on. In this case, the causal powers of s are not identical to the causal powers of its different physical realizers, but a subset of them. The problem with this way of understanding the subset account of realization is that it finally eliminates the mr of the higher level properties and, in consequence, their irreducibility. This follows from the application of the principle 24

It seems that this is the idea of the subset account that Shoemaker has in 2001 79.

143

Juan Diego Morales Otero

of causal individuation of kinds to the physical properties realizing s. The question is whether under this principle each of these realizers can be understood as a unitary property or really as a conjunction or cluster of different properties. Recalling that according to the principle of causal individuation of kinds, the similarity and difference of the various properties correspond to the similarity and difference of their causal powers. Let us suppose that CPA is the physical causal power to transmit an electrical signal that allows the realization of certain mental property s. Following the version of the subset perspective under consideration, let us assume that the power to transmit this electrical signal is shared by the different physical properties P1, P2… Pn of different physical structures, such as neurons, silicon chips, and others. Now, since B, CPC… CPZ of P1, P2… Pn are lower level causal powers not relevant CP to the fulfillment of the causal role of S, then they must be understood as causal powers that determine other different dynamics of the distinct structures that possess them. For example, we can say that B is the causal power that the property P1 bestows on neurons to CP release certain chemicals, CPC is the causal power that P2 bestows on silicon chips to reflect certain light waves, and so on. But this leads us to affirm that CPA (the causal power to transmit an electrical signal) and CPB (the causal power to release certain chemicals) are different causal powers that can be instantiated separately. By the principle of causal individuation of kinds, it follows that CPA and CPB individuate different properties. And the same claim applies to CPA and CPC, CPA and CPD… CPA and CPZ. If we now consider, as seems natural, that the realizer of a higher level property is the lower level property that is causally relevant and sufficient for the satisfaction of its causal role, then we must say that in each of these cases, the property individuated by CPA, say certain electrical property PE (not P1), should be the real property that realizes S, while the other properties that are causally irrelevant for the satisfaction of this role must be considered irrelevant for its realization. The consequence is that on this version of the subset account of realization, S is not MR; it is realized by a single physical property (PE)

144

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

that is instantiated in different physical structures, such as neurons, silicon chips, and so on. But, can we not claim that the realizers of mental properties are different physical structures such as neurons and silicon chips? This is precisely the fundamental thought that lies at the base of the emergence and development of the very idea regarding the MR of the special properties. Let us recall with Shapiro that Kim asserts the possibility that the causal powers of a mental property are a subset of the causal powers of one of its realizers because, «[f]or instance, neural cells might realize mental states, but presumably a mental state does not inherit all the causal powers of the mass or color of neural cells.» (Shapiro 2010 601, note 7) In this case, we take different physical structures such as neurons, silicon chips, and others, as the multiple realizers of mental properties. On this proposal, the subset account of realization can be articulated in two different ways: first, we can follow Audi’s interpretation and affirm that the various physical structures that realize a special property S are constituted by different causal properties, that there is a subset of these properties that is shared by all the realizers, and that this subset of properties includes those which occupy the causal role of S. Following our example, this interpretation argues that a mental property S is realized by different physical structures such as neurons and silicon chips; that these different realizers have different causal properties (neurons possess some electrical property PE1, certain chemical property PQ1, and certain property of luminance PL1, silicon chips possess PE1, PQ2, and PL2, and so on with respect to the other realizers of S); and that within these different sets of physical causal properties there is a subset that is shared by all the realizers, a physical property which occupies the causal role of S and, therefore, makes it possible for the different physical structures to realize S (in our example, the electric property PE1). The second way of understanding the subset account of realization on the basis of the idea that different physical structures can realize a higher level property s, is arguing that the relevant subset of the causal properties of these structures, that is, the properties which occupy the causal role of s, are not shared by the different physical

145

Juan Diego Morales Otero

structures realizing s. Following our example, on this interpretation the subset account of realization asserts that the mental property S is realized by different physical structures such as neurons, silicon chips, and others; that these different realizers have different causal properties (neurons possess PE1, PQ1, and PL1, silicon chips possess PE2, PQ2, and PL2, and so on with respect to the other realizers of S); and that there is no subset of causal properties that is shared by the different structures realizing S. That is, that in each case the subset of properties whereby each structure realizes S is different: PE1 in the case of neurons, PE2 in the case of silicon chips, and so on. In fact, this is the idea that most philosophers, both supporters and critics of the MR thesis, have had in mind since its appearance. It is in this sense, for example, that Fodor claims: The reason it is unlikely that every natural kind corresponds to a physical natural kind is just that (a) interesting generalizations (e.g., counter-factual supporting generalizations) can often be made about events whose physical descriptions have nothing in common, (b) it is often the case that whether the physical descriptions of the events subsumed by these generalizations have anything in common is, in an obvious sense, entirely irrelevant to the truth of the generalizations, or to their interestingness, or to their degree of confirmation or, indeed, to any of their epistemologically important properties, and (c) the special sciences are very much in the business of making generalizations of this kind. (1974 400, original italics)

The problem with the proposal that different physical structures can be the realizers of higher level properties (that, for example, neurons, silicon chips, and other various physical structures can be the realizers of a mental property S) is that, strictly speaking, we cannot consider that physical structures as such can realize higher level properties: it is the physical structures insofar as they possess certain causally relevant properties. The argument is developed by Shapiro who comments: Some of the properties of the realizers of multiply realizable kinds are relevant to the purpose, activity, or capacity that define the kind and some are not. I asked in the introduction whether two corkscrews, alike in constitution and mechanism but distinct in color, count as

146

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

alternative realizations of the kind corkscrew. I claim that they do not because the only property by which they differ -color- is not a property that contributes to their capacity to remove corks. To say that a kind is multiply realizable is to say that there are different ways to bring about the function that defines the kind. But, if two particulars differ only in properties that do not in any way affect the achievement of the defining capacity of a kind, then there is no reason to say that they are tokens of different realizations of the kind. […] This moral makes sense of why the waiter’s corkscrew and the winged corkscrew do seem to count as multiple realizations of a corkscrew. The waiter’s corkscrew relies on a lever to pry the cork out of the bottle whereas the winged corkscrew uses a rack and two pinions to do the same job. Levers and rack and pinions are different mechanisms that require different manipulations, they are described by different laws, and so on. The causally relevant properties of these two devices differ; a fortiori they qualify as different realizations of a corkscrew. (Shapiro 2000 643-644)

What these considerations are pointing out is the fact that physical structures (with all their physical properties) are not what in the strict sense can or cannot be counted as the physical realizers of higher level properties. What we have to say is that the actual realizers of higher level properties are some of the physical causal properties of these physical structures, that is, the physical properties which are the relevant causal properties for the satisfaction of the higher level causal roles. In this sense, it becomes clear that the realizer of a higher level property S is the lower level property (or properties) that is (are) metaphysically sufficient for the instantiation of S, whether this (these) property(ies) is (are) instantiated or not along with other physical properties in different structures or mechanisms. These considerations have direct negative consequences for the subset account of realization. In the first case, if we say that the physical causal property in virtue of which the higher level causal role is fulfilled is the same throughout the different physical structures, then we have to affirm that the higher level property is not really MR. In our example, if the property PE1, which plays the causal role of (and thus realizes) S, is the physical property that is shared by neurons, silicon

147

Juan Diego Morales Otero

chips, and other possible physical structures that could instantiate S, then we have to affirm that S is not MR. This is the same result that Shapiro reaches when he comments: But why should we suppose that the switch from neurons to silicon chips marks an ‘interesting’ physical difference? If each neuron’s contribution to psychological capacities is solely its transmission of an electrical signal, and if silicon chips contribute to psychological capacities in precisely the same way, then the silicon brain and the neural brain are not distinct realizations of a mind. The difference between neurons and silicon chips, in such a case, is no more interesting relative to their contribution to psychological function than would be the difference between the contributions of neurons gray in color and neurons stained purple. (2000 645)

But the second case is not more favorable. If we claim with Fodor and the tradition that the physical causal property in virtue of which the higher level causal role is fulfilled is different throughout the different physical structures, then we could in principle support the MR of the higher level properties, but only at the cost of their elimination as causally unitary properties. This follows from the fact that this perspective is not compatible with the subset account of realization: on this perspective, the causal powers of a higher level property are not a subset of the causal powers of each of its realizers, but identical to them. This can be seen in the following arguments. Let us return to our example and say that the mental property S is realized on a given occasion by the physical property PE1. And let us recall that CPA (the causal power to transmit an electrical signal) and CPB (the causal power to release certain chemicals) are different causal powers that can be instantiated separately and so, by the principle of causal individuation of kinds, they individuate different properties. Let us say then that the physical property PE1 is individuated by CPA, the causal power in virtue of which it realizes S. Now let us ask about the causal powers of S. The subset account of realization says that the causal powers of S on this occasion are a subset of the causal powers of its realizer PE1. In particular, that the causal powers of S on this occasion are the causal

148

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

powers in virtue of which PE1 occupies the causal role of S, that is, the causal powers of PE1 which are necessary for the fulfillment of this causal role. But we have said that PE1 is precisely individuated by the causal power in virtue of which it occupies the causal role of S (that is, by CPA). It follows that S’s causal powers and PE1’s causal powers are the same. Let us now examine a complementary situation. Suppose that CPA, the causal power to transmit a specific electrical signal, is not sufficient for the fulfillment of the causal role of S; that it should be along with B, the causal power to release certain chemicals, in order to fulfill CP this causal role and, so, to realize S. In this case the realizer of S would be a conjunctive property having two different causal powers, say the property PA&B , which comprises the causal powers CPA and CPB of the physical properties PE1 and PQ1. Let us then consider that the mental property S is realized on a given occasion by the physical property PA&B. But also consider that on this occasion the structure instantiating PA&B also instantiates the property PL1, which has CPC, the causal power to reflect certain light waves that is irrelevant for the realization of s. Let us now ask about the causal powers of S. The subset account of realization says that the causal powers of S on this occasion are a subset of the causal powers of its realizer PA&B. In particular, that the causal powers of S on this occasion are the causal powers in virtue of which PA&B occupies the causal role of S; that is, the causal powers of PA&B which are necessary for the fulfillment of this causal role. But we have said that the causal powers of PA&B (the causal powers that individuate this property) are those in virtue of which it occupies the causal role of S, namely, CPA and CPB. In this situation it follows the same result: S’s causal powers and its realizers’ causal powers are identical. In spite of these results, the argument assumes an idea that we have not clarified: that the causal powers of the physical realizer of a special property are those in virtue of which the realizer occupies the causal role of the special property. Although this idea seems to follow from the fact that the realizer of a special property is the relevant property for the satisfaction of the higher level causal role, and in this sense we can say that the physical realizer of a special property is the property in virtue of which some physical structure can instantiate a

149

Juan Diego Morales Otero

special property, surely the subset account will have to deny it. This is because this perspective affirms that, on a given occasion, the causal powers of a higher level property S are a subset of the causal powers of its physical realizer, namely, the causal powers in virtue of which the physical property fulfills the functional role of and, so, realizes s. It follows that on this account the realizer of S should have more causal powers than those in virtue of which it occupies the causal role of S. Our last scenario involves the realization of a higher level property by a physical conjunctive property, that is, a property that is composed of other different physical properties. In this example, the special property S is realized by the conjunctive property PA&B which comprises the causal powers CPA and CPB of the physical properties PE1 and PQ1. The question we have to face is the following: why do we have to say that S is physically realized by PA&B (with the causal powers A and CPB), and why can’t we say that S is physically realized by the CP property that comprises the causal powers CPA, CPB, and CPC (or CPA, B, CPC, and CPD, or…)? CP What we have here is the application of the principle of causal individuation of kinds not to the individuation of unitary natural causal properties as such, but to the individuation of physical realizers. What this principle says is that a realizer of S is the lower level property individuated by the lower level causal power(s) in virtue of which it occupies the causal role of S. Just as the causal powers of the conjunction of PA&B & PL1 (the causal powers CPA, CPB, and CPC) are not, because they are more than, the causal powers in virtue of which a physical property occupies the causal role of S, we should not consider that this conjunction can be a physical realizer of S. But as the causal powers of the conjunction of PE1 & PQ1 (the causal powers CPA and CPB) are the causal powers in virtue of which a physical realizer occupies the causal role of S, we must consider that this conjunction (namely, PE1 & PQ1 = PA&B) is a physical realizer of S. The rationale is that we cannot consider a conjunctive physical property which has more than the necessary causal powers for the satisfaction of the functional role of S (e.g. the conjunctive property PA&B & PL1, or PA&B & PL1 & PM1 or PA&B & PL1 & PM1 & PN1…) as a realizer of

150

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

S because we can separate the necessary causal powers (CPA and CPB) from the irrelevant causal powers (CPC or CPC & CPD or CPC & CPD & E…) and still have the realization of S. This is the same argumenCP tative move that we made with regard to the competition that can exist between a physical structure and some of its causally relevant properties for the individuation of the realizer that a special property S can have on a given occasion: what is counted as the realizer of S are the relevant causal factors for the instantiation of S just because these causal factors could be instantiated separately (in the same or in other different structures) and would still be realizing S. The result is that the causal powers of the realizer of S and the causal powers of S are the causal powers the realizer needs in order to fulfill the functional role of S; that is, the same. These considerations lead us to affirm that even if the physical realizer of a special property S is constituted by different properties with different causal powers, we should still claim that on the NRPist view the causal powers of S must be identical with the causal powers of each of its realizers. From these ideas we have to apply the argument of causal individuation of natural kinds which takes the first interpretation of the causal inheritance principle (i.e., the claim that the causal powers of each instance of S are identical with the causal powers of each of its realizers), the argument that we have developed above and that concludes that higher level properties cannot be regarded as real natural properties that can be instantiated through different physical realizers and that, in consequence, they should be reduced to the latter. The final result of the argument is that, as Kim has told us, the NRPist claim about the autonomy and irreducibility of the special sciences, «a doctrine heavily promoted in the wake of the MR-inspired antireductionist dialectic, [is] in fact inconsistent with the real implications of MR» (Kim 1992b 3-4). We have reached the idea that a higher level property must maintain a causal unity through its different instances in order to be considered as a real causal factor. There are two mutually exclusive ways to accomplish this: either the special property is instantiated in similar physical structures and maintains its causal unity in virtue

151

Juan Diego Morales Otero

of the causal unity of its physical realizers, or it is MR by different physical properties and maintains its causal unity providing some causal power or relevance that is not given by their physical realizers. This means that if it is true that higher level properties are MR and, therefore, irreducible, then we must accept Kim and Papineau’s idea that these properties should provide an additional causal relevance to that of their realizers. Since this is impossible under the assumption of the metaphysical supervenience of the special properties on their physical realizers, to vindicate the irreducible causal responsibility of the higher level properties we must then accept a different kind of non-causal dependency between these properties and their realizers or microphysical necessary conditions. That is, a type of metaphysical dependency that does not involve metaphysical determination, an emergentist view of the special-physical connection. And since the theory of the higher level properties as second order properties implies the metaphysical supervenience thesis, then we have to replace this understanding. Before concluding this chapter, I would like to make some final remarks about the exclusion argument. We have seen that the purpose of the interventionist approach is to show that the exclusion argument fails, particularly because its crucial exclusion principle is an empirically false statement, insofar as there are crucial situations that can be understood as allowing the existence of events having more than one sufficient cause occurring at one specific time, which are not genuine cases of causal overdetermination. We saw that this approach shows that the exclusion argument is not conclusive. But we also saw that this perspective does not succeed in its attempt to vindicate the causal relevance of the special properties because the interventionist requirements used to demonstrate the exclusion argument’s failure are insufficient for evaluating the causal relevance of the higher level properties. Finally, we have found good reasons to argue that the NRPist approach cannot give an account of the necessary unity that the higher level properties must maintain throughout their different physical realizations in order to be considered as causally irreducible properties. If this argument is sound, we have to conclude that, after all, the exclusion principle can be considered as an empirical claim that can be reformulated in this way:

152

Non-Reductive Physicalism and the Problem of Causation in the Special Sciences

Exclusion principle: No event can have more than one sufficient cause occurring at any given time -unless it is a genuine case of causal overdetermination. As special properties and their physical realizers cannot be understood as causally over-determining their effects because they are not independent properties, the principle should say: a subvenient property F and a supervenient property F* cannot both be causally relevant for (the objective probability of) another property G.

What this principle simply says is that a higher level property cannot have a causal power or relevance different from the causal relevance of its (physical) realizer. But from the above considerations we have to affirm that this principle is just an empirical claim that crucially and necessarily depends on the empirical and contingent assumption, assumed by both NRP and reductive physicalism, about the metaphysical supervenience of the macro-properties on their microphysical realizers or conditions. As we have seen in chapter 2, this assumption is plausibly an empirically false claim. Summary

In this chapter, I have analyzed the problem faced by NRP when defending its claim about the reality and irreducibility of the causal power (influence, responsibility) of mental properties and properties of the special sciences. I also examined Kim’s well-known causal exclusion argument according to which there are only two options for the physicalist with respect to the causal status of special p ­ roperties: either reductionism or epiphenomenalism. On the basis of the interventionist approach to causation, some philosophers argue that the exclusion argument turns out to be incorrect when we consider it from an empirical point of view. Although I agree that the exclusion argument is not conclusive, I have argued that there is another important argument set forth by Kim to show that NRP cannot account for the reality of the causal status of higher level properties: what I call the argument of causal individuation of natural kinds, which affirms that on the NRPist proposal higher level properties cannot maintain a necessary unity through their different physical realizations and, for this reason, cannot be considered as real natural properties or kinds.

153

Juan Diego Morales Otero

These criticisms of the NRPist approach have led some philosophers to support the functional reductive proposal (see Appendix), arguing that to avoid eliminativism about higher level properties, we must reduce them ‘locally’ to each of their physical realizers. The problem is that, on this view, special properties should be finally eliminated because the categories of the special sciences are not truly referring to real special properties but only to sets of multiple and dissimilar basic physical properties in a somehow contingent and indirect way. So far, the overall conclusion is that neither NRP nor the functional reductive proposal seems to have a satisfactory account of the causal status of mental and special properties. On the one hand, NRP cannot explain the necessary unity that the special causal powers must maintain through their different physical realizers. On the other hand, although more metaphysically robust, the functional reductive proposal argues that insofar as the special properties arise from different physical conditions, they finally cannot be real, a consequence that I think is implausible. Since this is an empirical claim, what I find most problematic about the functional reductive proposal is that it seems to assume that any physicalist proposal entails the reduction of the higher level properties and the denial of their MR. But surely a physical world with MR higher level properties is an empirical possibility, not a metaphysically excluded possibility! This complicated picture arises from the single idea of the metaphysical supervenience of the macro-properties on their microphysical realizers or conditions. As I have said, this is an empirical question and it could be that the microphysicalist thesis is correct. If this were the case and we follow the arguments developed so far, we should say that the functional reductive proposal is the more plausible approach for understanding our empirical world. But I have also said that microphysicalism is a contingent thesis with deep conceptual and empirical deficiencies. This means that there are alternative ways to think about the connection between higher and lower physical causal powers within a thoroughly physicalist framework. This is what emergentism does and what I will evaluate in the fifth and final chapter.

154

CHAPTER 5 Emergent, Downward, and Mental Causation

In this chapter, I focus on a careful articulation of the concept of emergent causation and its application to the phenomenon of mental causation. We know that the notion of emergent causation refers to the idea that some irreducible macrophysical properties are causally relevant for the instantiation of both other macro or higher level properties and microphysical properties. Meanwhile, the concept of mental causation refers to the causal relevance or responsibility that mental or psychological properties, such as being in pain, believing that snow is white, and desiring ice cream, have on the subsequent instantiation of other properties, whether psychological, social, biological, or properties of any other organizational level. But why is mental causation so important? As Jaegwon Kim says, «[w]e care about mental causation because we care about human agency» (2007b 257), that is, because we care about our capability to act according to our needs, desires, thoughts, and emotions. To understand, which is maybe even more important than to merely explain, human agency in a recognizable basic physical world is and should, therefore, be a primary and unavoidable enterprise. In this chapter, I present and articulate the initial steps for an emergentist response to this problem.

155

Juan Diego Morales Otero

The Idea of Emergent Causation

Both nrp and emergentism affirm a higher level causal realism whereby some macrophysical properties, in particular the properties of the special sciences, are genuine causal factors for the instantiation of both higher level and basic physical properties. We have seen that the supervenience argument shows that any physicalist perspective, insofar as it argues that special properties are ontologically dependent on the instantiation of microphysical conditions, should claim that special-to-special causation (the causal action of a special property over another special property 1) implies the necessary occurrence of special to physical causation (the causal action of a special property over a basic physical or microphysical property), that is, the existence of downward causation. Emergentism is the non-reductive physicalist proposal that claims that special properties as chemical, biological, mental, and economic properties are irreducible because they are emergent. This approach argues that an emergent property is a causally relevant special organization or relational structure that the constituents of a system can acquire that is not directly determined by the lower level properties and relations these constituents could have when they compose other different systems or structures and that, in consequence, introduces a causal and dynamical difference that is not determined by the causal factors of these constituents. We have synthesized this idea in two general conditions that an emergent property should fulfill: (i) it does not metaphysically supervene on the properties, relations, and laws that hold at the level of the parts of the system, and (ii) it cannot be explained from these lower level conditions, even using contingent and a posteriori compositional principles that govern the aggregation of lower level properties (that is, linear or non-linear mathematical principles that deliver the global property directly and exclusively from the properties of the components2). Let us recall that to avoid unnecessary complications, in many cases I will affirm that a property can cause another property, by which I shall mean that the instantiation of the first can cause the instantiation of the second. I will make it explicit when the context so requires. 2 See chapter 3. 1

156

Emergent, Downward, and Mental Causation

As special properties are understood as emergent properties, then, according to emergentism, special causation must be understood as emergent causation. Following the supervenience argument, this means that this proposal, as McLaughlin affirms, «is committed to the nomological possibility of what has been called ‘downward causation’.» (1992 51) That is, to the idea that our physical world contains physical systems with emergent properties that are causally relevant for the subsequent instantiation of lower level properties, ultimately, microphysical properties from which other emergent properties could arise. It was Donald Campbell who in his 1974 article «“Downward Causation” in Hierarchically Organised Biological Systems» introduced the expression ‘downward causation’. This psychologist and philosopher, concerned primarily with problems of philosophy of biology and evolutionary epistemology, starts from the idea of a hierarchical organization of biological systems and advances the thesis that the higher level properties have some kind of causal influence on the lower level properties through the selection that the former exert on the latter. According to Campbell, we necessarily have to assume, as physicalist theorists, the following two principles: (1) All processes at the higher levels are restrained by and act in conformity to the laws of lower levels, including the levels of subatomic physics. (2) The teleonomic achievements at higher levels require for their implementation specific lower-level mechanisms and processes. Explanation is not complete until these micromechanisms have been specified. (Campbell 1974 180)

These two principles synthesize the implications of any physicalist theory. However, they are not sufficient. Campbell argues that in order to understand the hierarchical organization of nature, we need to add two non-reductive principles: (3) (The emergentist principle) Biological evolution in its meandering exploration of segments of the universe encounters laws, operating as selective systems, which are not described by the laws of physics and inorganic chemistry, and which will not be described by the future substitutes for the present approximations of physics and inorganic chemistry. (4) (Downward causation) Where natural

157

Juan Diego Morales Otero

selection operates through life and death at a higher level of organisation, the laws of the higher-level selective system determine in part the distribution of lower-level events and substances. Description of an intermediate-level phenomenon is not completed by describing its possibility and implementation in lower-level terms. Its presence, prevalence or distribution (all needed for a complete explanation of biological phenomena) will often require reference to laws at a higher level of organisation as well. Paraphrasing Point 1, all processes at the lower levels of a hierarchy are restrained by and act in conformity to the laws of the higher levels. (1974 180)

Although Campbell is focused on the causal relevance that the emergent properties of biological systems can have on the properties of their chemical and physical constituents, we have seen that the supervenience argument shows that the occurrence of downward causation must be present in every instance of emergent causation, from holistic quantum physics to psychology and economics. We know that the idea of downward causation is the idea of a higher level or emergent property having certain causal influence on the subsequent instantiation of lower level properties. Different authors have articulated this kind of causal influence through different concepts that finally point to the same underlying idea. Among these concepts, the most recurrent and important are selection (Campbell 1974, Popper 1978, Van Gulick 1993) and constraint (Kelso 1995, Schröder 1998, Juarrero 1998, 2009, Newsome 2009, Thompson 2007).3 The basic idea of downward causation is that there is certain causal underdeterminacy (and, so, certain causal probabilities which are left open) at the lower physical levels of the natural systems that the higher level properties should constrain and select as a result of the fulfillment of the higher level laws. This means that downward causation consists in the existence of two necessary and sufficient conditions: the necessary causal underdetermination given at the lower levels, and the instantiation of higher

3

158

Dobzhansky (1974) uses the idea of a channeling and Gillett (2006), following Alexander, talks about modulation.

Emergent, Downward, and Mental Causation

level properties and/or laws that constrain and select the lower level courses of events. In this sense, for example, Popper comments: I suggest that downward causation can sometimes at least be explained as selection operating on the randomly fluctuating elementary particles. The randomness of the movements of the elementary particles -often called ‘molecular chaos’- provides, as it were, the opening for the higher-level structure to interfere. A random movement is accepted when it fits into the higher level structure; otherwise it is rejected. (1978 348)

We should note that in the biological-evolutionary context, the idea of selection is understood as the higher level biological influence on the distribution of the different organisms insofar as they have fitness differences. In this case, downward causation functions as the causal influence between some higher level properties of a macro system that is composed of numerous subsystems and the evolution of the distribution of these subsystems. But downward causation operates within all kinds of causal interactions having causally relevant higher level, emergent properties, either with respect to the lower level distribution of sub-systems, or regarding the causal connection between the global properties of a macrophysical system and its particular dynamics over time (which is the case we will illustrate with the following examples). On the other hand, we have the complementary notion of c­ onstraint. This, in turn, follows the notion developed by the dynamic systems theory about the ‘state space’ of a system, that is, about the collection of all the possible states that a system can have through its existence and development (Juarrero 1998 240-1, Thompson 2007 11). In chapter 1, I said that one of the necessary conditions that an instantiation of a property C must meet to be a causally efficacious factor for the instantiation of another property E is that it should increase the probability of the instantiation of E as a nomological matter. The articulation of the concept of downward causation in the context of the dynamic systems approach is based on the idea that a causal factor C increasing the probability of the instantiation of E should be understood in terms of the constraint that C introduces on the state space (the collection of the possible states) of a system S, in such a way that it decreases or

159

Juan Diego Morales Otero

eliminates some possible alternative courses of events of S which are incompatible with E. On this account, causation is primarily understood as a decrease of the different nomological possible courses of events that can occur from certain particular conditions and, complementarily, as the probability increase of some of these nomological possible courses of events as a function of the different constraints acting on the entire nomological possibilities. It follows that, on this approach, a causally deterministic theory should claim the ideal situation that the dynamics of a system (e.g. the entire universe) is completely fixed by its initial conditions in accordance with natural laws. On the other hand, the state space of (the conceptually possible existence of) a completely uncaused system is the collection of all the metaphysically possible states that this system can have. Following the idea that a causal factor or influence is understood as a causal constraint of the state space of a system, the development or evolution of a system can have both lower level and higher level constraints. A lower level causal influence is the kind of causal constraint that a lower level property has on the development of the system; that is, the causal constraint that the instantiation of this lower level property has on the resulting development of the system. The same holds with regard to the higher level properties. The idea of downward causation is that the lower level constraints are nomologically insufficient for the complete fixation of (and, in this sense, that these lower level constraints under-determine) the state space of an emergent system that is further constrained by some higher level properties. In this sense, Newsome, focusing on the inter-level connection between the cells and their physicochemical constituents, affirms: As far as we know, nothing about the life of unicellular organisms violates the laws of physics or the chemical laws that govern the behavior of macromolecules. The cell cannot behave in any way that is not permitted by the lower levels of organization of its constituent parts; the behavior of the cell is thus constrained but not determined by the lower levels. (Newsome 2009 59)

160

Emergent, Downward, and Mental Causation

Let us analyze and articulate in a precise way the structure and functioning of emergent and downward causation through the following situation. Suppose in a simplified and formal way the following lower level, say neurological laws: (i) Pr(P1⇨P2) = 0.5 (that is, the probability of an instantiation of P1 causing an instantiation of P2 is 0.5); (ii) Pr(P1⇨P5) = 0.5; (iii) Pr(P2⇨P3) = 0.5; (iv) Pr(P2⇨P6) = 0.5; (v) Pr(P3⇨P4) = 0.5; and (vi) Pr(P3⇨P7) = 0.5. The following diagram shows the different causal possibilities admitted by these neurological laws: P4 P3 P7

P2 P1

P6 P5

t2

t1

t3

t4

Diagram 1. The different causal possibilities admitted by neurological laws.

Now, let us suppose that P1⇨P2 (we can say, the neurological property that consists in the instantiation of the causal process P1⇨P2; let us call it P1→2) is the neurological basis for the higher level, mental property M1, and that P3⇨P4 (P3→4) is the neurological basis for the mental state M2. In graphic terms we would have something like the following: M2

is the basis of

is the basis of

M1

P1

P2

P4

P6

P5

t1

P3

t2

P7

t3

t4

Diagram 2. The neurological basis for M1 and M2.

t

161

Juan Diego Morales Otero

If this is so, from a purely lower level point of view (that is, from the instantiation of its neurological basis P1→2), the instantiation of M1 could not yet change the probabilities of the lower level events and it will be nomologically compatible with the subsequent instantiation of different courses of events which are not necessarily mental; for example, with P2⇨P6. Precisely, this is a consequence of the existence of some indeterminacy at the lower levels: from a single neurological state (say, P2) several different courses of events can follow (say, either P3 or P6). But the probability of the arrangement and occurrence of the various events changes if the higher level properties M1 and M2 are governed by higher level laws that constrain and select the possibilities given at the lower basal level. Let us then suppose the higher level, psychological law: Pr(M1⇨M2) = 1.0. In this case, if we have an instantiation of M1 that emerges from P1→2, and we have the fulfillment of the psychological law, then we will necessarily nomologically have the lower causal chain P1⇨P2⇨P3⇨P4.4 In this case we would have the following diagram:

P1

wn

wa

P

rd

ca

us

al

co

ns

tra

in

t

Particle source

2

P3

P

P

P

t3

t4

5

t1

M2

causes do

emergence

emergence

M1

t2

P4

6

7

Diagram 3. Emergence and downward causation of M1 and M2. 4

162

As we will see with the examples of the quantum entanglement and the emergence of the unitary experience of pain, it is not necessary that the higher laws select a single course of lower level events. In order to have emergent and downward causation it is necessary, however, that the higher level laws introduce (at least some minimal) causal constraints that we cannot find at the lower levels.

Emergent, Downward, and Mental Causation

In this kind of circumstances we can ask the reductionist theorist why the causal process P1⇨P2⇨P3⇨P4 (and thus M1⇨M2) is instantiated, and no other different lower processes which are compatible with the lower level laws which are present, as for example P1⇨P2⇨P6⇨P8, P1⇨P2⇨P6⇨P9, or P1⇨P2⇨P3⇨P7, or other different ones. The nonreductive, emergentist answer is that the lower neuronal possibilities governed by the lower neurological laws are constrained and selected by the higher level law M1⇨M2, which increases the probability of the instantiation of P1⇨P2⇨P3⇨P4 over the others. Whether this kind of macro-causation ends up being a fact of our world or not is essentially an empirical question and consists in the existence of two conditions: the necessary underdetermination given at the lower levels, and the existence of the higher level properties and/ or laws that constrain and select the lower level courses of events. If this kind of phenomenon constitutes a fact of our world, there must be multiple levels of organization with their own laws and causal influences that end up complementing one another. As stated by many theorists, the higher level laws do not contradict, nor change, nor violate the lower ones.5 For this reason, it is emphasized that not only the special laws must conform to the lower ones, but that the laws of the lower levels must act in accordance with those of the higher levels. But the mere assertion of the existence of multiple causal laws and levels is not enough. To understand the relationship and dependency of the higher level laws vis-à-vis the lower ones, we must remember that the former only function as higher level constraints of the latter and, therefore, can only exist while the latter take place; without the existence of lower level laws involving different degrees of freedom or underdeterminacy, the occurrence of higher level laws acting as their constraints is impossible. The Principle of Physical Causal Closure of Emergentism

Let us now analyze the way in which emergentism understands the principle of physical causal closure. In chapter 4, I said that the most common formulation of the principle is as follows: 5

See, for example, Anderson 1972 222, Campbell 1974 180, Sperry 1986 268, Van Gulick 1993 252, Gell-Mann 1994 112, Dretske 2004 167, Gillett 2006 276, and Newsome 2009 59.

163

Juan Diego Morales Otero

PPCC: If a physical event has a sufficient cause at t, it has a physical event as sufficient cause at t.6

The basic idea behind this formulation is the assertion that the causal relevance of the physical properties is sufficient for the causal determination of the (objective probability of the) instantiation of any physical property. Now, the scope of the PPCC depends on both the meaning and the scope we give to the concept of the physical. As both NRP and reductive physicalism are different kinds of microphysicalism,7 that is, proposals claiming their physicalist commitments on the basis of the idea that the higher level and macro properties are metaphysically supervenient on the properties and relations of the ultimate and simplest microphysical elements, then these proposals should argue that the PPCC must be understood in terms of microphysical properties. In fact, this is the most common interpretation of the principle, which can be called the principle of microphysical causal closure (PMCC) and can be stated as follows: PMCC: If a microphysical event has a sufficient cause at t, it has a microphysical event as sufficient cause at t.

This is the principle that, for example, Kim defends and Gillett criticizes: Notice that neither the mental nor the biological domain is causally closed; there are mental and biological events whose causes are not themselves mental or biological events. A trauma to the head can cause the loss of consciousness and exposure to intense radiation can cause cells to mutate. (Kim 2005 16-7) The same is true of macrolevel physics and chemistry. It is only when we reach the fundamental level of microphysics that we are likely to get a causally closed domain. (Kim 2005 65) In its reference to ‘ laws of physics’, the Completeness of Physics is intended to concern simple laws that directly refer only to microphysical entities and which are discovered by studying isolated, simple systems of microphysical entities. Quantum mechanical theories are 6 7

164

See Kim 2009a 38. See chapter 3.

Emergent, Downward, and Mental Causation

often taken to supply such laws and the Completeness of Physics implies that these laws suffice to determine all microphysical events, whether these events are located in simple systems or complex aggregates, in so far as they are determined. (Gillett 2006 277)

But in chapter 2 we saw that microphysicalism is a contingent thesis with deep conceptual and empirical deficiencies: on the one hand, it allows us neither to fix nor to understand the use we make of the concept of the physical. For this reason, I said that all physical entities should meet some necessary conditions, and that it is because holistic physical, chemical, biological, and neurophysiological properties meet these conditions that they are physical entities in their own right. Moreover, on the other hand, microphysicalism seems to contradict results coming from both the physical science itself, as when we talk about holistic or systemic physical properties not reducible to their constituent conditions, and the special sciences’ greatly successful theories and experiments that provide explanations and predictions which, as far as we know, are not reducible to the microphysical laws and explanations from which they must arise. It follows that the failure of the PMCC is strongly supported by the current results of the different sciences. These conceptual and empirical reasons lead us to claim that the PPCC should be re-interpreted in macrophysical or emergentist terms, that is, as stating that PPCC: If a (micro or macro) physical event has a sufficient cause at t, it has a (micro or macro) physical event as sufficient cause at t.

Clarifying the connection between emergentism and physicalism in the context of the (exclusion) causal argument for physicalism that is based on the PPCC, Papineau suggests that this formulation of the principle in macrophysical terms could be its most plausible understanding:8 What about the argumentative rationale for Physicalism? Would this survive the existence of emergent Broad-laws? Again, there seems no immediate reason why Broad-laws should stop us arguing for Physicalism. Maybe they would if the only argument for Physicalism 8

See also Silberstein 2001 86-7 for an account of the PPCC in macrophysical terms.

165

Juan Diego Morales Otero

somehow proceeded via a demonstration that all physical laws supervene on microphysical ones. However the causal argument for Physicalism sketched above makes no such assumption. Rather it hinges on the causal completeness of the physical realm, which says nothing about microphysics, but only that every physical effect has a fully sufficient physical cause. Broad-laws seem in perfectly good accord with this assumption. True, such laws would mean that some physical effects essentially result from macroscopic physical causes in ways unpredictable on the basis of microphysical laws alone. But for all that, they are still physical effects with sufficient (macro)physical causes. And so the causal argument will still tell us that any mental [or special] causes of those physical effects cannot be metaphysically distinct from those (macro)physical causes. (Papineau 2008 136-7)

For emergentism, then, the ppcc affirms that all the physical events, whether microphysical, macrophysical-resultant, or macrophysical-emergent, are fully causally fixed by micro or macrophysical events. Just as on this perspective not only holistic physical, chemical, biological, and neurophysiological, but also psychological, economic, and social events can be understood as physical events insofar as they meet the necessary conditions for (macro) physicality (see chapter 2), so should the ppcc claim that for the causal determination of every physical event the conjunctive action of the different properties at the different levels of organization working as necessary multilevel constraints is necessary. Emergent and Lower Level Causal Powers

Let us now focus on the relationship between the higher level and the lower level causal powers. Following our example, let us ask about the connection between the causal powers of M1 and the causal powers of its neurological realizer P1→2.9 Let us recall the reductionist principle that Kim has postulated to understand this connection: 9

166

Henceforth, I will use the concept of realizer (and, therefore, the concept of realization) in an emergentist sense. In this sense, a realizer of a higher level property is the lower level (ultimately microphysical) basis from which the higher level property emerges, that is, the lower level

Emergent, Downward, and Mental Causation

[The Causal Inheritance Principle] If M is instantiated on a given occasion by being realized by P, then the causal powers of this instance of M are identical with (perhaps, a subset of) the causal powers of P. (Kim 1993a 355, see also 1992b 18)

In chapter 4, we saw that the argument of causal individuation of natural kinds shows that, against the subset account of realization, the NRPist perspective should claim that the causal powers of a higher level property are identical with the causal powers of each of its realizers. In a nutshell, the idea is that NRP has to say that the physical causal powers in virtue of which a physical realizer occupies the functional role of a special property on a given occasion (the causal powers that the subset account considers that are a subset of the entire set of this realizer’s causal powers) are the causal powers that individuate both the higher level realized and the lower level realizer properties. On the one hand, the physical causal powers in virtue of which a physical realizer occupies the functional role of a special property on a given occasion individuate the higher level property because, as the subset account claims, these are the necessary and sufficient causal powers for the fulfillment of the functional role of the higher level property. That these are necessary causal powers for the higher level property follows from the fact that they are the minimal causal powers that a lower level property can have to occupy the higher level causal role. And that these causal powers are sufficient for the higher level property follows from the supervenience theory of functionalism, that is, from the idea that the higher level powers are metaphysically supervenient on, and determined by, the realizer’s causal powers. And here we find a crucial difference between NRP and emergentism: as we will see below, emergentism argues that the higher level causal powers, although metaphysically dependent on, are additional to and so different from the lower level causal powers from which they emerge; this means that on this proposal, basis that determines the existence and instantiation of the higher level property only along with certain trans-ordinal laws.

167

Juan Diego Morales Otero

the lower level causal powers are necessary but insufficient for the instantiation of the higher level, emergent property. On the other hand, NRP is committed to the idea that the physical causal powers in virtue of which a physical realizer occupies the functional role of a special property on a given occasion individuate the lower level property because, again, these are the necessary and sufficient causal powers for having the lower level realizer property. That these are necessary for the realizer property follows from the fact that these are the causal powers that the realizer property should have (that is, the minimal causal powers) for the fulfillment of the functional role of the higher level property. That these causal powers are sufficient for the realizer property follows from the idea that we can separate these causal powers from the causal powers that are irrelevant for the fulfillment of the higher level functional role, and still have the realization of S; as we have the realization of S from these separable causal powers, we have its realizer, so these separable causal powers are sufficient for the realizer. This means that on the NRPist perspective, the causal powers of a higher level property are identical with the causal powers of each of its realizers and that, from this premise, the argument of causal individuation of natural kinds concludes that higher level properties cannot be regarded as real natural properties. And here we find the crucial difference between NRP and emergentism. For the emergentist theory, the causal powers that a higher level property has on a given occasion are different from the causal powers of its realizer because they are not within and so, in a special and important sense, they are additional to the causal powers of the latter. It is for this reason that the causal powers of the higher level property cannot be deduced from or reduced to the causal powers of its realizer. Let us recall that the realizer of M1, that is, P1→2, can cause alternative causal chains of neurological events in accordance with the neurological causal laws that govern it; for example, P3⇨P4, P3⇨P7, or P6⇨P8. More specifically, P1→2 can cause at t3 either the instantiation of P3 with a probability of 0.5, or the instantiation of P6 with a probability of 0.5. This means that the causal power of the neurological property that realizes M1, that is, P1→2, is the causal power that this property has

168

Emergent, Downward, and Mental Causation

to cause either the instantiation of P3 with a probability of 0.5 or the instantiation of P6 with a probability of 0.5. But the causal power of the emergent property M1 is quite different. We have said that in our example M1 is governed by an emergent law that affirms that the instantiation of this property should cause as a nomological matter the subsequent instantiation of M2. We can say, then, that the causal power of M1 is the power to cause the instantiation of M2 with a probability of 1.0. But we have also said that on a given occasion the property M1 emerges from P1→2. This implies that there is a trans-ordinal law with the following form: if at the lower level we have the instantiation of P1→2, then at the higher level we should synchronously have the emergence of the property M1 with such and such causal powers. As a higher level property, M1 can have different bases of emergence (say P1→2, P11→12, P21→22…) which are capable of functioning as the lower level mechanisms necessary for its instantiation and for the fulfillment of the higher level laws that govern it; this implies the possible existence of different trans-ordinal laws connecting M1 with its multiple bases of emergence (Kim 2009b 99-100). And this also means that M1 is multiply realizable by different lower level properties. However, whether in fact M1 is multiply realized in our world is an empirical question that should be the object of the specific empirical sciences investigating its laws, both about its own organizational level (mental) and about its connections with other levels (Shapiro 2008 514). We have said that the causal power of P1→2 is the power that this property has to cause either the instantiation of P3 with a probability of 0.5 or the instantiation of P6 with a probability of 0.5, and that M1 has the power to cause the instantiation of M2 with a probability of 1.0. But M1 has not only this causal power; it must have the causal power to constrain and select the different lower level causal chains because the fulfillment of the higher level, emergent, causal law is not compatible with all the nomologically possible lower level events. Furthermore, these causal powers are not separable. M1 has the power to cause the instantiation of M2 in virtue of and because it has the causal power to constrain and select the different possible lower level causal chains. That is, as the supervenience argument states, M1 has the power to cause the instantiation of M2only through the power to cause the instantiation of the (possibly

169

Juan Diego Morales Otero

different) lower level property (properties) that is (are) the lower level basis (bases) or realizer(s) of M2. On a given occasion, then, when M1 is realized by P1→2, M1 must have the causal power to constrain and select the different lower level possibilities that the instantiation of P1→2 leaves open (that is, P3⇨P4, P3⇨P7, P6⇨P8, P6⇨P9) because the fulfillment of the emergent causal law Pr(M1⇨M2) = 1.0 is only compatible with one of these possible lower level causal chains, namely, with the causal chain that realizes M2, that is, P3⇨P4, while it is incompatible with the other: P3⇨P7, P6⇨P8, P6⇨P9. To put it in other words, M1 has the causal power to decrease the probability that P1→2 has of causing the instantiation of P6 and, complementarily, the causal power to increase the probability that P1→2 has of causing P3 so that the higher level causal process can be instantiated. In general terms, we can say that the causal power that a higher level property has on a given occasion is different from the causal power of its realizer because the causal power of the second allows a multiplicity of different causal chains, some of which are nomologically incompatible with the law that governs the higher level property, so the causal power of the latter must function as a constraint upon this multiplicity that causes the instantiation of the necessary realizer that allows the fulfillment of the higher level law. But these causal powers are not simply different. They are metaphysically related for the reason that the causal power of the emergent property is metaphysically dependent on the instantiation of the lower causal power: without the existence of multiple possible courses of lower level events that the causal power of the realizer leaves open, the action of a higher level causal power that can only function as a causal constraint and selector of these different possible chains of lower level events is impossible. Testing the Causal Relevance of Higher Level Properties

We have seen that the interventionist approach to causation affirms that to test whether B is a cause of C, we must not only make an intervention on B and analyze its connection to C, but also hold fixed the common causes of B and C. To illustrate this requirement, we find a recurrent example in those medical cases in which a pathological condition A causes various symptoms occurring one after the other,

170

Emergent, Downward, and Mental Causation

say B and C. Although we can find a nomological and projectible regularity between B and C, B can be shown to be a non-causally relevant factor for the instantiation of C, simply because an intervention on B that holds fixed their common cause A will make no difference to C. But the interventionist account applies this test not only to the evaluation of the causal connection of properties that are diachronically related as C, E, and F, but also to the evaluation of the causal relevance of properties that synchronously supervene on lower level properties that are at least nomologically sufficient for them, that is, to the evaluation of the causal relevance of the higher level properties. Nonetheless, this test was shown to be inefficacious to evaluate the causal relevance of the higher level properties because the empirical results of its application (that consist in the invariable intervention on the subsequent instantiation of a certain lower level property) end up being compatible with, and so equally supporting, different positions precisely on the status of the causal powers of those higher level properties. We saw that these results are compatible with NRP, reductionism, and epiphenomenalism. Even emergentism, which claims that the causal powers of the emergent properties are different from and additional to the causal powers of their lower level bases of emergence (of their realizers), is compatible with these empirical results. Let us return to the example set forth in chapter 4 and consider that M is emergently realized by P and that we make an intervention on M holding fixed the common causes of M and one of its alleged effects, P*. Just as the only possible way to intervene on M is intervening on P, and because within the emergentist setup P is synchronously nomologically sufficient for M, and M is causally nomologically sufficient for the instantiation of another emergent property M* that is realized by the lower level property P*, so too an intervention on P will result in a change in the value of P*, that is the same result that was expected by NRP, reductionism, and epiphenomenalism in their own accounts. We have seen that this shows that irrespective of the causal status we give to M (irreducible, reducible, epiphenomenal, or emergent), an intervention on M that holds fixed the common causes of M and P* will invariably produce a change in the value of P*. It follows that this

171

Juan Diego Morales Otero

kind of intervention cannot evaluate the causal relevance of higher level properties. I have suggested that we have to recognize that the method to test the causal relevance of higher level properties, insofar as they synchronously supervene on lower level properties that are at least nomologically sufficient for them, is different from and much more complex than the method to test this relevance for diachronically instantiated properties. The basic emergentist idea is that the causal relevance of a higher level property H of a macrophysical system S1 can be empirically tested from the evaluation of the causal powers that (the lower level properties of) the various components C1, C2… Cn of S1 have across the different systems S2, S3... Sn that can be constituted by these components. On the one hand, if the causal powers that these components have throughout the different systems are (at least in principle10) sufficient to build, account for, deduce, or predict the causal power of H, then H will be an aggregative property of S1 and, therefore, a property that is reducible to the lower level properties and relations of the constituents C1, C2… Cn of S1. A clear example of an aggregative property is the global mass MG of a system S1, which is a result of the linear addition of the masses M1, M2... Mn of its constituents C1, C2… Cn. This implies that the causal power of the resultant mass MG is a causal aggregative power that can be known from the causal powers of the masses M1, M2... Mn that each of the constituents C1, C2… Cn has in other systems different from S1.11 On the other hand, if the causal powers that the lower level properties of the components C1, C2… Cn of S1 have throughout the different 10 11

172

See chapter 3. We should note that the compositional principles that allow us to connect the values and properties of parts in order to yield the values and properties of wholes in a (linear or nonlinear – see chapter 3) mathematical way are both contingent and a posteriori, so the different aggregative properties of the macrophysical systems should be considered as contingent products, that is, as aggregative, reducible but metaphysically contingent and metaphysically non-supervenient on the properties and relations of the systems’ constituents. This implies that, as Hüttemann and Papineau (2005) argue, even the simple and common properties such as mass and density cannot be properly accounted for within the microphysicalist metaphysical proposal of NRP and reductive physicalism, at least in their customary formulations.

Emergent, Downward, and Mental Causation

systems S2, S3... Sn are insufficient to build, calculate, or explain the causal powers of the global property H of S1, then H has causal powers which are irreducible to the causal powers of the constituents of S1 and, therefore, H is a global emergent property. Let us return to the example given above. We can say that there is a psychological system S1 constituted by the neurological component C1 having at t1 the property P1 and the neurological component C2 having at t2 either the property P2 or the property P5. We have also said that this system has the global mental property M1 throughout the times t1 and t2. Let us consider the causal powers that the properties P1, P2, and P5 of the constituents C1 and C2 of the system S1 can have on a given occasion: we saw that P1 has the power to cause at t2 either P2 or P5 with the same probability, and that P2 has the power to cause at t3 either P3 or P6 also with the same probability; although irrelevant for the example, we can add that P5 has the power to cause at t3 the instantiation of Pn. This simply means that we have evaluated the causal powers that the properties P1, P2, and P5 of the components C1 and C2 can have within the different systems S2, S3... Sn that C1 and C2 can constitute (wherein these components are related to other components or to the same components in different organizations), and we have found that P1 has the power to cause at t2 either P2 or P5 with the same probability, that P2 has the power to cause at t3 either P3 or P6 also with the same probability, and that P5 has the power to cause at t3 the instantiation of Pn. If M1 were an aggregative property of S1 then its causal powers should be deducible, explainable, or predictable in a logical or mathematical way from the causal powers that the properties of the constituents C1 and C2 can possess. We have said that S1 can have the property M1 when its constituents instantiate the causal process P1⇨P2 but not when they instantiate the causal process P1⇨P5 and that, for this reason, the neurological property that consist in the instantiation of P1⇨P2 (i.e., the lower level property that we called P1→2) is the realizer of M1 on this occasion. This means that when S1 instantiates M1 its components C1 and C2 should instantiate the causal process P1⇨P2, and that the aggregative or resultant causal powers that S1 has on this occasion are the resultant causal powers of this process, that is, the resultant causal powers of the property P1→2. It follows that if M1 is an aggregative property of S1,

173

Juan Diego Morales Otero

then its causal powers should be buildable, deducible, explainable, or predictable from the causal powers of its realizer P1→2. As we have said before, the causal power of the property P1→2 is the power to cause either the instantiation of P3 with a probability of 0.5 or the instantiation of P6 with a probability of 0.5, and the causal power of M1 is the causal power to constrain and select the different lower level possibilities that the instantiation of P1→2 leaves open in order to fulfill the higher level law Pr(M1⇨M2) = 1.0, that is, the power to cause the instantiation of P3 with a probability of 1.0. But there is no way (to construct a mathematical function) to derive or calculate the power to cause the instantiation of P3 with a probability of 1.0 from the power to cause either the instantiation of P3 with a probability of 0.5 or the instantiation of P6 with a probability of 0.5 without assuming contextual or holistic information, that is, assuming that these causal powers apply when certain particular macrophysical conditions are met. It is clear, then, that the causal power of M1 cannot be known or solely derived from the knowledge that we can have of the causal powers that the constituents C1 and C2 of S1 have in other systems in which they behave in a different and less determined way, enabling different chains of events which are inconsistent with the fulfillment of the emergent law that governs M1. So, we can say that, in general terms, the way to make an empirical evaluation of the causal relevance of the global or higher level properties of a macrophysical system is to decompose the system into subsystems, test the causal powers that these subsystems have within other systems, and analyze whether the causal powers that these subsystems have in other systems can be logically or mathematically added or computed in order to build, understand, and predict the global causal power of the first system.12 Let us now examine a particular case that gives empirical support to the idea of the existence of causally relevant macrophysical properties that cannot be understood according to the reductionist perspective (that is, on the assumption that their causal powers are aggregative 12

174

For the analysis of this empirical method from a general hierarchical and emergentist perspective of science see Bechtel & Richardson 2010.

Emergent, Downward, and Mental Causation

and reducible to the causal powers of the lower level constituents) and that, therefore, in order to understand it from a dynamic and causal point of view, we must articulate and explain it applying the model of emergent causation that we have developed here. This example is particularly important for our conceptual understanding of the relationship between the different levels of organization of our physical world because it allows us to clarify and articulate in a very concrete way the nature and structure of the phenomenon of emergent and downward causation. Emergent Causation at the Quantum Level

One of the most recurrent examples that different theorists have used to analyze the seeming failure of the microphysical supervenience theory is the phenomenon of the quantum states of entanglement.13 Furthermore, we agree with Silberstein and McGeever that this particular phenomenon of «quantum mechanics (QM) provides the most conclusive evidence for the existence of ontological emergence.» (1999 187) An entangled quantum state is a state of a physical system characterized by having a state vector (or wave function) that is not factorizable in terms of the state vectors of its components; that is, a quantum state which cannot be deduced or explained from the knowledge of its components and that, in this sense, contains information beyond that of its components. Quantum entanglement is an empirically verified (Aspect, Grangier & Roger 1981) and ubiquitous phenomenon of our world. It is not only present in physical systems composed of elementary particles but, as Papineau says, also «present in systems comprising basic physical persisting objects, like atoms and molecules.» (2008 146) Furthermore, it is not only ubiquitous but indispensable for our current understanding of the physical world. Schrödinger, who introduced the notion of entanglement, said: «I would not call [entanglement] one

13

See, for example, Teller 1986, Healy 1991, Maudlin 1998, Esfeld 1999, Silberstein & McGeever 1999, Hüttemann 2005, Papineau 2008, Schaffer 2010, and Bernal 2012.

175

Juan Diego Morales Otero

but rather the characteristic trait of quantum mechanics, the one that enforces its entire departure from classical lines of thought.» (1935 555) Although all theorists agree about the empirical success that quantum mechanics and its notion of quantum states of entanglement have in the statistical dynamics of the quantum realm, there is a controversy about the possible physical principle that should be at the basis of and explain the empirical results; particularly, the empirical results that seem to show some kind of ‘non-local’ correlations between spatially separated particles. In one of the simplest and most common examples, we can have a pair of electrons with a conjoint spin S = 0 which after interacting are sent off in opposite directions. Electrons are particles with spin S = ±½, which means that their spin will always show one of two possible values, namely, +½ or -½. We can suppose that these electrons are a light year apart. On the basis of the holistic property of the system of having a spin S = 0, quantum theory successfully predicts that the behavior of the electrons will show a necessary correlation: if the left hand electron is shown to have S=+½, the right hand one will be shown to have S=-½, and vice versa. As Ismael and Schaffer comment, The type of coordinated randomness borne by the spatially separated components of a system in an entangled state is not just a straightforward mathematical consequence of the quantum formalism, but moreover is itself an empirically verified phenomenon. The difficulties only arise in trying to arrive at a physical understanding of how entangled components manage to exhibit such coordinated randomness. (Ismael & Schaffer forth 8-9)

These authors examine three different physical principles that have been postulated as the basis and explanation of the empirical results; they call them incompleteness, nonlocality, and nonseparability. Incompleteness says that quantum mechanics is an incomplete theory because it does not describe or take into account some intrinsic or local variables (some hidden variables) that the particles should have throughout the described physical process and explain the correlations in question. What lies behind this idea is the denial of the existence of the quantum entanglement phenomenon, that is, the denial of

176

Emergent, Downward, and Mental Causation

the idea that the wave function of some quantum systems cannot be deduced or explained from the wave function of the constituents of these systems. For this perspective, as microphysicalism affirms, all global physical properties must be explained from the local properties and relations of the constituents. This is precisely the insight of EPR (Einstein, Podolsky & Rosen 1935). The problem with the principle of incompleteness is that it goes against the empirically confirmed (Aspect, Grangier & Roger 1981) Bell’s Theorem (1964), which shows that a hidden variable theory can only match the statistical predictions of quantum mechanics if it assumes some form of non-local influence or correlation that is not completely determined by the local properties and connections that the particles can have. This leads us to the so called nonlocality principle, which I prefer to call action at a distance in order to differentiate it from the principle of nonseparability that also affirms a certain kind of non-locality. This principle intends to explain the dynamics of the entangled systems in a particular way; according to this proposal, there are some additional parameters that quantum mechanics has not taken into account (some hidden variables) allowing the functioning of certain mechanism whereby «[t]he measure result on some component(s) of an entangled system causes [in an instantaneous form] the other component(s) (no matter how far distant) to go into the coordinated state.» (Ismael & Schaffer forth 11)14 The primary problem with the idea that the nonlocal correlations of the entangled systems result from some instantaneous causal connections between spatially separated particles that are in the entangled state, and also the reason why this proposal has been accepted by few theorists, is that it violates the principle of relativity that says that no body, signal, or causal influence can travel faster than light. But its problem may be not only empirical. If we examine the situation following the analysis of the concept of causation that I developed in chapter 1, we have to say that the nonlocal correlations of the entangled systems cannot result from some instantaneous causal interactions between the spatially separated particles that are 14

See, for example, Bohm 1952 and Dürr, Goldstein & Zanghì 1992.

177

Juan Diego Morales Otero

in an entangled state because such correlations could not classify as causal. This is because the supposed causal interaction between the spatially separated particles will not meet some causally necessary requirements: in particular, the condition of spatiotemporal connectivity that is necessary for the transference of energy or another conserved physical quantity between the causally related events, and the requirement of a diachronic interaction between these events, which is what allows the concept of causation to explain the sequential order of the world. We can say, then, that the principle of action at a distance has deep empirical and conceptual problems that should be resolved before we can consider it as a plausible account of the quantum entangled correlations. But perhaps this problematic proposal is unnecessary, since we have a solution that seems to be both empirically and conceptually superior. This leads us to the most accepted physical principle for the explanation of the nonlocal quantum correlations. Ismael and Schaffer call it nonseparability, and I would call it the principle of macrophysicalism or emergentism: Nonseparability says that the coordinated randomness of entangled systems corresponds to a failure of mereological supervenience. The whole system has an intrinsic state that fails to supervene on the intrinsic states of its proper parts plus their spatial relations: Nonseparability: Entities a and b are nonseparable if and only if fixing the intrinsic state of a, the intrinsic state of b, and the spatial relations between a and b fails to fix the intrinsic state of a+b. (Ismael & Schaffer forth 11)

And this is precisely the idea of emergence. Let us recall our characterization of an emergent property as the property that (i) does not metaphysically supervene on the properties, relations, and laws that hold at the level of the parts of the system, and (ii) cannot be explained from these lower level conditions, even using contingent and a posteriori compositional principles that govern the aggregation of lower level properties (that is, linear or non-linear mathematical principles that deliver the global property directly and exclusively from the properties of the components). This perspective excludes

178

Emergent, Downward, and Mental Causation

the idea that to understand the holistic or ‘non-local’ correlations of the components of entangled systems we need to resort to the postulation of any action at a distance or instantaneous causation between the separated particles. Nonseparability says that the postulation of this kind of implausible action at a distance is unnecessary. In this sense, for example, Papineau affirms that entanglement must be understood as [A] ‘non-local’ fact about the joint system, in the sense that it cannot be viewed as the sum of local facts about the separated ­electrons. This kind of non-locality needs to be distinguished from the non-local action at a distance that some interpretations of quantum mechanics posit to explain what happens when measurements are made on distant ‘entangled’ objects. [… This kind on non-locality] arises directly from the fact that the wave function for multiple particles can contain information beyond what it implies for any localized properties of the particles. This species of non-locality is thus unavoidable in any interpretation of quantum mechanics that views the quantum wave function realistically. (2008 145)

But what causal workings underlie this phenomenon? As far as I know, there is no analysis of this kind. And I think that the reason is that there is no careful analysis of the nature and structure of the emergent and downward causation that is at the basis of this kind of holistic phenomenon. Let us then examine one of the simplest cases based on the kind of examples of epr (Einstein, Podolsky & Rosen 1935) in Bohm’s version (1951). As we have said, electrons are particles with spin s = ±½, which means that a measurement on the spin of an electron will always show one of two possible values, namely, +½ or -½, with equal probability.15 As usual, let us call the first value ‘spin up’ (Sup) and the 15

This does not mean that the electrons do not have a determinate spin value (either +½ or -½) before the measurement, as the orthodox Copenhagen interpretation of quantum mechanics affirms. It only means that there is an objective probability of 0.5 that the measurement shows the value s = +½, and an objective probability of 0.5 that the measurement shows the value s = -½. Many-worlds interpretations (see, e.g., Everett 1957 and Vaidman 2015) and

179

Juan Diego Morales Otero

second ‘spin down’ (Sdown). Suppose then, that at time t1 an electron E1 is emitted, traveling through empty space to, and interacting with, a Stern-Gerlach magnet (M1) at time t2. This interaction allows us to measure the value of the spin of E1 at t2 because these magnets are devices having an orientation that determines a deflection on the particles with spin S = ±½. When one of these particles enters the detectors, it is deflected either in the magnet’s direction of orientation (whereby the electron’s spin = Sup) or in the opposite direction (whereby the electron’s spin = Sdown). The graphic of the experimental setup is as follows:

E1 with the property S = ±

Particle source

M1

Diagram 4. The quantum experimental setup with M1.

Quantum mechanics predicts that there is a probability of 0.5 that the value of the spin of E1 at t2 be Sup, and a probability of 0.5 that this value be Sdown. Let us call ‘Ev 1’ the event that consists in the emission of E1 with the property S = ±½ at t1, ‘Ev2’ the event that consists in the interaction between E1 and M1 that allows us to measure the value Sup of the spin of E1 at t2, and ‘Ev3’ the event that consists in the interaction between E1 and M1 that allows us to measure the value Sdown of the spin of E1 at t2. In causal terms we would have that the event Ev 1 can (that is, has the power to) cause either the event Ev2 with a probability of 0.5, or the event Ev3 also with a probability of 0.5. Graphically we would have the following:

modal interpretations (see, e.g., van Fraassen 1973 and Lombardi & Dieks 2014) of quantum mechanics argue that these values are determined before the measurements, although, in many cases, holistically determined, that is, as a function of quantum states of entanglement.

180

Emergent, Downward, and Mental Causation

Ev 2 Ev 1

E 1 with S up

bability of 0.5

causes with a pro

E 1 with S=±

causes with

Ev 3 a probabilit

y of 0.5

E 1 with S down

t1

t2

Diagram 5. Quantum causation with E1.

Ev2

Evm

Ev3

Evn

t2

t3

Ev1

t1

Diagram 6. Quantum causation of Ev1.

Similarly, with a different electron E2 emitted at time t3 from the same source to another Stern-Gerlach magnet (M2) that is located in the opposite direction. The graphic of the experimental setup is as follows:

E2 with the property S = ±

M2

Particle source

Diagram 7. The quantum experimental setup with M2.

Quantum mechanics predicts that there is a probability of 0.5 that the value of the spin of E2 at t 4 be Sup, and a probability of 0.5 that this value be Sdown. We can call ‘Ev4’ the event that consists in the emission of E2 with the property S = ±½ at t3, ‘Ev5’ the event that consists in the interaction between E2 and M2 that allows us to measure the value Sup of the spin of E2 at t 4, and ‘Ev6’ the event that consists in the interaction between E2 and M2 that allows us to measure the value Sdown of the spin of E2 at t 4. In causal terms we would have that

181

Juan Diego Morales Otero

the event Ev4 can (has the power to) cause either the event Ev5 with a probability of 0.5, or the event Ev6 also with a probability of 0.5. Graphically we would have the following: Ev 5 Ev 4

E 2 with S up

bability of 0.5

causes with a pro

E 2 with S=±

causes with

Ev 6 a probabilit

y of 0.5

E 2 with S down

t3

t4

Diagram 8. Quantum causation with E2.

Ev5

Evo

Ev6

Evp

t4

t5

Ev4

t3

Diagram 9. Quantum causation of Ev4.

Consider now that E1 and E2 are simultaneously emitted from the same source, E1 towards M1 and E2 towards M2. The graphic of the experimental setup is as follows: E2

M2

with the property S = ±

.

E1 with the property S = ±

Particle source

.

M1

Diagram 10. E1 and E2 are simultaneously emitted.

Quantum theory predicts that the event Ev 7 at t1 (which consists inthe simultaneous emission of E1 and E2, both with the property S = ± ½, towards M1 and M2 correspondingly) can (has the power to) cause with a probability of 0.25 the instantiation of one of the following events at t2:

182

Emergent, Downward, and Mental Causation

A (Ev2& Ev5), consisting in the interaction between E1 and M1 that allows us to measure the value Sup of the spin of E1, and the interaction between E2 and M2 that allows us to measure the value Sup of the spin of E2. B (Ev2& Ev6), consisting in the interaction between E1 and M1 that allows us to measure the value Sup of the spin of E1, and the interaction between E2 and M2 that allows us to measure the value Sdown of the spin of E2. C (Ev3& Ev5), consisting in the interaction between E1 and M1 that allows us to measure the value Sdown of the spin of E1, and the interaction between E2 and M2 that allows us to measure the value Sup of the spin of E2. D (Ev3& Ev6), consisting in the interaction between E1 and M1 that allows us to measure the value Sdown of the spin of E1, and the interaction between E2 and M2 that allows us to measure the value Sdown of the spin of E2. In graphical terms the causal situation is as follows: A (Ev 2 & Ev 5)

Ev7

causes

with a

ility of

probab

E 1-S up & E 2-S up

0.25

B (Ev 2 & Ev 6 )

ty of 0.25

E 1-S up & E 2-S down

causes with a probabili

E1 with S = ± & E2 with S = ±

causes with cause

s wit

C (Ev 3 & Ev 5)

a probabilit

y of 0.25

hap

roba

bility

E 1-S down & E 2-S up D (Ev 3 & Ev 6)

of 0.2

5

E 1-S down & E 2S down

t1

t2

Diagram 11-A. Quantum causation of Ev7.

Ev7

t1

A

Evm

B

Evn

C

Evo

D

Evp

t2

t3

Diagram 11-B. Quantum causation of Ev7.

183

Juan Diego Morales Otero

But the situation changes when the electrons are in a quantum state of entanglement. As we have said, an entangled quantum state is a state of a physical system characterized by having a state vector (or wave function) that is not factorizable in terms of the state vectors of its components: Ψsystem≠ Ψcomponent-1 ⊗ Ψcomponent-2 ⊗… Ψcomponent-n This means that the system has a holistic or ‘non-local’ property, in the sense that this property can be neither deduced nor explained from the properties and relations of the components. This holistic property introduces some systemic dynamics which cannot be anticipated from the individual behavior of the parts. One can prepare a system of two electrons (E1 and E2) in a state of entanglement, for example, through the decay of an original particle.16 In this process the electrons are simultaneously sent to M1 and M2 respectively. In this case, quantum mechanics predicts that the probabilities that have been projected above about the causal dynamics of the separable electrons (electrons which are not in quantum states of entanglement) will change due to the state of entanglement of the system that influences the causal dynamics of the constituents. Quantum theory successfully predicts that the behavior of the electrons will show a necessary correlation: if the left hand electron is Sup, the right hand one will be Sdown, and vice versa. Let us call ‘Ev8’ the event which consists in the simultaneous emission of the electrons E1 and E2 in a particular organization at t1, and let us consider that Ev8 is the microphysical basis for the emergence of the entangled quantum state of the system of these two electrons, which we can call event ‘Ev9’. In causal terms, the theory predicts that Ev9 at t1 can cause with a probability of 0.5 the instantiation of one of the following events at t2: B (Ev2& Ev6),consisting in the interaction between E1 and M1 that allows us to measure the value Sup of the spin of E1, and the interaction

16

184

Entanglement is usually created by direct interactions between subatomic particles or by particle decay, that is, the spontaneous process of one particle transforming into other particles.

Emergent, Downward, and Mental Causation

between E2 and M2 that allows us to measure the value Sdown of the spin of E2. C (Ev3& Ev5), consisting in the interaction between E1 and M1 that allows us to measure the value Sdown of the spin of E1, and the interaction between E2 and M2 that allows us to measure the value Sup of the spin of E2. This means that the holistic property of entanglement downwardly constrains and selects the underdetermined possibilities given at the lower level, the nomological possibilities of the electrons that constitute the system. In particular, the state of entanglement that the system has on a given occasion acts as a downward causal influence which determines that the probabilities of the subsequent dynamics of the constituent electrons be modified: from a probability of 0.25 for each of the events A, B, C, and D, it determines the probability of 0.5 for B and 0.5 for C; eliminating with this any probability for the instantiation of both A and D. In graphical terms, we are going from the situation in which the event Ev 7 causes, with a probability of 0.25, the instantiation of one of the A, B, C, or D events,

Ev7

t1

A

Evm

B

Evn

C

Evo

D

Evp

t2

t3

Diagram 11-B. Quantum causation of Ev7.

to the situation in which the event Ev9 causes, with a probability of 0.5, the instantiation of either B or C:

Ev9

t1

A

Evm

B

Evn

C

Evo

D

Evp

t2

t3

Diagram 12. Quantum causation of Ev9.

185

Juan Diego Morales Otero

Ev9 emission of electrons in an entangled state

con

stra

emergence

ins

the

inst

ant

iati

on

A (Ev 2 & Ev 5)

of

bability

h a pro uses wit

of 0.25

ca certain organization of E1 with S = ± & E2 with S = ±

Ev8

ity of 0.5 causes with a probabil

causes with cause

s wit

E 1-S up & E 2-S up B (Ev 2 & Ev 6 )

E 1-S up & E 2-S down C (Ev 3 & Ev 5)

a probabili

hap

ty of 0.5

roba

bility

E 1-S down & E 2-S up D (Ev 3 & Ev 6)

of 0.2

5

t1

E 1-S down & E 2S down

t2

Diagram 13. Emergence and quantum causation of Ev9.

This articulation shows that the holistic or emergent properties of quantum entanglement that the electrons can have on certain occasions are causally relevant. And it shows how this is possible, that is, as a downward causal influence that constrains and selects the different and underdetermined causal possibilities that the properties and laws that govern the electrons when they are in other kind of systems leave open. I now turn to an analysis of the causal nature of the mental states through a particular examination of the neurobiological basis of pain and the appearance of different levels of personal and experiential psychological phenomena. Mental Causation as Emergent Causation

In this final section, I analyze the neurobiological basis of pain and its connection to the existence of different levels of personal and experiential psychological phenomena. I describe the two different neurological nociceptive (relative to pain) subsystems, namely, the discriminative and the affective nociceptive neural structures that are primarily composed of the thalamus, the somatosensory cortex, and the anterior cingulate cortex, from which two different and corresponding nociceptive experiences arise: the discriminative

186

Emergent, Downward, and Mental Causation

and affective nociceptive experiences. I examine the different levels of composition and organization that are involved here to focus on the conceptual articulation of the causal dynamics that structure the interaction between the two experiential levels appearing in this phenomenon, that is, the level of the discriminative and affective nociceptive experiences and the level of our normal and unitary experience of pain. The International Association for the Study of Pain (IASP) defines pain as «An unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.» (IASP 1994 S217) Biologically speaking, the experience of pain is part of the organism’s defense system that produces a reflexive retraction from the painful stimulus, tendencies to protect the affected body part while it heals, and the avoidance of the harmful situation in the future. The set of processes in the body that are involved in the experience of pain constitute what is called nociception, because they allow the organism to perceive, encode, and process information about potential bodily harm. Nociceptive pain is the experience that is caused by the ongoing activation of pain receptors in either the surface or deep tissues of the body. Meanwhile, neuropathic pain is the experience that is normally caused when the nerve fibers themselves are damaged, dysfunctional, or injured, sending «incorrect» signals to other pain centers of the neural system. Nociceptive pain is divided into two types: somatic pain and visceral pain. The former is caused by injury to skin, muscles, bone, joint, and connective tissues, and it normally carries well defined information about the injured bodily part. Visceral pain originates in ongoing injury to the internal organs or tissues and is poorly localized. In the case of nociceptive pain, the signal of a potential harm travels from the periphery to the spinal cord along A-δ and C fibers. Because the A-δ fiber is thicker than the C fiber, and is thinly sheathed in an electrically insulating material (myelin), it carries its signal faster (5-30 m/s) than the unmyelinated C fiber (0.5-2 m/s). In consequence, A-δ fiber provokes what is known as first pain or fast pain, a sensation experienced immediately after an injury, transmitting and processing the location, intensity, duration, and nature (stabbing, burning,

187

Juan Diego Morales Otero

prickling) of the stimulus. Meanwhile, C-fibers produce what is known as second pain or slow pain, a sensation that follows first pain and can be characterized as a diffuse and persisting burning pain. This subsystem is responsible for the production of the unpleasant part of the painful sensation, that is, what is normally called by neurobiologists the affective-emotional component of pain and by philosophers, the phenomenal or qualitative aspect of the sensation. Pain impulses enter the spinal cord, where they synapse primarily on the dorsal horn neurons. Laminae I and V of the spinal cord project discriminative information (fast pain) to the lateral nuclei in the thalamus, and lamina VI projects affective-emotional information (slow pain) to medial nuclei in the thalamus (Hardcastle 2001). From this point, the lateral nuclei of the thalamus send the nociceptive information to the somatosensory cortex, which computes the location and characteristics of the stimulus, and the medial nuclei of the thalamus transmit the information to the anterior cingulate cortex which processes the affective-emotional part of the signal. Finally, there are important connections between the somatosensory cortex and the anterior cingulate cortex that produce the integrated experience of pain that we naturally undergo. The schematic diagram of the afferent sensory system of pain is as follows (based on Hardcastle 2001 299): anterior cingulate cortex Interconnection somatosensory cortex

medial nuclei in the thalamus

surface

lateral nuclei in the thalamus

C fibers A- fibers

dorsal horn

Diagram 14. The afferent sensory system of pain (based on Hardcastle 2001 299).

188

Emergent, Downward, and Mental Causation

Thalamus, somatosensory cortex, and anterior cingulate cortex are the principal components of the central nervous system that process pain information, but they are not the only ones. The following diagram of the ascending pathways, subcortical structures, and cerebral cortical structures shows the areas most notably activated by pain signals: PAG, periaqueductal gray; PB, parabrachial nucleus of the dorsolateral pons; VMpo, ventromedial part of the posterior nuclear complex; MDvc, ventrocaudal part of the medial dorsal nucleus of thalamus; VPL, ventroposterior lateral nucleus of thalamus; ACC, anterior cingulate cortex; PCC, posterior cingulate cortex; HT, hypothalamus; S-1 and S-2, first and second somatosensory cortical areas; PPC, posterior parietal complex; SMA, supplementary motor area; AMYG, amygdala; PF, prefrontal cortex (taken from Price 2000 1770):

Diagram 15. The areas most notably activated by pain signals (taken from Price 2000 1770).

Although there are different and, to some extent, separable subsystems of the entire neurophysiological nociceptive system, virtually all theorists agree about its complex and holistic status, in the sense that it is only through the orchestrated interaction between the

189

Juan Diego Morales Otero

multiple neural components that the whole system can deliver and produce both the singular and integrated experience of pain that we naturally undergo and its biological, and psychological functions.17 In this sense, Hardcastle affirms: Neither conscious experience of pain, the damaged tissue itself, nor our bodily or emotional reactions are fundamental to pain processing. Each is but one component of a larger processor. Hence, it is a mistake to try and claim one or the other as pain simpliciter. And it is equally erroneous to conclude that since we cannot identify one or the other with pain, there is no such thing. [… The] components of our pain system take pressure, temperature, and chemical readings of our surface (and interior) and use this information to track what is happening with our tissues. The Aδ cells and the C fibers do this, as do the spinothalamic tract and its connections to the cortex. In short, it appears that we have a complex but well-defined sensory system which monitors our tissues in order to promote the welfare of our bodies. (2001 303)

The claim that pain is a global phenomenon that is not identifiable with any of its neurological subsystems, however, does not imply the very plausible idea of a global emergent property of the entire neurophysiological nociceptive system that cannot be reduced to and, consequently, explained on the basis of its lower level or local properties in an aggregative way. As we have said, there is a difference between our discriminative pain processing (the subsystem responsible for our fast pain) and our affective pain processing (the subsystem responsible for our slow pain). Neurologists have found this distinction through various empirical situations of both natural accidents and purposeful interventions: ingestion of some opiates such as morphine, lesions to the medial thalamus, and prefrontal lobotomies all result in experiences of pain without a sense of suffering; in these cases, patients can localize the origin of the harmful signals but are not upset by the fact that they have 17

190

See, for example, Sarter, et al. 1996, Ingvar 1999, Price 2000, and Hardcastle 2001.

Emergent, Downward, and Mental Causation

them (Hardcastle 2001). But we can also have the contrary situations: specific cortical lesions can produce circumstances in which patients undergo painful experiences without being capable of discriminating and localizing the harmful stimulus (Davies et al. 1995). This situation leads us to wonder about the unity of our natural experience of pain and the possibility of its constituting an emergent global property that cannot be understood and explained only on the basis of the functioning of its neurological and even its experiential discriminative and affective components. Here we have at least three levels of composition and organization: first, the neurological level of the interaction between at least the thalamus, the somatosensory cortex, and the anterior cingulate cortex; second, the personal level of the appearance of two nociceptive experiences: the discriminative nociceptive experience (which has as its principal neural constituents the processing of both the lateral nuclei of the thalamus and the somatosensory cortex) and the affective nociceptive experience (which has as its principal neural constituents the processing of both the medial nuclei of the thalamus and the anterior cingulate cortex); and third, the level of our normal unitary experience of pain. From an emergentist point of view, it is very plausible that the level of the discriminative and affective nociceptive experiences is an emergent level that is not reducible to the components and interaction between their neural components, but the complexity of the empirical elucidation of this point exceeds the limits of this study. For this reason, I will focus on the conceptual articulation of the causal dynamics that should structure the interaction between the two experiential levels, that is, the interaction between the level of the discriminative and affective nociceptive experiences and the level of our normal and unitary experience of pain. Although the interconnection between these two levels of experiences is ultimately an empirical question that must be resolved through physiological, neurological, and psychological studies, I think it is quite plausible that the global and unitary experience of pain should be understood as a phenomenon that emerges from the interaction of its simpler experiential components. The following scenario can support this intuition or, at least, show its specific sense.

191

Juan Diego Morales Otero

It is necessary to recall that an emergent property has a causal power which is not determined by the causal powers of its constituents and, therefore, cannot be deduced, calculated, or anticipated, even with the use of contingent and a posteriori mathematical tools that connect the properties of the parts with those of the whole, in a compositional way, that is, delivering the value of the global property directly and exclusively from the values of the components. Let us suppose that John is an athlete who has recovered from a knee injury a few weeks ago and that he is about a mile from finishing a marathon. John inadvertently hits his shoulder against the branch of a tree that is in his way, a situation which under normal conditions would cause sharp pain. There are three possible scenarios in which this situation can cause various mental and behavioral states that depend on the type of the experience that John undergoes. In a first case, John-D has a neural impairment that allows only the working of his discriminative nociceptive system (he can have, for instance, a problem in his anterior cingulate cortex); so his hurt shoulder produces a psychological state that lets him know about the specific traits of the damage and causes the following situation: as John only has discriminative information about the tissue damage of his shoulder that does not affect him emotionally, and because he wants to finish the marathon as soon as possible, he does not diminish his step; in this situation, his sensory discriminative experiential state can cause either an increase of 10% or an increase of 20% in his anxiety levels. The diagram of the causal situation is as follows (BH: bodily harm; IDNE: instantiation of the discriminative nociceptive experience; I of 10% in AL: an increase of 10% in the anxiety levels; I of 20% in AL: an increase of 20% in the anxiety levels):

BH

t1

I of 10% in AL

Evm

I of 20% in AL

Evn

IDNE

t2

t3

t4

t5

Diagram 16. The causal situation of John-D.

192

t6

Emergent, Downward, and Mental Causation

In a second case, John-A is identical to John-D in all relevant physiological and psychological aspects, except that he has a neural impairment that allows only the working of his affective nociceptive system (he can have a problem in his somatosensory cortex); so his hurt shoulder produces a sharp painful experience that allows him to track neither the type nor the location of the damage and that causes the following subsequent events: John knows that there is damage to his body, he remembers that he recovered from a knee injury a few weeks ago, and he believes that his knee is causing his experience of pain; this situation causes John to diminish his step, and causes either an increase of 70% or an increase of 80% in his anxiety levels. The diagram of the causal situation is as follows (iane: instantiation of the affective nociceptive experience):

BH

t1

I of 70% in AL

Evo

I of 80% in AL

Evp

t5

t6

IANE

t2

t3

t4

Diagram 17. The causal situation of John-A.

Finally, in a third case John-N is identical to John-D and J­ ohn-A in all the relevant physiological and psychological aspects except that he has a functionally normal nociceptive system; so his hurt shoulder produces a sharp and well discriminated and localized pain that causes the following events: John recognizes that his pain is caused by something like a jab in his shoulder, he diminishes his step slightly, and there is either an increase of 35% or an increase of 40% in his anxiety levels. Now, the question is whether we can reconstruct or predict these causal consequences from the causal consequences of the cases of John-D and John-A. In particular, whether we can build, calculate, or predict the causal powers of the normal and unitary experience of pain that John undergoes with respect to the increase of 35 % or 40 % in his anxiety levels from the causal powers of the lower level interaction between his discriminative and affective nociceptive experiences, or

193

Juan Diego Morales Otero

whether the unitary painful experience possesses a causal factor that is irreducible to the causal power of these components. As we have said, this is an entirely empirical question that must be resolved through physiological, neurological, and psychological studies about the exact causal powers and consequences of the lower level nociceptive experiences when they are located in different psychological systems (e.g., John-D and John-A), and their relation to the exact causal powers and consequences of the normal and unitary nociceptive experiences (in systems like John-N). Given these circumstances, we can have two possible kinds of empirical results. On one hand, the unitary experience of pain may be realized in a non-emergent way, that is, it can be synchronously determined (through compositional contingent and a posteriori principles) exclusively from the lower level properties and relations of their closest components, namely, the discriminative and affective nociceptive experiences. In this case, both the property of having a normal and unitary experience of pain and its causal powers would be reduced to the (compositional connection of the) properties and causal powers of the lower level component experiences. The diagram of the causal situation would be as follows: Unitary experience of pain

IDNE

I of 35% in AL

Evq

I of 45% in AL

Evr

BH IANE

t1

t2

t3

t4

t5

t6

Diagram 18. The causal situation of John-N whether his pain is realized in a non-emergent way.

But if the unitary experience of pain is a phenomenon that emerges from the interaction between the discriminative and affective nociceptive experiences, then the situation implies the existence of a causal underdetermination on which the emergent experiential property should act as a causal constraint. This simply means that the causal powers that the discriminative and affective nociceptive experiences have through other psychological systems different from

194

Emergent, Downward, and Mental Causation

the normal (that is, different from the systems that undergo normal and unitary painful states) are causal powers that are insufficient for the causal determination of (the objective probability of) the causal chains that are further determined by the system’s action in virtue of the holistic and emergent property of having a unitary pain experience. The diagram of the lower level causal situation would be as follows: Unitary experience of pain

IDNE

I of 30% in AL

Evs

I of 35% in AL

Evq

I of 40% in AL

Evr

I of 45% in AL

Evt

BH IANE

t1

t2

t3

t4

t5

t6

Diagram 19. The lower level causal situation when the unitary experience is emergent.

And the corresponding diagram about the downward causal influence in terms of the higher level constraint and selection that the emergent property exerts over the lower level possibilities is this:

Unitary experience of pain

emergence

co ins nst tan rain tia s t tio he no f

IDNE

I of 30% in AL

Evs

I of 35% in AL

Evq

I of 40% in AL

Evr

I of 45% in AL

Evt

BH IANE

t1

t2

t3

t4

t5

t6

Diagram 20. The downward causal influence of the emergent experience over the lower level possibilities.

195

Juan Diego Morales Otero

I have tried to broadly outline the most important conditions that must articulate the structure of the macrophysical causation if it is irreducible to the lower levels causation from which it must arise. We have seen that the question about the existence of emergent causation, in general, and mental causation, in particular, is necessarily an empirical issue that must be resolved through scientific work. We have also seen that within the mental level there can be multiple levels of composition and organization that can finally be emergent. For example, it is quite plausible that cognitively complex phenomena like imagination and solving theoretical problems, or even mental states like desire, belief, and thought, constitute mental phenomena that emerge from simpler personal informational states like sensations and perceptions. The emergentist theory then argues not only that the mind can emerge from atoms, molecules, cells, and neural informational processing, but that the very mental states can become organized in a hierarchical, emergent, and irreducible way. In fact, this can be the ascending path towards the emergence of self.18 If the overall emergentist picture on mental causation is correct, then human agency should function as the complex and highly contextual higher level constraints and selections that our minds impose on the underdetermined causal chains that are left open by the governance of the lower level microphysical, chemical, biological, and neurological laws. These hypotheses, however, must be analyzed and tested from the empirical point of view, a complex task to be performed given the invaluable empirical and philosophical fruits that may come to revolutionize the way we see both the physical world and our place in it. Summary

In this chapter, I have focused on a careful articulation of the concept of emergent causation and its application to the phenomenon of mental causation, clarifying and interconnecting the prominent concepts of downward causation, lower level causal underdetermination, and higher level causal constraint and selection. I developed 18

196

See Thagard 2012 on this line of thought.

Emergent, Downward, and Mental Causation

this articulation through the elucidation of three examples: in the first place, I analyzed and reconstructed in detail an abstract example that helped us understand the general mechanism and structure of downward causation which is based on the constrictive and selective action of the higher level laws on the underdetermined possibilities of the lower levels. Then I carried out a conceptual analysis of the causal structure of one of the most recurrent examples of the seeming failure of the microphysical supervenience theory, that is, the phenomenon of the quantum states of entanglement. I argued that if our current scientific understanding of the quantum world is correct, then we have a primary empirical example that can help us both to claim the failure of microphysicalism, and to clarify and articulate in a very concrete way the nature and structure of the phenomenon of emergent and downward causation. In the final section, I examined the neurobiological basis of pain and its connection to the existence of different levels of experiential psychological phenomena. I described the two different neurological nociceptive subsystems, namely, the discriminative and the affective nociceptive neural structures, from which two different and corresponding nociceptive experiences arise: the discriminative and affective nociceptive experiences. I examined the different levels of composition and organization that are involved here, focusing on the conceptual articulation of the causal dynamics that should structure the interaction between the two experiential levels involved in this phenomenon, that is, the level of the discriminative and affective nociceptive experiences, and the level of our normal and unitary experience of pain. Finally, I showed that on the basis of this kind of examples, the emergentist proposal argues that the different mental states can become organized in a hierarchical, emergent, and irreducible way.

197

Appendix: Functional Reduction of the Special Properties

In different opportunities, I have argued that microphysicalism, the physicalist theory affirming the metaphysical supervenience of the higher or macroscopic properties on the microphysical conditions, is insufficient both conceptually and empirically: on the one hand, it allows us to understand neither the scientific nor the daily use of the notion of physical entity and, on the other hand, it seems to be incompatible with different results of the sciences. Nevertheless, the last point is completely empirical: it could be that all macroscopic properties finally turn out to be metaphysically supervenient on their microphysical realizers. In chapter 4, I saw good reasons to claim that if this were the case, we should affirm with Kim and Papineau that neither epiphenomenalism nor NRP could end up being acceptable perspectives about the higher level properties. The rationale is that the MR of these properties, together with the fact that they are completely determined by their dissimilar physical realizers, implies that every instance of a higher level property can have nothing in common with another of its instances. In this case, a higher level property does not seem to have anything objective in virtue of which its various instances are actually its instances and, therefore, anything objective in virtue of which it is a real property.

199

Juan Diego Morales Otero

Several philosophers affirm that to avoid eliminativism about higher level properties we must reduce them ‘locally’ to each of their physical realizers.1 The idea is that a special property S in beings of a kind A (say humans) is the property which occupies the functional role specified by S, that is, the property that realizes S in normal members of A; the property S in beings of kind B (say reptiles) is the property that realizes S in normal members of B; the property S in beings of kind C (say Martians) is the property that realizes S in normal members of C; and so on. And if there is more than one property which can realize S within normal members of (for example) kind A, or even within individual members of this kind, then S should be relativized to these different realizers. This local reduction movement is based on the idea of reduction that philosophers such as Armstrong, Lewis, Chalmers, Jackson, Kim, and Levine have developed in recent years, which Kim calls functional reduction and Block (2015) the Canberra plan. In a somewhat paradoxical way, this reductive model takes the functionalization of the property to be reduced as its first and fundamental step, a functionalization that was precisely considered by Putnam and Fodor as the necessary means to understand both the MR and the irreducibility of special properties. Once the property to be reduced has been functionalized, the second step for reduction is to find the physical property that fills the specified functional role. As Levine says, [O]n this view explanatory reduction is, in a way, a two-stage process. Stage 1 involves the (relatively? quasi?) a priori process of working the concept of the property to be reduced ‘into shape’ for reduction by identifying the causal role for which we are seeking the underlying mechanisms. Stage 2 involves the empirical work of discovering just what those underlying mechanisms are. (1993 132)

Kim adds a final step, that is, the explanation of the mechanisms by which the realizers accomplish the specified function. In his account, To reduce a property, say being a gene, on this model, we must first ‘functionalize’ it; that is, we must define, or redefine, it in terms 1

200

See, for example, Lewis 1980, Enç 1983, Kim 1992b, Bickle 1998, Papineau 2010.

Appendix: Functional Reduction of the Special Properties

of the causal task the property is to perform. Thus, being a gene may be defined as being a mechanism that encodes and transmits genetic information. That is the first step. Next, we must find the ‘realizers’ of the functionally defined property —that is, properties in the reduction base domain that perform the specified causal task. It turns out that DNA molecules are the mechanisms that perform the task of coding and transmitting genetic information— at least, in terrestrial organisms. Third, we must have an explanatory theory that explains just how the realizers of the property being reduced manage to perform the causal task. In the case of the gene and the DNA molecules, presumably molecular biology is in charge of providing the desired explanations. (Kim 2005 101)

Let us suppose that we can give a functional analysis of pain in terms of its causal role: as the state that is normally caused by tissue damage and that, in turn, causes aversive behavior. Let us say that this is the causal role C of pain. We can claim, then, that having pain = def. having one or another property satisfying causal role C. Now, consider that P1 is the neural property which occupies C in humans, P2 the neural property which occupies C in reptiles, and P3 the hydraulic property which occupies C in Martians. By the chapter 4’s arguments against NRP, we have to say that having pain is not really a unitary and natural property instantiated by its different realizers; we also have to deny that having pain is a disjunction of its physical realizers, just because disjunctions of properties are not in themselves real properties; but it is clear that the property of having pain cannot be identified with any of these three realizers. The functional reduction idea is that we should identify the property of having-pain-in-humans with the property which occupies C in humans, that is, P1, the property of having-pain-inreptiles with P2, and the property of having-pain-in-Martians with P3. This strategy takes predicates of the special sciences, such as ‘having pain’, ‘being a mammal’, and ‘being a chemical bond’ as analogous to expressions like ‘being a can opener’ and «the property of being one or another property that annoys Jones,»2 that is, as 2

To use an example cited by Jackson 2002 646-7.

201

Juan Diego Morales Otero

expressions which refer in a contingent form to the different first order properties that occupies the causal role that these expressions define. In this case, predicates of the special sciences are not rigid designators, that is, expressions that refer to the same properties in every possible world, but descriptions that pick out the contextually indicated properties that occupy the causal roles in question. ­Following our example, ‘having pain’ would be the special predicate that refers to the contextually indicated property that occupies the causal role C, that is, the predicate that refers to P1 in the context of human beings, P2 in the context of reptiles, and P3 in the context of Martians. As this predicate refers in a contingent form, it is possible that it could refer to other different physical properties in our world or in other nomologically different worlds. But this approach leads to immediate problems. Although the functional reduction idea that mental and special properties must be relativized to species or structures was envisioned as a proposal to salvage these properties from elimination, many philosophers argue that its paradoxical result is that these properties should be finally eliminated.3 As Pereboom and Kornblith affirm, The suggestion that there are species-specific reductions of pain results in the claim that pains in different species have nothing in common. But this is just a form of eliminationism. If we generalize this view to other mental states, then we would be forced to reject the legitimacy of psychological theories that quantify over mental states which are instantiated in more than one species. (1991 135-6)

But the problem does not stop here. As Heil comments in this passage, The possibility that individuals in pain even within species might have ‘nothing in common’ (or at least nothing in common in virtue of which it would be true to describe them as being in pain) would, of course, make matters much worse. (Heil 2011 47)

3

202

See, for example, Block 1990, 2015, Pereboom & Kornblith 1991, Antony & Levine 1997, Horgan 1997, Shapiro & Sober 2007, Shapiro 2008, Heil 2011.

Appendix: Functional Reduction of the Special Properties

Let us see Shapiro and Sober’s treatment of the problem: Suppose, as Kim maintains, that M [the belief that martinis should be shaken], is P1 in human beings and P2 in Martians, where P1 and P2 are different physical properties. If so, they cannot both be identical to M. This suggests that Kim’s idea is that P1 is type identical with the property of being the human belief about martinis and P2 is type identical with the property of being the Martian belief about martinis. If this interpretation of Kim’s proposal is correct, then he apparently has given up on M. (2007 13) In fact, if the belief that martinis should be shaken is multiply realized within species, and not just between them, then the human belief must be causally inert by Kim’s argument. And if the belief is multiply realized within the life time of a single organism, then your believing that martinis should be shaken is also shown to be causally inert. (2007 note 16)

The consequence is that if the mental property of (say) having a headache is not only realized by different physical properties between and within different species, but also by different physical properties (say, different neural states) through different moments of your life, then your state of having a headache will not be real; this is just because what would be real will be your-having-a-headache-at-moment-1, your-having-a-headache-at-moment-2, your-having-a-headache-atmoment-3, and so on. In response Kim affirms: Given an extreme diversity and heterogeneity of realization, it would no longer be interesting or worthwhile to look for neural realizers of mental states for every human being at every moment of his/her existence. If psychology as a science were possible under these circumstances, that would be due to a massive and miraculous set of coincidences. We may conclude then that pushing multiple realizability to extremes does not impugn the idea of local reduction. It only makes local reductions more fine grained and atomistic and perhaps renders them practically not worthwhile. (Kim 1998 95)

On this reading, what could turn out to be worthwhile is not looking for physical realizers of mental and special properties, but

203

Juan Diego Morales Otero

getting rid of them and developing the different physical explanations behind all these heterogeneous phenomena.4 There is currently a great controversy about the actual multiple realization of mental properties between and within different species, within individual humans, and even about the multiple realization of biological and chemical ­properties.5 But on the functional reductive perspective, this no longer has a real importance. In strict sense, insofar as the categories of the special sciences pick out multiple and dissimilar physical properties, these are not (in fact there are no) scientific categories truly referring to real special properties but only to sets of basic physical properties in a somehow contingent and indirect way; that is, through second order and non-rigid designators.6 We can say that the overall conclusion is that the properties of the special sciences constructed as second order properties are not real natural properties and that the talk of multiple realization or rrealizability does not make real sense.7 What makes sense, on the functional reductive proposal, is the talk of second order designators that pick out different physical properties in a contingent way, designators that we could get rid of when we finally find and explain the functioning of the real and heterogeneous causal and ontological phenomena behind them. But we could still say that the functionalist strategy to identify first order properties with real causal powers is quite useful; that this can provide us with the necessary methods for the identification of the causal factors within the various sciences, both basic physical and special. What we must not overlook is that second order desigOn eliminativism regarding mental properties, see, for example, P.M. Churchland 1981 and P.S. Churchland 1986. 5 See, for example, Bickle 1998, Polger 2004, Gillett 2007, Shapiro 2008, Aizawa & Gillett 2009 and 2011. 6 In this sense, Kim says that «it is less misleading to speak of second order descriptions or designators of properties, or second-order concepts, than second order properties.» (Kim 1998 104) See also Jackson 2002 646-647, McLaughlin 2006 44, and Shoemaker 2007 67. 7 As Shapiro comments, «[i]f pain is a property, then we can understand what it means to say that pain is multiply realized in mammals and mollusks. But if the predicate ‘pain’ names no property, then there is nothing that mammals and mollusks multiply realize when each is said to be in pain» (2008 520). 4

204

Appendix: Functional Reduction of the Special Properties

nators can only refer to first order properties in a merely contingent and provisional way, until we can clarify the nature of the first order properties which we really want to pick out. This in no way implies that if we refer to higher level or special properties through functional or second order concepts these properties could not be real natural properties with all the necessary causal legitimacy and the corresponding necessary inclusion within our scientific ontology. The claim of the exclusion or elimination of the special properties as real natural properties is an additional step that is empirical and based on the plausibly empirically false assumption of the metaphysical supervenience of the macro-properties on their microphysical realizers or conditions, an assumption that is shared by both the functional reductive proposal8 and by NRP. To summarize, we can say that neither nrp nor the functional reductive proposal seem to have a satisfactory account of mental and special properties. On the one hand, nrp cannot vindicate the reality of special causal powers because it cannot explain the necessary unity they must maintain through their different physical realizers (see chapter 4). On the other hand, although more metaphysically robust, the functional reductive proposal affirms that insofar as the special properties arise from different physical conditions, they finally cannot be real, a consequence that I think is implausible.

8

Expressed in its second step for the accomplishment of the functional reduction; that is, its idea that we have to find the basic physical (micro-physical) property that fills the specified functional role of the higher level property to be reduced.

205

Conclusion

I have said that the main purpose of this dissertation is to discuss and respond to the reiterated critiques leveled by some reductionist philosophers, especially Jaegwon Kim, at the non-reductive physicalist explanation of the causal power or relevance of the higher level properties, that is, those of the special sciences, such as chemistry, biology, and, especially, psychology. We have seen that most of contemporary analytic philosophy is mistaken in assuming a physicalist proposal on the basis of the metaphysical supervenience theory of the higher level and macro properties on their microphysical bases, the theory that we have called microphysicalism. As we have argued, this proposal has not only deep empirical but conceptual problems: on the one hand, it allows us neither to fix nor to understand the use we make of the notion of the physical and, therefore, of the notion of a physicalist theory; this only means that there are alternative ways to understand and formulate physicalism which are not based on the microphysical supervenience theory. And, on the other hand, it seems to be incompatible with results coming both from physical science itself, as when we talk about holistic or systemic physical properties not reducible to their constituent conditions, and from the special sciences’ greatly successful

207

Juan Diego Morales Otero

theories and experiments that provide explanations and predictions which, as far as we know, are not reducible to the microphysical laws and explanations from which they must arise. Because the most important and accepted version of non-reductive physicalism -the current functionalist proposal- is committed to the metaphysical supervenience of the higher level properties on their microphysical bases, Kim and the reductionists are right in their conceptual criticisms of the inability to account for the reality and irreducibility of the causal powers of the higher level properties within this physicalist framework. But they are wrong in claiming that the failure of the current functionalist proposal implies a general failure of any non-reductive physicalism. Emergentism has been articulated as a type of macrophysicalist theory because it considers that special properties such as biological and mental ones are metaphysically dependent on but not metaphysically determined by and, therefore, not reducible to their microphysical bases. This proposal claims that higher level causation is articulated combining the underdetermination of the lower level causal processes, along with the instantiation of higher level causal laws that constrain and select from the different and underdetermined causal alternatives that the properties and laws that govern the lower processes leave open. As a final and general conclusion we can say that macrophysicalism or emergentism is not only a coherent and well suited conceptual proposal about the causal functioning of the different levels of composition and organization of our physical world, but that as far as we know it can be its most plausible empirical articulation.

208

References Aizawa, K. & Gillett, C. «The (Multiple) Realization of Psychological and other Properties in the Sciences». Mind & Language, vol. 24, n.° 2, 2009, pp. 181-208. Aizawa, K. & Gillett, C. «The Autonomy of Psychology in the Age of Neuroscience». Causality in the Sciences, edited by P. M. Illari, F. Russo, and J. Williamson. New York: Oxford University Press, 2011, pp. 203-223. Alexander, S. Space, Time, and Deity. Vol. 2. London: Macmillan, 1920. Anderson, P. W. «More is different». Science n.° 177, 1972. Reprinted in Emergence: Contemporary Readings in Philosophy and Science, edited by M. Bedau & P. Humphreys. Massachusetts Institute of Technology. Antony, L. & Levine, J. «Reduction with Autonomy». Noûs, vol. 31, n.° 11, Supplement: Philosophical Perspectives, Mind, Causation, and World, 1997, pp. 83-105. Armstrong, D. M. A World of States of Affairs. Cambridge: Cambridge University Press, 1997. Aronson J. J. «Unentangling Ontology from Epistemology in Causation». Erkenntnis, 18, 1982, pp. 293-305. Aspect, A., Grangier, P., & Roger, G. «Experimental Tests of Realistic Local Theories via Bell’s Theorem». Phys. Rev. Lett., 47, 1981, p. 460. Audi, P. «Properties, Powers, and the Subset Account of Realization». Philosophy and Phenomenological Research, vol. lxxxiv, n.° 3, 2012a. Audi, P. «Grounding: Toward a Theory of the In-Virtue-Of Relation». The Journal of Philosophy, vol. cix, n.° 12, 2012b. Ayer, A. J. Language, Truth and Logic. London: Victor Gollancz, 1936. Barnes, E. «Emergence and Fundamentality». Mind, vol. 121, 2012, p. 484. Baumgartner, M. «Interventionist Causal Exclusion and NonReductive Physicalism». International Studies in the Philosophy of Science, vol. 23, n.° 2, 2009, pp. 161-178. Baumgartner, M. «Interventionism and Epiphenomenalism». Canadian Journal of Philosophy, 40, 2010, pp. 359-383.

209

References

Bechtel, W. & Richardson, R. Discovering Complexity: Decomposition and Localization as Strategies in Scientific Research. Massachusetts Institute of Technology, 2010. Bedau, M. «Weak Emergence». Philosophical Perspectives, 11, 1997, pp. 375-399. Bell, J. S. «On the Einstein Podolsky Rosen Paradox». Physics, 1, 1964, pp. 195-200. Bennett, J. Events and Their Names. Oxford: Clarendon Press, 1988. Bennett, K. «Why the Exclusion Problem Seems Intractable, and How, Just Maybe, to Tract It». Noûs, 2003, pp. 471-497. Bennett, K. «Exclusion Again». Being Reduced: New Essays on Reduction, Explanation, and Causation, edited by J. Kallestrup and J. Hohwy. New York: Oxford University Press, 2008. Bernal, R. E-Physicalism: A Physicalist Theory of Phenomenal Consciousness. Ontos Verlag, 2012. Bickle, J. Psychoneural Reduction: The New Wave. Cambridge, MA: mit Press, 1998. Block, N. «Troubles with Functionalism». Readings in the Philosophy of Psychology. Vol. 1. Edited by N. Block. Cambridge, MA: Harvard University Press, 1980, pp. 268-305. Block, N. «Can the mind change the world?» Meaning and method: Essays in honor of Hilary Putnam, edited by G. Boolos. New York: Cambridge University Press, 1990. Block, N. «Do Causal Powers Drain Away?» Philosophy and Phenomenological Research, vol. 67, n.° 1, pp. 133-150. Block, N. «The Canberra Plan Neglects Ground». Qualia and Mental Causation in a Physical World: Themes from the Philosophy of Jaegwon Kim, edited by T. Horgan, M. Sabates, and D. Sosa. Cambridge University Press, 2015. Block, N. & Stalnaker, R. «Conceptual analysis, dualism and the explanatory gap». The Philosophical Review, n.° 108, 1999, 1-46. Bohm, D. Quantum Theory. Prentice Hall, 1951. Bohm, D. «A suggested interpretation of the quantum theory in terms of hidden variables. I & II». Physical Review, n.° 85, 1952, pp. 166-79 and 180-93.

210

References

Broad, C. D. The Mind and its Place in Nature. London: Routledge & Kegan Paul, 1925. Brown, W. S. «The Emergence of Causally Efficacious Mental Function». Evolution & Emergence: Systems, Organisms, Persons, edited by N. C. Murphy, & W. R. Stoeger. Oxford: Oxford University Press, 2007, pp. 88-110. Bunge, M. Causality: the Place of the Causal Principle in Modern Science. Cambridge: Harvard University Press, 1959. Campbell, D. «“Downward Causation” in Hierarchically Organised Biological systems». Studies in the Philosophy of Biology. Reduction and Related Problems, edited by F. J. Ayala & T. Dobzhansky. Berkeley: University of California Press, 1974. Carnap, R. «Psychology in Physical Language». Logical Positivism, edited by A. J. Ayer. New York: The Free Press, 1959, pp. 165-198. Cartwright, N. «Causal Laws and Effective Strategies». Noûs, vol. 13, n.° 4, Special Issue on Counterfactuals and Laws, 1979, pp. 419-437. Cartwright, N. «Do the Laws of Physics state the Facts». Pacific Philosophical Quarterly, n.° 61, 1980, pp. 75-84. Cartwright, N. «Why Physics?». The Large, the Small and the Human Mind, edited by R. Penrose, A. Shimony, N. Cartwright & S. Hawking. Cambridge: Cambridge University Press, 1997. Cavendish, M. L. Observations upon Experimental Philosophy, edited by E. O’Neill. Cambridge: Cambridge University Press, [1668] 2001. Copi, I. & Cohen, C. Introduction to Logic. Eighth Edition. New York: Macmillan Publishing Company, 1990. Chalmers, D. The Conscious Mind. Oxford: Oxford University Press, 1996. Chalmers, D. «Consciousness and its Place in Nature» The Blackwell Guide to Philosophy of Mind, edited by S. P. Stich and T. A. Warfield. Wiley-Blackwell, 2003. Chalmers, D. «Strong and Weak Emergence». The Re-Emergence of Emergence. The Emergentist Hypothesis from Science to Religion, edited by P. Clayton and P. Davies. New York: Oxford University Press, 2006. Chalmers, D. & Jackson, F. «Conceptual Analysis and Reductive Explanation». Philosophical Review, n.° 110, 2001, pp. 315-60.

211

References

Churchland, P. M. «Eliminative Materialism and Propositional Attitudes». Journal of Philosophy, n.° 78, 1981, pp. 67-90. Churchland, P. S. Neurophilosophy: Toward a Unified Theory of the Mind/Brain. Cambridge, MA: mit Press, 1986. Collins, J., Hall, N., & Paul, L. A. «Counterfactuals and Causation: History, Problems, and Prospects». Causation and Counterfactuals, edited by J. Collins, N. Hall, and L. A. Paul. Cambridge, MA: mit Press, 2004. Crane, T. «The Significance of Emergence». Physicalism and Its Discontents, edited by C. Gillett, & B. Loewer. Cambridge: Cambridge University Press, 2001. Crane, T. & Mellor, D. H. «There is No Question of Physicalism». Mind, vol. 99, n.° 394, 1990. Davidson, D. «Mental Events». Essays on Actions and Events. Oxford: Clarendon Press, [1970] 1980. Davidson, D. Essays on Actions and Events. Oxford: Clarendon Press, 1980. Davies, K. D. et al. «Visceral Pain Evoked by Thalamic Microstimulation in Humans». Neuroreport, n.° 6, 1995, pp. 369-74. Davies, P. «The Physics of Downward Causation». The ReEmergence of Emergence. The Emergentist Hypothesis from Science to Religion, edited by P. Clayton and P. Davies. New York: Oxford University Press, 2006. Davies, P. «The epigenome and top-down causation». Interface Focus, n.° 2, 2012, pp. 42-8. Deacon, T. «Three levels of emergent phenomena». Evolution & emergence: Systems, Organisms, Persons, edited by N. C. Murphy, & W. R. Stoeger. Oxford: Oxford University Press, 2007, pp. 88-110. Dehmelt, H. «Experiments on the Structure of an Individual Elementary Particle». Science, n.° 247, 1990, pp. 539-45. D’Holbach, P.-H. System of Nature, translated by H. D. Robinson. New York: Burt Franklin, [1770] 1970. Demaret, J., Heller, M., & Lambert, D. «Local and Global Properties of the World». Foundations of Science, n.° 2, 1997, pp. 137-76.

212

References

Dobzhansky, T. «Chance and Creativity in Evolution». Studies in the Philosophy of Biology. Reduction and Related Problems. Berkeley: University of California Press, 1974. Dowe, P. «Causality and Conserved Quantities: A Reply to Salmon». Philosophy of Science, n.° 62, 1995, pp. 321-33. Dretske, F. «Psychological vs. Biological Explanations of Behavior». Behavior and Philosophy, n.° 32, 2004, pp. 167-77. Ducasse, C. J. «On the Nature and Observability of the Causal Relation». Reprinted in Causation, edited by E. Sosa, & M. Tooley. New York: Oxford University Press, [1929] 1993, pp. 125-36. Dupré, J. Processes of Life. Essays in the Philosophy of Biology. New York: Oxford University Press, 2012. Dürr, D., Goldstein, S. & Zanghì, N. «Quantum Equilibrium and the Origin of Absolute Uncertainty». J. of Statistical Physics, n.° 67, 1992, pp. 843-907. Earley, J. «How Philosophy of Mind Needs Philosophy of Chemistry». International Journal for Philosophy of Chemistry, vol. 14, n.° 1, 2008, pp. 1-26. Earman, J., Roberts, J. & Smith, S. «Ceteris Paribus Lost». Ceteris Paribus Laws, edited by J. Earman, et al. Erkenntnis, n.° 52 (Special Issue), 2002, pp. 281-301. Einstein, A., Podolosky, B., & Rosen, N. «Can Quantum-Mechanical Descriptions of Physical Reality be Considered Complete?» Physical Review, n.° 47, 1935, pp. 777-80. Elder-Vass, D. The Causal Powers of Social Structures. New York: Cambridge University Press, 2010. Enç, B. «In Defense of the Identity Theory». Journal of Philosophy, n.° 80, 1983, pp. 279-98. Esfeld, M. «Quantum Holism and the Philosophy of Mind». Journal of Consciousness Studies, n.° 6, 1999, pp. 23-38. Evans, G. «Things Without the Mind». Collected Papers. Oxford. Clarendon Press, 1985. Everett, H. «“Relative State” Formulation of Quantum Mechanics». Review of Modern Physics, n.° 29, 1957, pp. 454-62. Fodor, J. «Special Sciences (Or: The Disunity of Science as a Working Hypothesis)». Emergence: Contemporary Readings in Phi-

213

References

losophy and Science, edited by M. Bedau, & P. Humphreys. ­Massachusetts Institute of Technology, 1974. Fodor, J. «Making Mind Matter More». A Theory of Content and Other Essays. Massachusetts Institute of Technology, 1990. Fodor, J. «Special Sciences: Still Autonomous After All These Years». Noûs, vol. 31, 1997, pp. 149-63. Fodor, J. Hume Variations. New York: Oxford University Press, 2003. Frith, C. D. «Free Will and Top-Down Control in the Brain». Downward Causation and the Neurobiology of Free Will, edited by N. Murphy, G. Ellis, & T. O’Connor. Berlin: Springer, 2009. Gell-Mann, M. The Quark and the Jaguar: Adventures in the Simple and the Complex. New York: W. H. Freeman and Company, 1994. Gillett, C. «The Metaphysics of Realization, Multiple Realization, and the Special Sciences». The Journal of Philosophy, n.° 11, 2003, pp. 591-603. Gillett, C. «Samuel Alexander’s Emergentism: Or, Higher Causation for Physicalists». Synthese, vol. 153, n.° 2, 2006, pp. 261-296. Gillett, C. «Understanding the New Reductionism: The Metaphysics of Science and Compositional Reduction». The Journal of Philosophy, vol. 104, n.° 4, 2007, pp. 193-216. Gillett, C. & Loewer, B. (eds.). Physicalism and Its Discontents. Cambridge: Cambridge University Press, 2001. Hall, N. «Two Concepts of Causation». Causation and Counterfactuals, edited by J. Collins, N. Hall, and L. A. Paul. Cambridge, MA: mit Press, 2004. Hardcastle, V. «The Nature of Pain». Philosophy and the Neurosciences: A Reader, edited by W. P. Bechtel, P. Mandik, J. Mundale, & R. S. Stufflebeam. Blackwell, 2001, pp. 295-311. Healey, R. «Holism and Nonseparability». The Journal of Philosophy, vol. 88, n.° 8, 1991, pp. 393-421. Heathcote, A. & Armstrong, D. «Causes and Laws». Noûs, n.° 25, 1991, pp. 63-73. Heil, J. «Powers and the Realization Relation». The Monist, vol. 94, n.°1, 2011, pp. 34-53. Hellman, G. «Determination and logical truth». Journal of Philosophy, vol. 82, n.° 11, 1985, pp. 607-16.

214

References

Hendry, R. F. «Is There Downward Causation in Chemistry?» Philosophy of Chemistry: Synthesis of a New Discipline, edited by D. Baird, E. R. Scerri, L. C. McIntyre. Berlin: Springer, 2006. Hitchcock, C. «Probabilistic Causation». The Stanford Encyclopedia of Philosophy (Winter 2012 Edition), edited by E. N. Zalta. Hoefer, C. «Causal Determinism». The Stanford Encyclopedia of Philosophy (Spring 2010 Edition), edited by E. N. Zalta. Hoffmann, R. «What Might Philosophy of Science Look like if Chemists Built It?». Synthese, n.° 155, 2007, pp. 321-36. Horgan, T. «Mental Quausation». Philosophical Perspectives, n.° 3, 1989, pp. 47-76. Horgan, T. «Kim on Mental Causation and Causal Exclusion». Noûs, vol. 31, n.° 11, Supplement: Philosophical Perspectives, Mind, Causation, and World, 1997, pp. 165-184. Hume, D. [tnh] Tratado de la Naturaleza Humana, traduced by Félix Duque. Barcelona: Orbis, [1740] 1984. Hume, D. [ehu] An Enquiry concerning Human Understanding, edited by T. L. Beauchamp. New York: Oxford University Press, [1748] 1999. Hume, D. [thn] A Treatise of Human Nature, edited by D. Fate Norton and M. J. Norton. Oxford: Clarendon Press, [1740] 2007. Humphreys, P. «How Properties Emerge». Philosophy of Science, n.° 64, 1997, pp. 1-17. Hüttemann, A. «Explanation, Emergence, and Quantum Entanglement». Philosophy of Science, vol. 72, n.° 1, 2005, pp. 114-27. Hüttemann, A. & Papineau, D. «Physicalism Decomposed». Analysis, vol. 65, n.° 1, 2005, pp. 33-9. iasp. Classification of chronic pain: descriptions of chronic pain syndromes and definitions of pain terms. Suppl. 3. Seattle: iasp Press, 1994. Ingvar, M. «Pain and Functional Imaging». Philosophical Transactions: Biological Sciences, vol. 354, n.° 1387, 1999, pp. 1347-58. Ismael, J. & Schaffer, J. «Quantum Holism: Nonseparability as Common Ground», to appear in Synthese. Downloaded from: http://www.jonathanschaffer.org/quantumholism.pdf , forthcoming publication.

215

References

Jackson, F. «Armchair Metaphysics». Philosophy in Mind, edited by M. Michael and J. Hawthorne. Amsterdam: Kluwer, 1994. Jackson, F. From Metaphysics to Ethics. Oxford University Press, 1998a. Jackson, F. «Causal Roles and Higher-Order Properties. Review of Ten Problems of Consciousness by Michael Tye». Philosophy and Phenomenological Research, vol. 58, n.° 3, 1998b, pp. 657-61. Jackson, F. «From Reduction to Type-Type Identity». Philosophy and Phenomenological Research, vol. 65, n.° 3, 2002, pp. 644-47. Jackson, F. «From H2O to Water. The Relevance to a priori Passage». Real Metaphysics. Essays in honour of D. H. Mellor, edited by H. Lillehammer and G. Rodriguez-Pereyra. New York: Routledge, 2003. Jackson, F. «A Priori Physicalism». Contemporary Debates in Philosophy of Mind, edited by B. P. McLaughlin and J. Cohen. Oxford: Blackwell, 2007. Jackson, F. «Response to Barbara Gail Montero». The Fourth Online Consciousness Conference, February 17th-March 2nd 2012, downloaded from: http://consciousnessonline.files.wordpress.com/2012/02/jacksoncomments-on-montero.pdf Juarrero, A. «Causality as Constraint». Evolutionary Systems, edited by G. Van de Vijver et al. Kluwer Academic Publishers, 1998, pp. 233-42. Juarrero, A. «Top-Down Causation and Autonomy in Complex Systems». Downward Causation and the Neurobiology of Free Will, N. Murphy, G. Ellis, & T. O’Connor. Berlin: Springer, 2009. Kauffman, S. The Origins of Order. New York: Oxford University Press, 1993. Kellert, S. In the Wake of Chaos. Chicago: University of Chicago Press, 1993. Kelso, S. Dynamic patterns: the self-organization of brain and behavior. Cambridge, MA: The MIT Press, 1995. Kim, J. «Causes and Events: Mackie on Causation». Reprinted in Causation, edited by E. Sosa, and M. Tooley. New York: Oxford University Press, 1993.

216

References

Kim, J. «Causation, Nomic Subsumption, and the Concept of Event». Reprinted in Supervenience and Mind. Selected Philosophical Essays. New York: Cambridge University Press, [1973a] 1993. Kim, J. «Causes and Counterfactuals». Reprinted in Causation, edited by E. Sosa, and M. Tooley. New York: Oxford University Press, [1973b] 1993. Kim, J. «Noncausal Connections». Reprinted in Supervenience and Mind. Selected Philosophical Essays. New York: Cambridge University Press, [1974] 1993. Kim, J. «Events as Property Exemplifications». Reprinted in Supervenience and Mind. Selected Philosophical Essays. New York: Cambridge University Press, [1976] 1993. Kim, J. «Explanatory Realism, Causal Realism, and Explanatory Exclusion». Reprinted in Essays in the Metaphysics of Mind. New York: Oxford University Press Inc., [1988] 2010. Kim, J. «The Myth of Nonreductive Materialism». Proceedings and Addresses of the American Philosophical Association, vol. 63, n.° 3, 1989, pp. 31-47. Kim, J. «Events: Their Metaphysics and Semantics». Philosophy and Phenomenological Research, vol. 51, n.° 3, 1990, pp. 641-646. Kim, J. «“Downward Causation” in Emergentism and Nonreductive Physicalism». Emergence or Reduction?: Essays on the Prospects of Nonreductive Physicalism, edited by A. Beckermann, H. Flohr, and J. Kim. Berlin: Walter de Gruyter & Co, 1992a. Kim, J. «Multiple Realization and the Metaphysics of Reduction». Philosophy and Phenomenological Research, vol. 52, n.° 1, 1992b, pp. 1-26. Kim, J. Supervenience and Mind. Selected Philosophical Essays. New York: Cambridge University Press, 1993a. Kim, J. «Mental Causation in a Physical World». Philosophical Issues, vol. 3, 1993b, pp. 157-76. Kim, J. «Supervenience». A Companion to the Philosophy of Mind, edited by S. Guttenplan. Cambridge, MA: Blackwell, 1994. Kim, J. «The Mind-Body Problem: Taking Stock After Forty Years». Noûs, vol. 31, n.° 11, Supplement: Philosophical Perspectives, Mind, Causation, and World, 1997, pp. 185-207.

217

References

Kim, J. Mind in a Physical World: An Essay on the Mind-Body problem and Mental Causation. Cambridge MA: A Bradford Book, The mit Press, 1998. Kim, J. «Making Sense of Emergence». Reprinted in Essays in the Metaphysics of Mind. New York: Oxford University Press Inc, [1999] 2010. Kim, J. Physicalism, Or Something Near Enough. New Jersey: Princeton University Press, 2005. Kim, J. «Emergence: Core Ideas and Issues». Synthese, vol. 151, n.° 3, 2006a, pp. 547-559. Kim, J. «Being Realistic about Emergence». The Re-Emergence of Emergence. The Emergentist Hypothesis from Science to Religion, edited by P. Clayton, and P. Davies. New York: Oxford University Press Inc., 2006b. Kim, J. «The Causal Efficacy of Consciousness». The Blackwell Companion to Consciousness, edited by M. Velmans, and S. Schneider. Oxford: Blackwell Publishing Ltd., 2007a. Kim, J. «Causation and Mental Causation». Reprinted in Essays in the Metaphysics of Mind. New York: Oxford University Press Inc., [2007b] 2010. Kim, J. «Mental Causation». The Oxford Handbook of Philosophy of Mind, edited by B. McLaughlin, A. Beckermann, and S. Walter. New York: Oxford University Press, 2009a. Kim, J. «“Supervenient and Yet Not Deducible”: Is There a Coherent Concept of Ontological Emergence?». Reprinted in Essays in the Metaphysics of Mind. New York: Oxford University Press Inc., [2009b] 2010. Kim, J. Essays in the Metaphysics of Mind. New York: Oxford University Press Inc., 2010a. Kim, J. «Two Concepts of Realization, Mental Causation, and Physicalism». Essays in the Metaphysics of Mind. New York: Oxford University Press Inc., [2010b] 2010. Kim, J. Philosophy of Mind. Third Edition. Westview Press, 2011. Kistler, M. «The Necessity of Laws, Quiddity and the Causal Criterion of Reality». Argument und Analyse – Sektionsvorträge. Mentis: Paderborn, 2002.

218

References

Kistler, M. Causation and Laws of Nature. New York: Routledge, 2006a. Kistler, M. «Reduction and Emergence in the Physical Sciences: Reply to Rueger». Synthese, n.° 151, 2006b, pp. 347-54. Kistler, M. «The Interventionist Account of Causation and Noncausal Association Laws». Erkenntnis, n.° 78, 2013, pp. 65-84. Kistler, M. «Interventionism and Non-Causal Dependence», downloaded from: http://www.philosophie.ens.fr/IMG/Interventionism-Autonomy-Special-Sciences4.pdf Klee, R. «Micro-Determinism and Concepts of Emergence». Philosophy of Science, n.° 51, 1984, 44-63. Lange, M. «Natural Laws and the Problem of Provisos». Erkenntnis, n.° 38, 1993, 233-48. Leggett, A. J. The Problems of Physics. New York: Oxford Clarendon Press, 1987. LePore, E. & Loewer, B. «Mind Matters». Journal of Philosophy, n.° 93, 1987, pp. 630-42. LePore, E. & Loewer, B. «More on Making Mind Matter». Philosophical Topics, n.° 17, 1989, pp. 175-92. Levin, J. «Is Conceptual Analysis Needed for the Reduction of Qualitative States?». Philosophy and Phenomenological Research, vol. 64, n.° 3, 2002, pp. 571-591. Levine, J. «Materialism and Qualia: The Explanatory Gap». Pacific Philosophical Quarterly, n.° 64, 1983. Levine, J. «On Leaving Out What it’s Like». Consciousness, edited by M. Davies, and G. Humphreys. Oxford: Blackwell Publishers, 1993. Levine, J. Purple Haze: The Puzzle of Consciousness. New York: Oxford University Press, 2001. Lewes, G. H. Problems of Life and Mind. London: Kegan Paul, Trench, Turbner & Co, 1875. Lewis, D. «Causation». Journal of Philosophy, n.° 70, 1973a, pp. 556-67. Lewis, D. Counterfactuals. Cambridge, MA: Harvard University Press, 1973b. Lewis, D. «Mad pain and Martian pain». Reading in the Philosophy of Psychology, vol. I., edited by N. Block. Cambridge, MA: Bradford Book.

219

References

Lewis, D. «New Work for a Theory of Universals». Australasian Journal of Philosophy, vol. 61, n.° 4, 1983. Lewis, D. «Events». Philosophical Papers 2. Oxford: Oxford University Press, 1986, pp. 241-69. Lewis, D. «Reduction of mind». A Companion to the Philosophy of Mind, edited by S. Guttenplan. Oxford: Blackwell, 1994. pp. 412-431. List, C. & Menzies, P. «Non-Reductive Physicalism and the Limits of the Exclusion Principle». The Journal of Philosophy, vol. 106, n.° 9, 2009, pp. 475-502. Loewer, B. «Mental Causation, Or Something Near Enough». Contemporary Debates in Philosophy of Mind, edited by B. P. McLaughlin and J. Cohen. Oxford: Blackwell, 2007. Lombardi, O. & Dieks, D. «Modal Interpretations of Quantum Mechanics». The Stanford Encyclopedia of Philosophy (Spring 2014 Edition), edited by E. N. Zalta. Luchinsky, D. & Atefanovska, A. «Definition of Nonlinearity». Encyclopedia of Nonlinear Science, edited by A. Scott. New York: Routledge. Lycan, W. Consciousness. Cambridge, MA: mit Press, 1987. Mackie, J. L. «Causes and Conditions». Causation, edited by E. Sosa & M. Tooley. New York: Oxford University Press, 1993 [1965]. Maudlin, T. «Part and whole in quantum mechanics». Interpreting Bodies: Classical and Quantum Objects in Modern Physics, edited by E. Castellani. Princeton, NJ: Princeton University Press, 1998. McLaughlin, B. «The Rise and Fall of British Emergentism». Emergence or Reduction?: Essays on the Prospects of Nonreductive Physicalism, edited by A. Beckermann, F. Flohr, & J. Kim. Berlin: Walter de Gruyter & Co, 1992. McLaughlin, B. «Emergence and Supervenience». Reprinted in Emergence: Contemporary Readings in Philosophy and Science, edited by M. Bedau & P. Humphreys. Massachusetts Institute of Technology, [1997] 2008. McLaughlin, B. «Is Role-Functionalism Committed to Epiphenomenalism?». Journal of Consciousness Studies, vol. 13, 2006, pp. 39-66.

220

References

McLaughlin, B. «On the Limits of A Priori Physicalism». Contemporary Debates in Philosophy of Mind, edited by B. P. McLaughlin and J. Cohen. Oxford: Blackwell, 2007. McLaughlin, B. «Why Rationalization Is Not a Species of Causal Explanation». Reasons and Causes: Causalism and anti-Causalism in the Philosophy of Action, edited by J. D’Oro and C. Sandis. London: Palgrave McMillan, 2012. pp. 97-123. McLaughlin, B. & Bennett, K. «Supervenience». The Stanford Encyclopedia of Philosophy (Spring Edition), edited by E. N. Zalta, 2014. Melnyk, A. A Physicalist Manifesto: Thoroughly Modern ­Materialism. Cambridge: Cambridge University Press, 2003. Menzies, P. «Counterfactual Theories of Causation». The Stanford Encyclopedia of Philosophy (Fall Edition), edited by E. N. Zalta, 2009. Melden, A. I. Free Action. London: Routledge and Kegan Paul, 1961. Mettrie, J. O. Man a Machine, traduced by G. C. Bussey. La Salle, IL: Open Court Press, [1748] 1912. Mill, J. S. System of Logic. London: Longmans, Green, Reader, and Dyer, 1843. Montero, B. «Physicalism in an Infinitely Decomposable World». Erkenntnis, vol. 64, n.° 2, 2006, pp. 177-91. Montero, B. «Must Physicalism Imply the Supervenience of the Mental on the Physical?». The Journal of Philosophy, vol. 110, n.° 2, 2013, pp. 93-110. Montero, B. & Papineau, D. «A Defense of the Via Negative Argument for Physicalism». Analysis, vol. 65, n.° 3, 2005, pp. 233-37. Morgan, C. L. Emergent Evolution. London: Williams & Norgate, 1923. Morris, K. Physicalism and a priori Connections, Realization and Reduction: Two Issues in the Metaphysics of Mind, PhD Thesis in the Program in Philosophy at Brown University. Downloaded from: https://repository.library.brown.edu/studio/item/bdr:11194/ Murphy, N. & Brown, W. Did My Neurons Make Me Do It? Philosophical and Neurobiological Perspectives on Moral Responsibility and Free Will. New York: Oxford University Press Inc. Nagel, E. The Structure of Science. New York: Harcourt, Brace and World, 1961.

221

References

Nagel, T. «What is it Like to Be a Bat?». Reprinted in The Nature of Consciousness, edited by N. Block, O. Flanagan, and G. Guzeldere. Cambridge, MA: mit Press, [1974] 1997. Newman, D. «Emergence and Strange Attractors». Philosophy of Science, n.° 63, 1996, pp. 245-61. Newsome, W. «Human Freedom and “Emergence”». Downward Causation and the Neurobiology of Free Will, edited by N. Murphy, G. Ellis, and T. O’Connor. Berlin: Springer. O’Connor, T. & Wong, H. Y. «The Metaphysics of Emergence». Noûs, n.° 39, 2005, pp. 658-78. O’Regan, J. K. Why Red doesn’t Sound Like a Bell: Understanding the Feel of Consciousness. Oxford: Oxford University Press, 2011. Pap, A. An Introduction to the Philosophy of Science. New York: Free Press, 1962. Papineau, D. «The Rise of Physicalism». Physicalism and Its Discontents, edited by C. Gillett, and B. Loewer. Cambridge: Cambridge University Press, 2001. Papineau, D. «Must a Physicalist be a Microphysicalist?». Being Reduced: New Essays on Reduction, Explanation, and Causation, edited by J. Kallestrup and J. Hohwy. New York: Oxford University Press, 2008. Papineau, D. «Can any Sciences be Special?». Emergence in Mind, edited by C. Macdonald and G. Macdonald. Oxford: Oxford University Press, 2010. Papineau, D. «Causation is Macroscopic but not Irreducible». Mental Causation and Ontology, edited by S. Gibb, R. Ingthorsson, and E. Lowe. Oxford: Oxford University Press, 2013. Pereboom, D. & Kornblith, H. «The Metaphysics of Irreducibility». Philosophical Studies, n.° 63, 1991, pp. 125-45. Pettit, P. «Microphysicalism without Contingent Macro-Micro Laws». Analysis, n.° 54, 1994, pp. 253-57. Polger, T. Natural Minds. Cambridge, MA: mit Press, 2004. Polger, T. «H₂O, “Water”, and Transparent Reduction». E ­ rkenntnis, vol. 69, n.° 1, pp. 109-30. Popper, K. «Natural Selection and the Emergence of Mind». Dialectica, vol. 32, n.° 3-4, 1978, pp. 339-55.

222

References

Primas, H. Chemistry, Quantum Mechanics, and Reductionism. Berlin: Springer-Verlag, 1983. Primas, H. «Emergence in the Exact Sciences». Acta Polytechnica Scandinavica, n.° 91, 1998, pp. 83-98. Price, D. «Psychological and Neural Mechanisms of the Affective Dimension of Pain». Science, vol. 288, n.° 5472, 2000, pp. 1769-1772. Putnam, H. «On Properties». Essays in Honor of Carl G. Hempel, edited by N. Rescher, et al. Dordrecht: D. Reidel, 1970. Reichenbach, H. The Rise of Scientific Philosophy. Berkeley and Los Angeles: University of California Press, 1951. Reutlinger, A., Schurz, G., & Hüttemann, A. «Ceteris Paribus Laws». The Stanford Encyclopedia of Philosophy (Spring Edition), edited by E. N. Zalta, 2011. Robb, D. & Heil, J. «Mental Causation». The Stanford Encyclopedia of Philosophy (Spring Edition), edited by E. N. Zalta, 2013. Rothschild, L. J. «The Role of Emergence in Biology». The ReEmergence of Emergence. The Emergentist Hypothesis from Science to Religion, edited by P. Clayton and P. Davies. New York: Oxford University Press, 2006. Rupert, R. «The Causal Theory of Properties and the Causal Theory of Reference, or How to Name Properties and Why It Matters». Philosophy and Phenomenological Research, vol. lxxvii, n.° 3, 2008. Ryle, G. The Concept of Mind. New York: Routledge, [1949] 2009. Salmon, W. Scientific Explanation and the Causal Structure of the World. Princeton: Princeton University Press, 1984. Salmon, W. «Causality without Counterfactuals». Philosophy of Science, n.° 61, 1994, pp. 297-312. Sarter, M., Berntson, G. G., & Cacioppo, J. T. «Brain Imaging and Cognitive Neuroscience. Toward Strong Inference in Attributing Function to Structure». American Psychologist, vol. 51, n.° 1, 1996, pp. 13-21. Schaffer, J. «The Metaphysics of Causation». The Stanford Encyclopedia of Philosophy (Fall Edition), edited by E. N. Zalta, 2008. Schaffer, J. «Monism: The Priority of the Whole». Philosophical Review, vol. 119, n.° 1, 2010, pp. 31-76.

223

References

Schaffner, K. F. «Genes, Behavior, and Developmental Emergentism: One Process, Indivisible?». Philosophy of Science, n.° 65, 1998, pp. 209-52. Schrödinger, E. «Discussion of Probability Relations between Separated Systems». Proceedings of the Cambridge Philosophical Society, n.° 31, 1935, pp. 555-63. Schröder, J. «Emergence: Non-Deducibility or Downwards Causation?». The Philosophical Quarterly, vol. 48, n.° 193, 1998, pp. 433-452. Scott, A. «On Quantum Theories of the Mind». Journal of Consciousness Studies, n.° 3, 1996, pp. 484-91. Scott, A. «Nonlinear Science and the Cognitive Hierarchy». Evolution & emergence: Systems, Organisms, Persons, edited by N. C. Murphy and W. R. Stoeger. Oxford: Oxford University Press, 2007. Shapiro, L. «Multiple Realizations». The Journal of Philosophy, vol. 97, n.° 12, 2000, pp. 635-54. Shapiro, L. «How to Test for Multiple Realization». Philosophy of Science, n.° 75, 2008, pp. 514-525. Shapiro, L. «Lessons from Causal Exclusion». Philosophy and Phenomenological Research, n.° 81, 2010, 594-604. Shapiro, L. «Mental Manipulations and the Problem of Causal Exclusion». Australasian Journal of Philosophy, n.° 90, 2012, pp. 507-524. Shapiro, L. & Sober, E. «Epiphenomenalism. The Do’s and Don’ts». Thinking About Causes: From Greek Philosophy to Modern Physics, edited by G. Wolters and P. Machamer. Pittsburgh: University of Pittsburgh Press, 2007. pp. 235-64. Shoemaker, S. «Absent Qualia are Impossible». Philosophical Review, n.° 90, 1981, pp- 581-599. Shoemaker, S. «Realization and Mental Causation». Physicalism and Its Discontents, edited by C. Gillett and B. Loewer. Cambridge: Cambridge University Press. Shoemaker, S. Physical realization. New York: Oxford University Press, 2007. Shoemaker, S. «Reply to my Critics». Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition, vol. 148, n.° 1, 2010.

224

References

Silberstein, M. «Converging on Emergence: Consciousness, Causation and Explanation». Journal of Consciousness Studies, vol. 8, n.° 9-10, 2001, pp. 61-98. Silberstein, M. & McGeever, J. «In Search for Ontological ­Emergence». Philosophical Quarterly, n.° 49, 1999, 182-200. Sklar, L. «The Reduction(?) of Thermodynamics to Statistical Mechanics». Philosophical Studies, vol. 95, n.° 1-2, 1999, pp. 187-202. Sosa, E. «Varieties of Causation». Causation, edited by E. Sosa and M. Tooley. New York: Oxford University Press, 1993. Sosa, E. & Tooley, M. Introduction. Causation, edited by E. Sosa and M. Tooley. New York: Oxford University Press, 1993. Sperry, R. W. «Mental Phenomena as Causal Determinants in Brain Functions». Process Studies, vol. 5, n.° 4, 1975, pp. 247-256. Sperry, R. W. «Macro- versus Micro-Determinism». Philosophy of Science, vol. 53, n.° 2, 1986, pp. 265-270. Stalnaker, R. «A Theory of Conditionals». Studies in Logical Theory, edited by N. Rescher. Oxford: Blackwell, 1968. Stalnaker, R. «Varieties of Supervenience». Noûs, vol. 30, n.° 10, Supplement: Philosophical Perspectives, Metaphysics, 1996, pp. 221-241. Stalnaker, R. «What is it Like to be a Zombie». Conceivability and Possibility, edited by T. Szabo Gendler and J. Hawthorne. New York: Oxford University Press, 2002. Stoljar, D. «Distinctions in Distinction». Being Reduced: New Essays on Reduction, Explanation, and Causation, edited by J. Kallestrup and J. Hohwy. New York: Oxford University Press, 2008. Stoljar, D. Physicalism. New York: Routledge, 2010. Strawson, P. F. Individuals: An Essay in Descriptive Metaphysics. London: Methuen, 1959. Strevens, M. «Mackie Remixed». Causation and Explanation, edited by J. K. Campbell, M. O’Rourke, and H. Silverstein. Cambridge: Massachusetts Institute of Technology, 2007. Teller, P. «Relational Holism and Quantum Mechanics». The British Journal for the Philosophy of Science, vol. 37, n.° 1, 1986, pp. 71-81. Thagard, P. «The Self as a System of Multilevel Interacting Mechanisms». Philosophical Psychology, n.° 2, 2012, pp. 1-19.

225

References

Thompson, E. Mind in Life. Biology, Phenomenology, and the Sciences of Mind. Cambridge: Harvard University Press, 2007. Tooley, M. «Causation: Reductionism versus Realism». Philosophy and Phenomenological Research, n.° 50, Supplement, 1990, pp. 215-236. Tye, M. Ten Problems of Consciousness. Cambridge, MA: mit Press, 1995. Vaidman, L. «Many-Worlds Interpretation of Quantum Mechanics». The Stanford Encyclopedia of Philosophy (Spring Edition), edited by E. N. Zalta, 2015. Van Cleve, J. «Emergence vs. Panpsychism: Magic or Mind Dust?». Philosophical Perspectives, edited by J. E. Tomberlin. Vol. 4. Atascadero, Cal.: Ridgeview Publishing Company, 1990. pp. 215-26. Van Fraassen, B. C. «Semantic Analysis of Quantum Logic». Contemporary Research in the Foundations and Philosophy of Quantum Theory, edited by C. A. Hooker. Dordrecht: Reidel, 1973. pp. 180-213. Van Gulick, R. «Who’s in Charge Here? And Who’s Doing All the Work?». Mental Causation, edited by J. Heil and A. Mele. Oxford: Clarendon Press, 1993. pp. 233-256. Van Gulick, R. «Reduction, Emergence, and the Mind/Body Problem: A Philosophic Overview». Evolution & Emergence: Systems, Organisms, Persons, edited by N. C. Murphy and W. R. Stoeger. Oxford: Oxford University Press, 2007. Van Gulick, R. «Functionalism». The Oxford Handbook of Philosophy of Mind, edited by B. McLaughlin, A. Beckermann, and S. Walter. New York: Oxford University Press, 2009. Velmans, M. «How Could Conscious Experiences Affect Brains?». Journal of Consciousness Studies, Special issue, vol. 9, n.° 11, 2002, pp. 3-29. Wimsatt, W. Re-Engineering Philosophy for Limited Beings. Piecewise Approximations to Reality. Cambridge: Harvard University Press, 2007. Wilson, J. «Supervenience-based Formulations of Physicalism». Noûs, vol. 39, n.° 3, 2005, pp. 426-59. Wilson, J. «Non-Reductive Realization and the Powers-based Subset Strategy». The Monist, vol. 94, n.° 1, 2011.

226

References

Wilson, J. «Nonlinearity and Metaphysical Emergence». Metaphysics and Science, edited by S. Mumford and M. Tugby. Oxford: Oxford University Press, 2013. pp. 201-35. Wittgenstein, L. [pi] Philosophical Investigations, edited by G. E. M. Anscombe and R. Rhees. Oxford: Blackwell, 1953. Wolfram, S. Cellular Automata and Complexity. New York: Addison-Wesley, 1994. Woodward, J. «Mental Causation and Neural Mechanisms». Being Reduced: New Essays on Reduction, Explanation, and Causation, edited by Hohwy and Kallestrup. Oxford University Press, 2008. pp. 218-262. Woodward, J. «Interventionism and Causal Exclusion». Preprint, 2011. Downloaded from: http://philsci-archive.pitt.edu/8651/ Woodward, J. «Causation and Manipulability». The Stanford Encyclopedia of Philosophy (Summer Edition), edited by E. N. Zalta, 2013. Woodward, J. & Hitchcock, C. «Explanatory Generalizations, Part 1: A Counterfactual Account». Noûs, n.° 37, 2003, pp. 1-24

227

Index a posteriori: 56-57, 60, 62-68, 78, 106, 113, 115, 156, 172 n. 116, 178, 192, 194. affective nociceptive: 22, 186-187, 191, 193-194, 197. causal: 17-21, 27-29, 34-35, 60, 83, 109,

emergent causation: 21, 23, 72 n. 55, 155-158, 175, 186, 196. epiphenomenalism: 17, 19, 29, 35, 119, 125, 137-139, 153, 171, 198. exclusion argument: 19, 119, 125-127,

119-120, 126, 131, 135, 137, 139-145,

130-131, 136, 140, 152-153.

148-154, 160, 162 n. 167-175, 192-196,

functionalism: 121, 123, 167.

203-206.

fundamentality: 82, 85-86, 93, 110.

-constraint: 21, 160, 162 n. 109, 170, 194, 196. -individuation of natural kinds: 19, 120, 140-141, 151, 153, 167-168.

higher level causation: 39, 48, 51, 127, 131-132, 142, 206. holistic properties: 68-69, 87. internal context: 17, 23, 49, 53-54.

-power(s): 17, 19-21, 27-29, 34-35, 60, 83,

intervention: 46-48, 134, 136-139, 170-172.

119, 126, 131, 135, 137, 139-145, 148-154,

interventionist approach: 19, 119, 135-137,

166-175, 192-195, 203-206. causation: 17, 19, 21-36, 38-41, 43-49,

140, 152-153, 170. inus condition: 32, 47-48.

51-54, 69 n. 49, 72 n. 55, 81, 84, 110,

laws of nature: 25, 29, 35, 38, 45, 103.

112, 116, 119-120, 125-127, 133-137, 142,

local reduction: 198, 202.

153, 155-163, 170, 175, 178-179, 187, 196-

macrophysicalism: 22, 81, 91-92, 178,

197, 206. compositional principles: 105-106, 108, 113-115, 156, 172 n. 116, 178. contingent: 106, 108-109, 113, 120, 153154, 156, 165, 172 n. 116, 178, 192, 194, 200, 202-203. counterfactual dependence: 17, 23, 39-40, 44-45, 48, 58.

206. mathematically expressible behavior: 75, 78, 92. mental causation: 21, 32, 119, 127, 155, 186, 196. metaphysical: 15, 18, 20, 25 n. 5, 28, 38, 55-56, 63-64, 71, 71, 76, 86, 92-93, 95, 97-100, 103-105, 109, 112, 114, 116-117,

deducibility: 82-84, 88-90.

120-122, 124-125, 129, 131, 152-154, 197,

downward causation: 17, 21-22, 116,

203, 205-206.

126-127, 131, 156-162, 175, 179, 196-197. dualism: 16, 18, 28 n. 10, 59, 77, 81-82,

-dependence: 93, 98-100, 103-104, 109, 116.

92-93, 100-104, 109, 111, 117, 128, 139.

-supervenience: 18, 20, 55, 64, 71, 92-93,

eliminativism: 19, 120, 154, 198, 202

104-105, 109, 112, 114, 117, 120, 122,

n. 127.

152-154, 197, 203, 205-206.

229

Juan Diego Morales Otero

microphysicalism: 10, 11, 63-64, 66-69, 73, 91, 93, 98, 100, 102, 109, 111, 154, 164-165, 177, 197, 205. multiple realizability: 94, 96, 120-121, 124, 202.

quantum entanglement: 69 n. 49, 115, 162 n. 109, 175-176, 186. second order properties: 95, 121, 132, 152, 203. spatiality: 76.

natural properties: 19, 151, 153, 168, 203.

spin up/down: 179-180.

nomological regularities: 17, 23, 34,

subset account of realization: 140, 143-

38, 54. non-reductive physicalism: 17-19, 81-82, 84, 92-97, 111-112, 116, 119-121, 206.

149, 167. supervenience: 18, 20, 22, 35, 55, 58-59, 62, 64-66, 68-69, 71-73, 79, 92-94,

non-rigid designators: 203.

97, 99, 102-105, 108-109, 111-112, 114-

objective phenomena: 75-76.

115, 117, 120-123, 126-127, 131, 135-138,

ontological emergence: 18, 81-90, 92,

152-154, 156-158, 167, 169, 175, 178, 197,

112, 114-116, 175. physical causal closure: 21, 127-128, 163-164. physicalism: 15-20, 55-69, 72-79, 81-87, 92-97, 101-102, 109, 111-112, 114, 116, 119121, 123, 128, 139, 153, 164-166, 205-206. probability: 46-48, 51, 53, 128-132, 153, 159-164, 168-170, 173-174, 179-182, 184185, 195. property dualism: 18, 82, 93, 100-104, 109, 111, 117.

230

203, 205-206. token physicalism: 18, 82, 93, 101-102, 128. transference of a physical quantity: 23, 54. trans-ordinal laws: 105, 108, 167 n. 114, 169. under-determination: 21, 158.

The Emergence of Mind in a Physical World e di ta d o p or e l C e n t ro E di t or i a l de l a Fac u lta d de c i e n c i a s h u m a n a s de l a u n i v e r s i da d n ac io n a l de c ol om bi a , F or m a pa rt e de l a bi bl io t e c a a bi e rta , c ol e c c ió n g e n e r a l , s e r i e f i l o s of í a . e l t e x t o f u e c om p u e s t o e n c a r ac t e r e s m i n io n y f ru t ig e r . s e u t i l i z ó pa p e l H ol m e n b o ok de 70 g r a m o s y, e n l a c a r át u l a , C l a s s ic L i n e n N at u r a l W h i t e DE 2 16 g r a m o s . e l l i bro s e t e r m i n ó de i m p r i m i r e n b o g o tá , e n dig i p r i n t e di t or e s s a s , e n f e br e ro de 2 01 8.