Gender, Language and New Litera
This page intentionally left blank
Gender, Language and New Literacy
Edited by Eva-Maria Thune, Simona Leonard! and Carla Bazzanella
continuum
Continuum The Tower Building 11 York Road London SE1 7NX
80 Maiden Lane Suite 704 New York NY 10038
© Eva-Maria Thune, Simona Leonardi, Carla Bazzanella and contributors 2006 All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Eva-Maria Thune, Simona Leonardi and Carla Bazzanella have asserted their right under the Copyright, Designs and Patents Act, 1988, to be identified as Authors of this work. First published 2006 British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library. ISBN: 978-0-8264-3218-6 Library of Congres s Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress. Typset by RefineCatch Limited, Bungay, Suffolk Printed and bound in Great Britain by Biddies Ltd, King's Lynn, Norfolk
Contents
List of tables
vii
Contributors
viii
1 Introduction Carlo Bazzanelia, Eva-Maria Thiine and Sinuma Leonardi
1
2 Czech: friendly to women? Svetla Cmejrkovd
24
3 Treatment of 'woman' and 'man' in the Dutch Word Thesaurus Ingrid van Alphen and Akssandra Corda
41
4 Gender in the English Word Thesaurus Camilla Bettoni
62
5 The German Word Thesaurus and socio-cultural models Eva-Maria Thiine and Simona Leonardi
79
6 Gender construction and the Word Thesaurus for Greek Marianna Katsoyannou and Dionysis Goutsos
95
7 Gender on-line in the Italian Word Thesaurus Manuela Manera and Carlo Bazzanelia 8 The lexicographical representation of gender in the Microsoft Word Thesaurus: Polish Johanna Miecznikowski 9 The perspective on gender in the Portuguese Word Thesaurus Maria Aldina Marques 10 Word's Spanish Thesaurus: some limits of automaticity Pura Guil
107
123
139 153
vi
CONTENTS
11 Gender on-line in Hebrew: new technology, old language ZoharLivnat
169
12 Lexical gender in Hungarian: reflections of social structures in an electronic corpus Marcell Nagy and Viviana Patti
182
13 A world of difference: gender disparity in the Turkish Word Thesaurus Marina Castagneto and Rosita D'Amora
196
14 Gender in Chinese and new writing technologies Antonetta Ceccagno Index
213 231
List of tables
1.1 'Meanings' and 'synonyms' in the English Word Thesaurus
11
4.1 4.2 4.3 4.4
69 71 72 74
Comparison in the treatment of two entries in three thesauri Quantitative comparison between Word 6 and Word 11 Qualitative comparison between Word 6 and Word 11 Qualitative comparison between US English and UK English
5.1 Lexical gender in German 5.2 Non-gender-specific tenns in German
79 80
7.1 The pair compagno - covipagna in the Word Thesaurus 2002
114
9.1 Morphemic variation in masculine and feminine nouns in Portuguese 9.2 Gender as a syntagmatic process marked by determiners
140 141
11.1 Masculine and feminine pronouns in Hebrew
171
12.1 Unmarked and marked profession nouns in Hungarian 12.2 Occurrence of MPN-creating methods in the nouns of the type tandr' teacher' 12.3 Frequency of various female profession nouns at http://www.google.co.hu 12.4 Occurrence of MPN-creating methods in the nouns of the type miniszterelnok 'prime minister' 12.5 Frequency of various female profession nouns at http://www.google.co.hu 12.6 Occurrence of MPN-creating methods in the nouns of the type pilota 'pilot'
183
14.1 Hownet en tries including nu
185 185 185 186 186 223
Contributors
Carla Bazzanella is Professor of Linguistics at the University of Turin, Italy. Her main areas of research are interactional linguistics, pragmatics and gender studies, e-mail:
[email protected] http://hal9000.cisi.unito.it/ wf/DIPARTIMEN/Disciplinel/Professori/Carla-Bazz Camilla Betton i is Professor of Applied Linguistics at the University of Verona, Italy. Her main research areas are language contact, language maintenance and language shift, language acquisition and language attrition, and language education, e-mail:
[email protected] http:// www.lingue.uniw.it/fol/main?ent=persona&id=853&lang=en Marina Castagneto is a Researcher in Applied Linguistics at the University of Cagliari, Italy. Her main areas of research are conversation analysis and lexical semantics in Turkish, e-mail:
[email protected] http:// www.unica.it/ dip-ling/ricercatori/ castagneto. htm Antonella Ceccagno is Professor of Chinese Language and Literature at the University of Bologna, Italy. Her main areas of research are Chinese language and linguistics, and migration from China/the Chinese diaspora in Europe, e-mail:
[email protected] Svetla Cmejrkov a is a Researcher at the Czech Language Institute of the Czech Academy of Sciences. Her main research areas are discourse studies and sociolinguistics. e-mail:
[email protected] http://www.ujc.cas.cz/ lide/people.php?man=cmejrkova Alessandra Corda coordinates the ICT educational projects in the Faculty of Arts, Leiden University, The Netherlands. She has published in the fields of computer assisted language learning, foreign language teaching, and lexicography and was involved in computer-based lexicographic projects, e-mail:
[email protected] http://www.ilo.uva.nl/Ontwikkeling/imictil/ persons/ Alessandra.html Rosita D'Amor a is a Lecturer in Turkish Language and Literature at the University of Naples 'L'Orientale', Italy. Her main areas of research are Turkish as a foreign language, contemporary Turkish women writers, and Ottoman social history, e-mail:
[email protected]
CONTRIBUTORS
ix
Dionysis Goutsos is Assistant Professor in Textlinguistics at the University of Athens, Greece. His main areas of research are discourse analysis, corpus linguistics and translation studies, e-mail:
[email protected] http:// www.cc.uoa.gr/linguistics/PP/Goutsos/indexEN.html Pura Gui l is Professor of Linguistics at the Complutense University of Madrid, Spain. Her main areas of research are Italian-Spanish contrastive linguistics, textual linguistics and pragmatics, e-mail:
[email protected] Marianna Katsoyannou is Assistant Professor of General Linguistics at the University of Cyprus, Dept. of Byzantine and Modern Greek Studies. Her main areas of research are general linguistics, lexicography, LSP translation & terminology, e-mail:
[email protected] Simona Leonard! is Associate Professor of Germanic Philology at the University Federico II of Naples, Italy. Her main research areas include historical semantics, pragmatics and cognitive linguistics, e-mail: simona.leonardi @unina.it http://www.docenti.unina.it/simona.leonardi Zohar Livnat is a Senior Lecturer in the Department of Hebrew and Semitic Languages at Bar-Han University, Israel. Her main areas of research are semantics, pragmatics and discourse analysis. She is the editor of the journal Hebrew Linguistics, e-mail:
[email protected] http://www.biu.ac.il/JS/hb/ zlivnathtml Manuela Manera is a PhD student at the University of Turin, Italy. Her main areas of research are history of the Italian language, linguistics and Italian literature, e-mail:
[email protected] Maria Aldina Marques is Professor of Portuguese Linguistics at the Institute de Letras e Ciencias Humanas at the University of Minho, Braga, Portugal. Her main areas of research are pragmatics and discourse analysis, e-mail:
[email protected] Johanna Miecznikowsk i is a Guest Researcher at the University of Turin, Italy, with a postdoctoral fellowship of the Swiss National Foundation of Scientific Research. Her main areas of research are interactional linguistics, plurilingualism and the pragmatics of scientific discourse, e-mail: Johanna.
[email protected] Marcell Nagy is a PhD student in Romance Linguistics at the University ELTE in Budapest, Hungary. His main area of research is diachronic morphology of Italian, e-mail:
[email protected] Viviana Patt i is a Researcher in the Computer Science Department at the University of Turin, Italy. Her main research interests include computational logic, communication protocol design and implementation, e-learning and personalization by reasoning in the semantic web. e-mail:
[email protected] http://www.di.unito.it/~patti/ Eva-Maria Thune is Professor of German Linguistics at the Faculty of Foreign
x
CONTRIBUTORS
Languages and Literature at the University of Bologna Italy. Her main areas of research are German as a foreign language, conversation analysis and gender studies, e-mail:
[email protected] http://www.lingue. unibo.it/linguaparlata Ingrid C . va n Alphen is Assistant Professor of General Linguistics at the University of Amsterdam, The Netherlands. Her main areas of research are interactional sociolinguistics, pragmatics & gender studies, e-mail:
[email protected] http://home.medewerker.uva.n1/i.c.vanalphen/
1 Introduction Carlo, Bazzanella, Eva-Maria Thune and Simona Leonardi
1 Gender , language and culture in new technologie s Computer technology has brought with it a whole series of changes, not just those that are very visible in the world around us. More intimate cognitive practices such as reading and writing have been affected by these radical transformations and as a matter of fact, in recent years, increasingly studies (see e.g. Gee 1996; Lankshear 1997; Leu et al 2004) have stressed that the notion of literacy itself needs to be reconceived; we need to recognize the existence of a 'new literacy', which actually encompasses a plurality of literacies (Unsworth 2002:63). It is thanks to sufficient competence in these multiliteracies (see Cope & Kalantzis 2000) that one is able to adequately cope with the various textual practices we are constantly confronted with (traditional print media and CD-ROMS, e-mails, Internet, etc.). The very way we write has completely changed since the use of word-processing programs became widespread, because, for example, these allow 'cut £ paste' techniques, which not only permit a constant manipulation of our own text, but also make it a straightforward matter to insert other texts. Moreover, various reference and research tools, such as thesauri, dictionaries or spell-checkers, are now available in the most common wordprocessing programs. The use of such tools in writing practice is obviously not new, since there have always been plenty of print dictionaries, thesauri, etc. What is new is the ease with which these tools may now be used, involving no more than a couple of clicks with the mouse; also new is the simple fact that these tools are available to everyone: anybody writing with these programs can access them, while the use of the 'traditional' dictionary or thesaurus was more often than not something that required a trip to the local library (for those in the habit of using such facilities). As everyone knows, the tools created by new technologies offer the advantage of immediate and apparently unproblematic use and this means that users tend to have a sort of reflexive trust in them: they tend to accept them as objective and efficient and thus as a positive innovation. This presumption of objectivity and efficiency in turn means that by actually using the information supplied by these devices, users are in danger of not giving adequate rein to their critical faculties: they tend in fact to use the information in a
2
GENDER, LANGUAGE AND NEW LITERACY
more or less automatic fashion, something which the sheer simplicity of use only encourages. Of course, this practical simplicity is more often than not only apparent, and may indeed be dangerously deceptive, because in the end it has been gained at the price of the systematic omission of appropriate context and examples. As Romaine (2001: 160) states, [T] he connotations of words do not arise from words themselves, but from how they are used in context. The meanings of words are constructed and maintained by patterns of collocation. Collocations transmit cultural meanings and stereotypes which have built up over time.
This acritical use of recent technology appears all the more dangerous if one considers the full implications of the fact that (as we will see) highly advanced technological devices actually end up conveying - and thus helping to perpetuate - a traditional or even 'reactionary' socio-cultural picture (see below, §4). The aim of the present work is to provide a multilingual and cross-cultural approach to the problem of how gender is lexically and socially categorized in certain electronic word-processing software. We will approach the problem both from a theoretical point of view (with regard to the cross-cultural gender issue), and from an applied one. For all the languages we will consider (with the exception of Chinese), our focus will be on die Microsoft Word Thesaurus, while, in the case of Chinese, the software we will be mainly dealing with is Microsoft Pinyin IME 2003, which allows a standard keyboard to be used for the Chinese non-alphabetic writing system and which has a database of lexical and phrasal entries (Ceccagno: this volume1). The various contributions focus on gender categorization in the Thesaurus of Microsoft JVorrf/Microsoft Pinyin. While the 'traditional', printed thesaurus is a 'reference book in which lists of words with similar meanings are grouped together' (Collins Cobuild 2003, s.v. 'thesaurus'), the Thesaurus of Microsoft Word, like other electronic thesauri, is a tool that exploits an internal database to make suggestions as required by the user. More specifically, our main goal is to describe gender categorization in different languages - both those that are closely related from a typological and cultural point of view and those that are not - as this is represented in the Thesaurus database, and furthermore to reflect on how (in the case of each language) the categorizations we find are related to culture2 and societv (e.g. socio-economic roles, as reflected in the names of various professions, see below, §4). Our work belongs to the (by no means fully consolidated) tradition of gender studies, and it is specifically related to works such as Hellinger (1990), Hellinger & BuBmann (2001-3), Romaine (1999), Unterbeck & Rissanen (2000) (to name but a few); what distinguishes our work, however, is the focus on how women and men are characterized in new software programs.
INTRODUCTION BAZZANELLA ET AL.
3
The substantive contribution of this research is intended to be threefold: 1 It aims to analyse the complex and stimulating semantic/pragmatic relationship between the lexicalization of roles and the social construction of gender identity. 2 It aims to provide a comprehensive picture of the social and cultural construction of the female and male roles in different languages and societies, by exploring how women and men are semantically categorized in new technologies, compared both to traditional ones (printed dictionaries and thesauri) and to gender categorization in actual societal settings. 3 It offers an opportunity to make readers and scholars aware of the different ways of categorizing gender identities in software programs, suggesting how the best use may be made of such programs. In the analysis of these tools, we keep in mind what Romaine (1999:117) said about dictionaries, i.e. that they are 'hardly a neutral storehouse of words and their meanings', because they inevitably reflect the society in which they were compiled, its main beliefs and its social structure. We are, of course, well aware of the danger of basing conclusions about the actual representation of societies on a restricted database (i.e. the Word Thesaurus and, for Chinese, the Microsoft Pinyin IME 2003 database); we nevertheless believe that these data can offer a basis for a better understanding of this complex and multifarious domain, i.e. the gender belief system (Braun 2001: 287) .3 At the same time, we think that there may be a real chance of exercising some practical influence on future versions of these programs - an influence that may eventually result in substantial improvements and thus more satisfactory programs. 2 Language , culture, and gender 2.1 Generalissues Without entering into the so-called Sapir-Whorf question, i.e. the relativity hypothesis (see Sapir 1921; Whorf 1956), the reciprocal influence of language and culture cannot be denied: 'the language habits of people are by no means irrelevant as unconscious indicators of the more important traits of their [social] personalities' (Sapir 1933:17). This is all the more so with regard to gender issues, which are both reflected in and reinforced by language use. In Romaine's (2001: 170) words: 'Language plays an active role in the symbolic positioning of women as inferior to men. It both constructs and perpetuates that reality, often in obvious ways, but at other times in subtle and invisible ways.' Different lexical organization can affect not only the way we speak, but also the way we think. Our lexicon is strictly related to our experience, which, in a way, moulds and conceptually organizes it (see Givon 1989; Violi 1997;
4
GENDER, LANGUAGE AND NEW LITERACY
Wierzbicka 1997; Dirven & Verspoor 1998). Is it a matter of chance, for instance, that kadtn 'woman' in Turkish is a problematic, if not deprecatory, designation (Braun 2001: 304), or that, in the Turkish Thesaurus the only meaning given for kadtnhk 'womanhood' is 'virginity' (see Castagneto & D'Amora)? As cognitive linguistics has made clear, semantic modulation is often a function of specific cultural models, which correspond to the frames associated with particular words and concepts; for example, in English the value of strong in a strong woman differs from the value we typically attribute to the same adjective in the expression a strong man: if referred to a woman, it is usually interpreted in terms of moral strength, while, if referred to a man, in terms of physical strength. As Lee (2001:60) states, specifically with regard to woman and man: There is a great deal of cultural modelling associated with terms such as man and woman, and a significant part of such belief systems comes into play in the interpretation of expressions containing these words.' This 'typical' association is explained by Lee (ibid.) on the basis of the fact that men are physically stronger than women, with the result that tasks that require physical strength are traditionally performed more often by men than by women. Consequently, physical strength is more highly prized in men than in women. Since physical strength belongs primarily to the male rather than the female domain, the word stwngin a strongman is naturally [??] interpreted in those terms.
While this particular argument is perhaps questionable, the relation between semantic modulation and cultural models (in this case, the English one) is, in our opinion, undeniable. One very long-established practice that tends to interfere with the perception of women's role in society is the use of the so-called masculine generics, or androcentric generics, i.e. the use of masculine gender terms to refer to groups of people that include women. A distinction has been made (Holmes 2001: 124) between generics in a narrow sense (i.e. where the intended referent is clearly 'humankind' rather than a specifically male human being, e.g. fearful of what man can do to man, the right of man to life) and generics in a broad sense (often called 'pseudo-generics', or 'false generics', see Heliinger & BuBmann 2001: 9), as in the taxman, the man in the street, where the form man actually signals 'male' in the minds of many speakers. More specifically, Holmes (2001: 133) used the following criteria to establish the 'broad' category of generics: 'i) substituting person would not lose relevant information about the referent; ii) substituting person would make it clear that this referent could be female, and that there was nothing inherently masculine about the reference to justify the use of man.' In fact, psycholinguistic tests (see, for example, Martyna 1978; Silveira 1980; Hardin & Banaji 1993, 1996; Braun et al 1998, 2005; for a survey, see Klann-Delius 2005: 49^-55) have shown that generic male terms are not really generic, but inevitably convey a male bias. 'It seems that, at least among
INTRODUCTION BAZZANELLA ET AL
5
Western populations, the male is perceived as the normative, and the use of the pronoun he is literally assumed to refer to a male referent' (Halpern 2000: 32). In English, this applies to the pronoun teand to the term man as well. Even in the case of the term people, Silveira (1980) was able to detect a degree of male bias. Analogous forms are found in many languages: see the Italian use of uomo/uomini to denote either specifically male beings or alternatively human beings, irrespective of gender (see Marcato & Thune 2002: 201). The standard defence of the use of masculine generics is based on the idea that the use of these forms does not involve any intention to exclude women. But the fact of the matter is - as our studies will confirm - that the strong grammatical and semantic asymmetries inherent in masculine generics both construct and maintain the 'male way of perceiving and organizing the world' (Romaine 2001:156). 2.2 Categorizing gender
Grammatical gender and sexual gender are actually different categorizations, although they are often confused and can intertwine: grammatical gender is not necessarily linked to the opposition female vs male, and is thus a broader category, encompassing not only the opposition female-male, but also the one between animate and inanimate. In languages such as those of the Romance family, for example, where there are, with the exception of Rumanian, two grammatical genders (masculine and feminine), the grammatical gender categorization seems to correspond very closely to the sexual categorization. While gender-construction comes about by various means, a distinction can be traced between three different gender categorizations, that is grammatical, lexical and social gender construction. What is particularly interesting is that languages totally devoid of morphological gender (e.g. Chinese, Turkish) can also construct gender-related messages through various linguistic means (for Chinese, see Ceccagno; for Turkish, see Castagneto & D'Amora). In such cases it is common to speak of covert gender. Covert gender includes: • grammatical gender-marking: in the so-called 'gender languages' or 'languages with grammatical gender', every noun, whether it relates to an animate or an inanimate entity, belongs to one of two 'gender classes' (feminine, masculine/common gender, neuter) or three (feminine, masculine, neuter) (e.g. in Italian lafiglia (f.) 'the daughter' - ilfiglio (m.) 'the son', but also il cucchiaio (m.) 'the spoon' - laforchetta (f.) 'the fork'). As Hellinger & Bufimann (2001-3: 7) pointed out, 'unlike case or number, grammatical gender is an inherent property of the noun which controls agreement between the noun itself (the controller) and some (gendervariable) satellite element (the target), which may be an article, adjective, pronoun, verb, numeral or preposition'. Among the linguistic means used to mark gender we find:
6
GENDER, LANGUAGE AND NEW LITERACY
• morphological phenomena such as agreement and anaphoric determiners, gender-marking suffixes (e.g. in Italian: la (f.) potinma (f.) rossa (f.) 'the red armchair' - il (m.) tavolo (m.) rosso (m.) 'the red table'; la (f.) zia (f.) 'the aunt' - lo (m.) zio (m.) 'the uncle'), formatives in compounds (e.g. -frau in German: Feuerwehrfrau '(female) fire-fighter' Feuerwehrmann'(male) fire-fighter'); • pronominal forms (e.g. she/he); • classifiers, i.e. separate that specify class membership in certain contexts (as in Japanese). • lexical gender, in the first place, this concerns animate and therefore sexed endues, but it extends to all entities which are culturally perceived as carrying the semantic property [+ female] or [+ male], 'which may in turn relate to the extra-linguistic category of referential gender (or "sex of referent")' (Hellinger & BuBmann 2001-3: 7). A clear example of this is to be found in English with daughter- son or cow- butt (see Bettoni). On the other hand, there are nouns which are lexically non genderspecific and can thus refer both to female and male animate entities, e.g. in German die Person 'the person' (grammatically feminine), but derSdugUng 'the newborn" (grammatically masculine) and das Individuum 'the individual' (grammatically neuter) (seeThune &Leonardi). The relationship between lexical and grammatical gender is actually quite complex and although correspondences can certainly be found, they are not as direct and clear as is often assumed. There is also another kind of gender: • social gender, this is based on stereotypical cultural and social assumptions about the respective roles of women and men in society. It is manifested, for example, in widely held expectations about an architect being a man, or a secretary being a woman (see Hellinger & BuBmann 2001-3: 11). As Rosch (1973) pointed out, the correlation between degrees of category membership and the frequency and order with which category members are named is significant: if people are asked to name exemplars of a category, they tend to mention the more prototypical members first (see Taylor 1989: 48). In the discussion about gender, the concept of doing gender has become more and more important; it was originally pointed out by Goffman (1977) that gender is not just an extralinguistic variable which influences linguistic production, since individuals actually 'perform' assigned gender roles, something which is in turn necessary for social interaction. Later it was suggested (West &r Zimmermann 1991) that gender differences are constructed rather than being simply 'given', and that they are continually reinforced. Gender rules are also clearly laid down and individuals have to obey and 'perform' the rules in order to be accepted as valid members of society. West & Zimmermann regard the performance of gender as a constant feature of every situation, because, in their opinion, every person - in whatever situation - has to act either as a woman or as a man; more
INTRODUCTION BAZZANELLA ET AL.
7
recently, Susanne Gunthner (1997) has argued for a more nuanced view: everybody decides on the basis of the specific situation they find themselves in whether gender issues are relevant or not. The norm is always established on the basis of the male, not only from a purely grammatical point of view, as in the principle of servitude grammaticale ('grammatical subservience'), which results in masculine agreement prevailing over feminine, or in the predominance of masculine generics (see above, and also Livnat's contribution on Hebrew). The social hierarchy is both reflected and constructed in 'sayings, proverbs, fairy tales and stereotypical expressions ..., which build on and reproduce a number of sexist stereotypes placing woman in an inferior place' (Katsoyannou & Goutsos). A further area of sexist usage deserving investigation is the use of metaphorical reference terms for women and men. In a pilot study in New Zealand focusing on terms used to refer to women and men as sex objects, the respondents (a group of students) provided for women the terms tort, fox, crumpet, and for men stallion, goat, stud, studmuffin. Clearly, terms such as tart are condemnatory and carry the added derogatory meaning of someone who provides sex for payment; things are different with stallion and stud, which might superficially be regarded as equivalent terms for males (since they too focus on intense sexual activity) but which in actual fact are not strongly pejorative (indeed, they generally convey some degree of admiration, see Holmes 2001:129). With studies of this type a degree of caution is always in order, as Graddol & Swann (1989: 111) have pointed out: different results might well be obtained from different speakers in different settings. As an illustration of how fluid things may be, they point to the fact that in recent years female students in US colleges have been reported as using terms such as bitch, whore and slut - usually assumed to be feminine - to refer to men. 2.3 Linguistic strategies towards visibility of women
As a consequence of the widespread male hegemony, women tend to become invisible, or 'deviant', in a way that is suggested by the following excerpt, taken from Austin (1990: 282), quoted in Holmes (2001: 119): Context: Bank manager to female customer A. Now that will be Miss, won't it?
B. No, Ms. A. Oh, one of those.
Actually, the title Ms was first promoted in the 1970s as an alternative to both Mrs and Miss, and was thus intended to be symmetrical to Mr- as a way of avoiding obligatory characterization of women according to their (often only presumed) marital status. Three decades later, though it has gained some
8
GENDER, LANGUAGE AND NEW LITERACY
currency, its uses and social meanings, whether as a term of address or reference or as a means of self-identification, are actually far from transparent (see Holmes 2001; Pauwels 2001; Bettoni). Romaine (2001:169) underlines that it 'is being used in ways its proposers never intended, to maintain the very distinctions it was supposed to replace'; in fact, it has not replaced the odier two terms but exists alongside them. Furthermore, paradoxically, it reflects identity and ideology, both at an individual and at a societal level: most married women prefer to be addressed as and to identify themselves as Mrs, because being married is still seen, even by women, as more prestigious than being unmarried. Ms tends therefore to be regarded as indicating that a woman is single, separated, divorced, widowed or in a de facto relationship, or simply that she is a feminist who thinks that her marital status is her own business and does not need to be disclosed publicly. As Ritchie Key (1996: xiv) rightly states: 'Male and female behaviour can only be understood within the matrix of the power structure of the total society', and linguistic changes are inescapably intertwined with social changes (see below, §5). With regard to the general linguistic policy (see the social outlines of the different countries presented in this volume), it is essential to obtain greater linguistic visibility for women, by moving in the two possible directions: 1 By further extending what has been called the 'feminization' of language (engendering or regendering: see Hellinger 1990; Hellinger & Bufimann 2001-3; Romaine 2001: 156; Bettoni), e.g. through the use of explicit gender markers (as, for example, in Spanish lajueza (f.) 'the (female) judge' vs el/la juez 'the judge' (m./f.) - see Guil), by the use of derivation (as, for example, in Czech itcitel 'male teacher' —» ucitelka 'female teacher'; wmisfr'male minister' —> wmisfrvn/'female minister', see Cmejrkova), or by using split forms, i.e. the use of a double form (feminine and masculine) both for nouns - for example, in Dutch docent/e 'university lecturer' (m.' f.) - and for personal pronouns - e.g. s/he. 'In the case of split forms, while this often tedious practice is certainly at times necessary to clarify reference, it tends to be received with scepticism through being associated with a certain brand of political rhetoric, heard especially at election times and aimed at the ballot box' (Guil). Also in Italian, splitting is rarely used (see Manera & Bazzanella). The problem with engendering is that the unmarkedness of the male forms tends to persist (see above §2.1 for the widespread use of the masculine generics; see in detail the case of Czech, Cmejrkova). Even worse, in Greek 'whereas different forms can often be created for the feminine gender, these forms are socially stigmatized, since they come from "low" social dialects and are used by uneducated people' (Katsoyannou & Goutsos). 2 By fostering gender neutralization (degendering), i.e. promoting the use of neutral terms like 'person' in 'gender languages' (e.g. Greek, Hebrew. Italian,4 Czech), or of the plural, as for example in Dutch, where - as Gerritsen (2002: 89) notes - plurals such as personen or mensen 'people' are
INTRODUCTION BAZZANELLA ET AL.
9
sometimes used as formatives in compounds, as in zeemensen 'sailors (lit. sea-people)', or as in German, where the substantivized participle forms (present and past) are available. These are semantically neutral in the plural and are therefore especially suited to generic reference (as in the case of die Studierenden 'students', which includes both Studenten 'male students' and Studentinnen 'female students', see Thune & Leonardi). But, as Bettoni points out, this option runs the risk of making women even more invisible. 3 Communicatin g gender in computer writing programs 3.1 The starting point The starting point for the present research was a previous study (Bazzanella et ai 2000), where four languages were examined: French (Catherine Camugli-Gallardo), English (Paloma Tejada), Italian (Manuela Manera & Carla Bazzanella), and Spanish (Pura Guil). The focus was on the analysis of 'meanings'5 and 'synonyms'6 which are provided by the Word Thesaurus in relation to a series of lexical pairs with special relevance to the linguistic categorization of woman and man. The data obtained in this study were surprising (in the negative sense), particularly if one considers everything that has happened in western countries in recent years: the feminist movement, the huge amount of research on language and gender, the whole discourse on political correctness and par condicio. The main surprise came from the striking tension that emerged between advanced technology, on the one hand, with all the natural expectations of innovation which this tends to arouse, and, on the other, the social-cultural picture actually conveyed by this technology, which is redolent of much more traditional views. On further reflection, the most worrying aspect of these negative results concerns the future: there is an evident danger that these electronic Thesauri will end up propagating an asymmetric categorization and representation of the roles of women and men in our societies. In view of this risk (which could be avoided by partially changing the Thesauri themselves, see below), and recognizing the obvious need to expand the scope of the research so as to encompass a wider and typologically more heterogeneous corpus (which also means referring to more widely dissimilar cultures, languages, countries), we decided to seek the collaboration of a range of additional co-authors, scholars who are either native speakers of the languages we are interested in or who have a high degree of second language competence in those languages; our aim is to obtain a more complete picture of relations between gender, language and cultural issues, and thus contribute to a better awareness of the problems involved, something which we venture to hope might lead to improvements in the thesauri themselves. As has frequently been pointed out, changing the language cannot in itself improve the situation of women in society; however, there can be no doubt that language has an enormous potential to help correct the imbalance
10
GENDER, LANGUAGE AND NEW LITERACY
associated with male bias, and that it can contribute to the construction of a more positive female identity (see Holmes 2001: 131). 3.2 The present research
The languages represented in our present research, which - as with the preceding one7 - is meant to be qualitative rather than quantitative, are: Chinese, Czech, Dutch, English (i.e. American, British and Australian English), German, Greek, Hebrew, Hungarian, Italian, Polish, Portuguese, Spanish, and Turkish. From a typological point of view, the following families are represented: • the Indo-European family. Czech (Svetla Cmejrkova), Dutch (Ingrid van Alphen, Alessandra Corda), English (Camilla Bettoni), German (Simona Leonardi, Eva-Maria Thune), Greek (Dionysis Goutsos, Marianna Katsoyannou), Italian (Carla Bazzanella, Manuela Manera), Polish (Johanna Miecznikowski), Portuguese (Maria Aldina Marques), Spanish (PuraGuil); • the Semitic family. Hebrew (Zohar Livnat); • the Finno-Ugric family. Hungarian (Marcel Nagy, Viviana Patti); • the Altaic family. Turkish (Marina Castagneto, Rosita D'Amora); • the Sino-Tibetan family. Chinese (Antonella Ceccagno). Given that in the present study the same theoretical frame (i.e. a functional semantic frame focusing on the lexicon) is used to analyse all the different languages examined, we venture to hope that the reader will find her/his task somewhat easier than would otherwise be the case. Furthermore, an effort has been made to ensure that the studies in this volume take account of the different cultural settings of the various languages, and devote considerable attention to cross-language and cross-cultural comparison; this should make it possible for the reader to get a good idea of the general issues, while at the same time allowing him/her to weigh the specific contribution of each language/culture, given the different linguistic resources and cultural characteristics that each brings with it. All the chapters are original contributions and follow common (necessarily flexible, in view of the various language systems and social settings involved) guidelines and terminology, namely: 1 a general outline of gender-marking and gender-construction in each language, i.e. more specifically, a brief outline of grammatical, lexical and social gender-construction (see above § 2.2); 2 a short view of language reform concerning gender matters where such reform has taken place, together with a survey of the achievements of the feminist movement and a sketch of the social structure of the society/ societies8 corresponding to the language under examination and social roles attributed to women/men;
INTRODUCTION BAZZANELLA ET AL.
11
3 an analysis of crucial pairs as they emerge from the specific database under consideration - see Table 1.1 for examples based on the English Ward Thesaurus - relating to the following classes of terms which are particularly worth considering in any overview of the construction of gender (see on-line tables for some chapters in this book, on www.lingue.unibo.it/wordthesaurus): (a) kinship terms, analysis of which has also proved important in anthropology, as a way of improving our understanding of family organization and of various family roles; (b) terms for person reference/address forms, greetings; (c) occupational terms, professional tides, since this is an area where - as a result of social developments and linguistic reforms - changes have been frequent in recent years and where clearly we may expect progress to continue. 4 a summary and (partial) conclusions in relation to the specific language analysed. The various factors that we thought it important to evaluate were not only the amount of information given or not given (e.g., in the English Thesaurus, host is registered, but hostess is not), but also the register selected (formal/ informal), the related stereotypes, and the order in which the terms are listed (i.e. how obvious the term is, its 'visibility',9 which is obviously higher at the top of a list). In other words, the linguistic coding (e.g. the abovementioned lexico-grammatical asymmetries) and the actual ordering and presentation in the Thesaurus were examined (though clearly with different degrees of attention). Other additional data are of course relevant to gender issues, and some of them are dealt with in this volume as a way of widening the analysis in a
Table 1.1 'Meanings' and 'synonyms' in the English Word Thesaurus Thesaurus
Word or Phrase
woman
Meanings
Synonyms
female
female, lady, girl, dame, matron, gentlewoman
womankind
maid, virgin, fern, femininity
human being
human being, humankind, mankind, humanity, mortals, populace, race, folk, people
male partner
male, fellow, chap, guy, gentleman, beau, swain, blade, yeoman
man
12
GENDER, LANGUAGE AND NEW LITERACY
multidimensional perspective: proverbs (for Italian, Manera & Bazzanella; for Turkish, Castagneto & D'Amora), metaphors and metaphorical expressions (e.g. for Hungarian, Nagy & Paid; for Portuguese, Marques), obscene expressions and verbal insults,10 idioms, diminudves and terms of endearment11 (for Spanish, see Guil), data taken from newspapers (see Holmes 2001 for an analysis based on New Zealand newspapers) and radio/TV programmes, and advertisements (e.g. for Italian, see Manera Be Bazzanella). Especially in a field like gender, one cannot avoid taking account of a number of different parameters, above all multicultural ones. As Hellinger & BuBmann (2001-3: 3) note: In communication, parameters like ethnicity, culture, social status, setting, and discourse functions mav in fact be as important as extra-linguistic gender, and none of these parameters is represented in a language in any direct or unambiguous way (see Bing & Bergvall 1996: 5). Only a multidimensional theory of communication will be able to spell out the ways in which parameters interact with linguistic expressions. By interpreting linguistic manifestations of gender as the discursive result of 'doing gender' in specific socio-cultural contexts, the analysis of gender across languages can contribute to such a theory.
If one takes a multicultural approach, there is a better chance of avoiding the risk of 'saturating L&G [language and gender] interpretations with middleclass-white-Anglo-American attributes' (Bergvall 1999). 4 Result s In the preceding research (Bazzanella et ai 2000), which was based on four languages, die following results turned out to be common: • the social 'superiority' of men as against the subordinate status of women, reflecting the historical domination of men; • a clear quantitative difference, as seen in Italian with regard to padrona padrone 'mistress' - 'master', where the male entry includes twice as many terms as the female one. This quantitative difference is accompanied by a qualitative imbalance which is all to the advantage of the male gender as regards the variety and complexity of the 'synonyms' provided for the male terms; • different structuring of the entries, with regard both to 'visibility' and frequency/order, which often suggests subordination of the female terms to the male ones (that is, when the former are not actually missing); • frequent dependence of the female term on the male one; • women are relegated to (minor) domestic affairs, sometimes even without anv recognition of their function in procreation (e.g. in Italian procreatare 'procreator' is given as a synonym of padre 'fadier', but, strangely enough, there is no female correspondent of 'procreator' in the entry for madre 'modier');
INTRODUCTION BAZZANELLA ET AL.
13
• pervasive stereotypes suggesting the strength of men and the delicacy of women. The various studies we present here (though the Thesauri considered may differ in a number of details, see the specific contributions) show a number of common features. In general, a hegemonic, patriarchal social system has been found to be dominant in all the languages considered. The 'woman' is not presented as an 'individual' in her own right, but only in relation to others, her identity deriving from being a wife, a daughter or mother.12 Lemmas with female reference are not listed - in van Alphen & Corda's words - 'on the basis of autonomous, intrinsic semantic features (like age or sex)'. As a result, a general male-dominant attitude, which reflects a common view of male superiority13 and female subordination, tends to emerge quite strongly. The order in which synonyms appear is also significant: in German, for example: the first term listed for Frau is the word for 'spouse' (Ehefrau), while for Mann this comes second (Ehemanri). This suggests the enduring influence of social stereotypes, namely that the primary semantic sphere associated with 'woman' is the family and above all that of being someone's wife. (Thune & Leonardi)
In sharp contrast, man turns out to be the point of reference, and the general noun denoting a person is often derived from the noun man or associated with maleness; interestingly, also with regard to the Chinese tool here analysed, the functioning of the database for Microsoft Word 2003 with Microsoft Pinyin IME 2003 word-processing is based on the identification of humankind with male gender (Ceccagno). Crucially, in Hebrew, the man is considered 'as the prototype from which woman is created'. . . . This [i.e. the Hebrew word adorn 'man'] is the first name of the first man, which in and of itself explains the source of the word's ambiguity: On the one hand, it refers to the human race as a whole (ben adam= 'descendant of the first man'), while on the other hand, it refers to the males of this group. (Livnat)
An androcentric bias is present in all the lexical representations of womanvs man. This takes different forms in different languages: for instance, in Dutch, as van Alphen and Corda report, 'man "man", singular and jongens "guys", plural, are commonly used in informal language to address both genders'. Preceding research on the presence of women in the English and German vocabulary14 (see Nielsen 1977; Breiner 1996) showed a considerable quantitative and qualitative gender asymmetry in the lexical entries, since on the whole there are fewer gender-specific terms referring expressly to women, while the percentage of derogative terms which are normally used in
14
GENDER, LANGUAGE AND NEW LITERACY
reference to women was found to be higher than those used for men (Linke 2002: 123). This was confirmed in the present analysis, where a clear imbalance favouring the male gender is found across the board: • in the amount of information provided (in Portuguese, for example, 'there is a prevalence of synonyms for the masculine in relation to the feminine, respectively 353 as against 230', as noted by Marques), and in the kind of information. In Turkish even neologisms built up to be symmetric, as Bay 'Mr' and Bayan 'Miss'/'Mrs', are not dealt with as such by the Thesaurus: to the word Bay the Thesaurus devotes twice as many meanings and synonyms as to the equivalent female term. (Castagneto & D'Amora) • in the selection of meanings and synonyms (in some cases the female lemmas are simply missing) and in the order in which they are given:15 The asymmetry derives also from the different sequentiality and ranking within the main entry, and from the distribution of meanings and synonyms; a term which should be registered as a synonym is given as a meaning, gaining extra visibility and introducing an imbalance between the female/male pair. (Manera & Bazzanella) • in the greater complexity of the synonyms and details provided for male terms; • in the stereotypes associated with female terms: feminine forms tend to prevail in semantic fields related to the physical characteristics of women or to sexual conduct (Manera & Bazzanella; Marques); furthermore, in the Hebrew Thesaurus, while terms of abuse used specifically for women are given, equivalent terms for men (which are no less plentiful in spoken Hebrew) do not appear (Livnat); • in the tendency for the feminine forms to carry a negative connotation as compared to the corresponding masculine forms; this is found in Greek (Katsoyannou &r Goutsos), in Hebrew (Livnat), in Hungarian (Nagy & Patti), in Portuguese (Marques), in Spanish (Guil), in Turkish (Castagneto & D'Amora). Though the more recent versions of the Thesauri examined in the present work (as compared to the earlier versions) definitely show a degree of improvement in the sense that male and female entries are now better balanced as regards both the quantity of synonyms/meanings given and the ordering of the entries (at least this is the case with the English, Italian, and Spanish Thesauri-for details, see respectively Bettoni; Manera & Bazzanella; Guil), asymmetry nevertheless remains pervasive. The lack of balance in the Thesauri examined here is also evident in the dependence of the female term on the male one, particularly in the area of
INTRODUCTION BAZZANELLA ET AL.
15
occupational terms (for example, in regard to Polish, Miecznikowski notes a tendency for the female terms given to be the result of simple mechanical application of derivational rules without any consideration given to the fact that the resulting female terms do not exist or are not used). Nagy & Patti establish an interesting subdivision of names of professions (see below), which can be usefully applied to the variegated landscape of professional nouns in the various languages: 1 Type tandr' teacher'. This group consists of names of professions/occupations traditionally associated with both men and women. 2 Type miniszterelnok 'prime minister'. In this group we find highly prestigious and relatively new occupational denominations. 3 Type pilota 'pilot'. This group contains words denoting professions/ occupations prevalently associated with men. 4 Type ovono 'kindergarten teacher'. In this group we find professions/ occupations traditionally associated with women. Gerritsen's (2002:85) statement- 'the professional term indicating a woman refers to a job with a lower social status than the term indicating a man' perfectly encapsulates the treatment of the professional lemmas in our corpus. Summarizing (albeit at the risk of simplifying somewhat), the cultural/ social background which (on the basis of the Thesauri16 analysed in our study) turns out to be associated with the lemmas related to 'woman' and 'man' tends to suggest a degree of male dominance which exceeds what we find in real life: as van Alphen & Corda claim: 'this twenty-first century application contains old-fashioned and very stereotypical assumptions about social roles for women and men in Dutch society'. From a technical point of view, the Microsoft Word Thesaurus, 'compared to other computer-mediated dictionaries, which form highly complex hypertextual structures, ... is a simple lexicographical tool' (Miecznikowski). In general, an electronic thesaurus presents inherent dangers (see Bettoni). What is more, the Ward Thesaurus seems to be affected by a pervasive carelessness which requires the utmost attention on the part of the user: formal and substantive imprecision, inconsistencies, omissions, and straightforward mistakes are all common (see the specific contributions). Bettoni considers the American, British and Australian versions of the Word 11 Thesaurus to be: above all, an unreliable tool. Is it really appropriate to use the term 'sexist' to describe a word-processing instrument which can assign to secretary the sole meaning of desk, forget to mention a female referent among the synonyms for millionaire. treat mother on\y as a verb with the sole meaning of look after, and father only as a noun with the sole meaning of priest? Rather than being 'sexist', it is simply a badly conceived instrument.
With regard to synonyms, which are of course the central device in the
16
GENDER, LANGUAGE AND NEW LITERACY
thesauri, many criticisms have been made: van Alphen & Corda, for instance, claim that: the use of synonyms to explain senses (which characterizes all versions of the Word Thesaurus] has many shortcomings, and these shortcomings are reinforced by the use of automatic procedures to generate synonym relations among lemmas. Short definitions, like those in bilingual dictionaries, would be much more helpful.... In the second place, not all the lemmas in the Thesaurus have synonyms, while, on the other hand, the Thesaurus is not a complete list of all words with synonyms. Why have lemmas without synonyms been listed? The answer is probably to be found in the automatic generation of senses, antonyms and synonyms.
According to Nag}- & Paid, fake synonymous terms are given which 'often do not fit easily into any semantic affinity such as hypernymy, hyponymy or synonymy. Furthermore, it is not uncommon to find the lemma itself proposed as the first synonym of the list, which is at best a useless repetition.' In the Polish version of the Microsoft Word Thesaurus, a 'large proportion of lexically female personal nouns turn out to owe their inclusion onlv to mechanical (and therefore error-prone) application of derivational rules, and have only lexically male synonyms, irrespective of the latter's referential properties' (Miecznikowski). The shortcomings of the thesauri - which, we would like to emphasize, remain a useful instrument in principle - are by no means limited to those we have just listed. Indeed, each of the Thesauri'we analysed (without exception) could not unfairly be characterized as an unsuccessful attempt at simplification of the system. The potential of these instruments for causing confusion and misunderstanding is only too apparent: • Sociolinguistic parameters in general, i.e. diachronic, diatopic, diastratic, diaphasic, which are highly significant for pragmatic appropriateness, are not provided: 'the fact that neither the "meanings" nor the "synonyms" are differentiated in the slightest poses serious problems when it comes to choosing the appropriate word' (Thune & Leonardi). Obsolete words are proposed, without being signalled as such; for example in Dutch 'The obsolete juffmuw and juffer "miss" are given as synonyms of dame "lady"' (van Alphen & Corda) - see also German (Thune & Leonardi) and Hungarian (Nagy & Patti). The kind of register is usually not signalled: in the case of the German Thesaurus, among the various 'synonyms' for Mddchen 'girl' we find Fratz, a term which implies affectionate closeness to the 'girl', and is therefore limited to the colloquial-familiar domain (Thune & Leonardi); registers are mixed up so that in the Turkish Thesaurus standard words appear side by side with archaisms (e.g. muattim 'teacher'), provincial terms (e.g. dados 'brother, young man', used in Eastern Anatolia), colloquial terms (e.g. dudu 'old Armenian woman') and slang tenns (e.g. bizdtk 'litde child') (Castagneto £ D'Amora). Regionalisms are not signalled: thus, in Italian, picciotto 'boy' (typically Sicilian) is given as a neutral choice (Manera & Bazzanella), and words belonging to different national
INTRODUCTION BAZZANELLA ET AL
17
varieties are given without warning; thus we find words from Brazilian Portuguese (sid, iaid, sinhd- familiar terms of address for women from the era of slavery) in a European Portuguese dictionary (Marques). The same failure to offer any sort of sociolinguistic guidance is found in all the thesauri, but of course the seriousness of such omissions in any given case depends to a great extent on the specific sociolinguistic parameter and on the specific language: for example in Greek, a language 'which draws on a plethora of sources from its particularly rich diachronic variation, relating to its long-standing history of diglossia' (Katsoyannou & Goutsos), the failure to take account of the diachronic variable is particularly serious. • Nor is the reader offered any contextual cues such as collocations and periphrastic idioms; this is exactly what we find in Czech with regard to pan - pant'lord', 'Mr' - 'lady', 'Ms', pan tvorstva 'lord of creation', pani domu 'lady of the house', lepst polovicka 'better half, 'good lady', etc. - which, as Cmejrkova notes, are all terms characteristic of traditional gender discourse. Analogously, metaphorical items are not explicitly indicated as such, thus opening the way for a series of errors on the part of the user (Manera & Bazzanella; Nagy & Patti). In Polish, where 'the extent to which . . . morphological and semantic resources are used to make female gender explicit... does not only depend on the referent's sex, but also on the type of noun, the syntactic context, textual constraints . . . and the communicative importance of gender as a category' (Miecznikowski), lack of contextual cues seems to be particularly significant. • Since the actual source of the data provided is not disclosed, as a matter of fact it is not possible to know which corpora a specific thesaurus was based on, its size and chronological composition, all of which information could be useful (at least to a scholar) in deciding in advance whether to use this tool for a given purpose or not. The general problem which devices such as the Word Thesaurus face, that of how to deal adequately with 'the complexity of both sociocultural settings and the ramifications of multiple roles that individuals play in their interaction and communication with others' (Chan 2002: 58), could be partially solved by exploiting their powerful resources so as to provide a variety of links to monolingual and bilingual dictionaries, encyclopedias (for instance, the Encarta World Dictionary, which is included with the latest versions of the US English Thesaurus (see Bettoni) and of the Spanish Thesaurus (see Guil)), as well as any other documents which could be relevant to a given item, anything, in short, that will serve to better elucidate its lexical or encyclopaedic meaning.17 To conclude with Chinese, it is noteworthy that, though this is in absolute terms a 'genderless' language, [it] is in actual fact governed by sociocultural factors, i.e. by socio-cultural assumptions and expectations about the relationship between women and men, and therefore can be gender-biased. ... covert male gender abounds in terms for person
18
GENDER, LANGUAGE AND NEW LITERACY reference in Chinese, as the neutral meaning is often being associated with male gender. (Ceccagno)
Once again, the asymmetric categorization and representation of the roles of women and men, particularly in the domain of person reference and occupational names (which were the main focus of the analysis), are both reflected in and reinforced by Microsoft Pinyin IME 2003, a software program that allows a standard keyboard to be used for the Chinese non-alphabetic writing system (see above §1). What is more, in comparison to the thesauri of other languages, in the case of Chinese, die effect of gender bias in word processing programs is - as Ceccagno succincdy puts it - 'even more pervasive, as their use is not confined to certain specific tasks but radier is absolutely indispensable as a way of getting Chinese characters on the screen'. 5 Conclusio n Though 'no movement or campaign has yet effected the total elimination of sexist language among its subject people', as Ettner (2002: 51) states, what changes could be made in Uiese thesauri so as to improve them? If it is to provide a more accurate, up-to-date18 and helpful instrument for its users with regard to gender issues, the Microsoft Word Thesauruswll need to fill the various gaps mentioned above (i.e. the lack of sociolinguistic guidance and contextual cues); it will also need to correct various mistakes, omissions, and imprecisions, and make lexicographical criteria (e.g. for morphological categories - see Bettoni; Nagy & Patti; Thune & Leonardi; Guil) more accurate, clear, and consistent, as well as removing the remaining asymmetry between the female and male entries. It will further need to deal with all die related problems which have been unearthed in this volume, and do so on the basis of the various existing recommendations and established guidelines, more specifically, following the possibilities given in the various languages (i.e. engendering and degendering, see §2.3). As Bettoni states widi regard to the English language, 'Microsoft is not responsible for sexism, but it is certainly responsible for failing to show awareness and providing available alternatives.' Unfortunately, linguistic reform limited to what Romaine (2001: 154) calls 'gendered bits of the language such as titles, forms of address, and androcentric generics', is not enough. 'More has to change', as Elisabeth Burr (2003: 133) claims: 'the predominance of the masculine over the feminine gender is part of our grammatical culture'. We would like to conclude our introductory remarks by stating that, if there is to be any substantial change in the androcentric position that is firmly entrenched in the electronic thesauri we have analysed, it is not enough for discourse and interaction between women and men to change;19 inevitably, there must also be a change in familial and social roles, in power
INTRODUCTION BAZZANELLA ET AL.
19
relationships, and, crucially, a change in the way of constructing and experiencing reality.20 This must necessarily involve the media (see Koniuszaniec & Blaszkowska 2003: 277), because they surely have an important role to play in disseminating 'non-sexist' language use. Notes 1 Henceforth all references are to studies contained in this volume, unless otherwise specified. 2 According to Curzan (2000: 572): The way in which language users make distinctions between male and female and between masculine and feminine in their culture will be reflected in the distinctions they made between masculine and feminine in their language, as long as the gender system is a semantic one. Like gender in society, gender in the English language represents a set of constructed categories, categories whose boundaries will change over time, reflecting the evolution of ideas about sex and gender. 3 As Braun (2001: 287) rightly states: This gender belief system would produce gender biases or gender expectations that remain hidden beneath the grammatical neutrality of the linguistic structure.' 4 Paradoxically, in the Italian Word Thesaurus a source of asymmetry derives from the gender-indefinite terms, exactly the ones which should be exploited by following a politically correct usage, but which, on the contrary, turn out to be only, or mainly, masculine, e.g. singl e provides the unique meaning of scapolo 'bachelor', celibe 'not married' (m.), and ammogUato"married' (m.) as antonyms. (Manera & Bazzanella) 5 'Meanings' is the term used in the Thesaurus, we keep it for the sake of simplicity. Actually, they rather work as hypernyms: by selecting them, one is provided with possible 'synonyms'. 6 Similar to 'meanings' (see note 5) they are not 'synonyms' strictu sensu; they are possible substitutions of a given lexeme in relation to a selected 'meaning'. 7 English, Italian, and Spanish are the languages analysed in both the research studies (see the specific contributions for comparisons between the two versions). 8 Despite the complexity of cultural/linguistic relations, in particular in a language such as English, which is spoken in many countries (American, Australian, British, New Zealand varieties are distinguished in Hellinger & BuBmann (2001-3)), a common discourse can nevertheless be drawn (see Bettoni). 9 On the contrary, 'invisibility' means not being mentioned at all as women: 'One of the sometimes more subtle forms of discrimination against women is that they are simply not mentioned at all!' (Romaine 2001: 154). 10 As Braun (2001: 300) reports in general: In many languages, insults aimed at women focus upon their sexuality as the target of the verbal attack. ... To the degree that recipients feel degraded or provoked, insults reconfirm the norms a society has built up for women and men. With specific regard to Turkish:
20
GENDER, LANGUAGE AND NEW LITERACY
Abuse and insults are not exclusively concerned with gender (see expressions such as salak 'idiot'), but from this conventionalized area of language usage much can be learned about the core conceptions of gender and the asymmetries in gender role operative in the Turkish language community. (Braun 2001: 303) 11 According to Kuntjara (2001: 214): Concerning terms of endearment, Wolfson (1989) maintains that in English women are often addressed with endearment terms like dear, hon, sweetheart or doll by men, irrespective of age and social status. At the same time, the use of terms of endearment is often non-reciprocal, which can be interpreted as a sign that females are generally held in less respect than males. 12 See the case of surnames in Greek: 'the fact that the official identity in the case of women is morphologically related to her father's or husband's identity suggests that it follows the identity of her "owner"' (Katsoyannou & Goutsos). 13 See, with specific regard to the dependence from man in the wife/husband relation in Polish, Koniuszaniec & Blaszkowska (2003: 264), in Serbian, Hentschel (2003:299). 14 For German, the latest edition of the Duden showed a serious attempt to balance the gender asymmetries (see Duden-Bedeutungsworterbuch (2002) and Duden, Deutsches Universalwarterbuch (2003), Kunkel-Razum (2004); for further details, see Thune & Leonardi). 15 In Portuguese, the words selected as 'synonyms' are listed in alphabetical order (see Marques). 16 Note that 'a similar failure to take account of current issues regarding women's role in society is to be found in other electronic devices, such as the Google searchengine' (Thune & Leonardi). 17 AsTeubert (2005) savs: Meaning is in the discourse. Once we ask what a text segment means, we will find the answer only in the discourse, in past text segments which help to interpret this segment, or in new contributions which respond to our question. Meaning does not concern the world outside the discourse. There is no direct link between the discourse and the 'real world'. It is up to each individual to connect the text segment to their first-person experiences, i.e. to some discourse-external ideation or to the 'real world'. 18 As Bettoni states with regard to the English Thesaurus:. the treatment of women in the three English versions of the Microsoft Word Thesaurus is in striking contrast to the international role of the English language as a model democratic institution, its glorious tradition of lexicography, and its important achievements in feminist studies and gender reforms. 19 Let us think about the system of address and self-reference; for a survey of the different (asymmetric) forms used to address women and men in Spanish public broadcast interviews, see Guil. 20 A small example: The overwhelming majority of the instances of chairman [in four English corpora] were identifiable as male, a sad reflection of the social reality that it was men who held this position most often' (Holmes 2001:125).
INTRODUCTION BAZZANELLA ET AL.
21
References Bazzanella, Carla; Gallardo, Catherine; Guil, Pura; Manera, Manuela & Tejada. Paloma (2000), 'Categorizzazioni del femminile e del maschile nelle nuove tecnologie: prime ricerche nel Thesaurus italiano, spagnolo, francese, inglese, danese di Word'. Cuadernos defilologia italiana, 7,193-245. Bergvall, Victoria L. (1999), 'An agenda for language and gender research for the start of the new millennium'. Linguistik online, 2.1,
Bing, Janet M. & Bergvall, Victoria L. (1996), 'The question of questions: beyond binary thinking', in Victoria L. Bergvall, Janet Bing & Alice F. Freed. Rethinking Language and Gender Research: Theory and Practice. Singapore: Longman. pp. 1-30. Braun, Friederike (2001), 'The communication of gender in Turkish', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins. pp. 283-310. Braun, Friederike; Gottburgsen, Anja; Sczesny, Sabine & Stahlberg, Dagmar (1998), 'Konnen Geophysiker Frauen sein? Generische Personenbezeichnungen im Deutschen'. Zeitschriftfur germanistische Linguistik, 26, 265-83. Braun, Friederike; Sczesny, Sabine & Stahlberg, Dagmar (2005), 'Cognitive effects of masculine generics in German: an overview of empirical finding'. Communications. 30,1-22. Breiner, Ingeborg (1996), Die Frau im deutschen Lexikm: Eine paradigmatische Untersuchung. Vienna: Edition Praesens. Burr, Elisabeth (2003), 'Gender and language politics in France', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 119-39. Chan, Marjorie K. M. (2002), 'Chinese: gender-related use of sentence-final particles in Cantonese', in Marlis Hellinger & Hadumod BuBmann (eds). Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 57-72. Collins Cobuild (2003), = Coltins COBUILD Advanced Learner's English Dictionary. Ed. John Sinclair. 4th edn. London: HarperCollins. Cope, Bill & Kalantzis, Mary (eds) (2000), MuUiliterades: Literacy Learning and the Design of Social Futures. London: Routledge. Curzan, Anne (2000), 'Gender categories in early English grammars: their message to the modern grammarian', in Barbara Unterbeck & Matti Rissanen (eds), Gender in Grammar and Cognition. Berlin: de Gruyter, pp. 561-76. Dirven, Rene & Verspoor, Marjolijn (eds) (1998, 2004), Cognitive Exploration of Language and Linguistics. Amsterdam and Philadelphia: Benjamins. Duden-Bedeutungsworterbuch (2002), (= Der Duden in zwolf Bdnden. vol. 10). 3rd edn. Mannheim: Dudenverlag. Duden Deutsches Universalworterbuch (2003), 5th edn. Mannheim: Dudenverlag. Ettner, Charles (2002), 'In Chinese, men and women are equal - or - women and men are equal?', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 29-55. Gee, James Paul (1996), SocioUnguistics and Literacies: Ideology in Discourses. 2nd edn. London: Taylor & Francis.
22
GENDER, LANGUAGE AND NEW LITERACY
Gerritsen, Marinell (2002), 'Towards a more gender-fair usage in Netherlands Dutch', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 81-108. Givon, Talmy (1989), Mind, Code and Context: Essays in Pragmatics. Hillsdale, NJ: Erlbaum. Goffman, Erwin (1977), The arrangement between the sexes'. Theory and Society, 4, 301-31. Graddol, David & Swann, Joan (1989), Gender Voices. Oxford: Basil Blackwell. Gunthner, Susanne (1997), 'Zur kommunikativen Konstruktion von Geschlechterdifferenzen im Gesprach', in Friederike Braun & Ursula Pasero (eds), Kommunikation von Geschlecht [Communication of Gender]. Pfaffenweiler. Centaurus, pp. 122-46. Halpern, Diane F. (2000), Sex Differences in Cognitive Abilities. 3rd edn. Mahwah, NJ: Erlbaum. Hardin, Curds & Banaji, Mahzarin R. (1993), 'The influence of language on thought'. Social Cognition, 11, 277-308. Hardin, Curtis & Banaji, Mahzarin R. (1996), 'Automatic stereotyping'. Psychological Science, 7, 136-41. Hellinger, Marlis (1990), Kontrastive feministische Linguistik. Mechanismen sprachlicher Diskriminierung im Engtischen und Deutschen. Munich: Hueber. Hellinger, Marlis & BuBmann, Hadumod (2001-3), 'The linguistic representation of women and men', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 1-25. Hellinger, Marlis & BuBmann, Hadumod (eds) (2001-3), Gender across Languages: The Linguistic Representation of Women and Men. Amsterdam and Philadelphia: Benjamins. Hentschel, Elke (2003), 'Serbian: The expression of gender in Serbian', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 287-309. Holmes, Janet (2001), 'A corpus-based view of gender in New Zealand English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 115-36. Klann-Delius, Gisela (2005), Spracheund Geschlecht: EineEinfuhrung. Stuttgart: Metzler. Koniuszaniec, Gabriela & Blaszkowska, Hanka (2003), 'Language and gender in Polish', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women- and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 259-85. Kunkel-Razum, Kathrin (2004), 'Die Frauen und der Duden - der Duden und die Frauen', in Karin M. Eichhoff-Cyrus (ed.), Adam, Eva und die Sprache: Beitrdge zur Geschlechterforschung. Mannheim: Dudenverlag [Thema Deutsch; 5], pp. 308-15. Kuntjara, Esther (2001), 'Indonesian: gender in Javanese Indonesian', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 199-227. Lankshear, Colin (ed.) (1997), Changing Literacies. Buckingham: Open University Press. Lee, David (2001), Cognitive Linguistics: An Introduction. Oxford: Oxford University Press.
INTRODUCTION BAZZANELIA ET AL.
23
Leu, Donald J., Jr; Kinzer, Charles K.; Coiro, Julie L. & Cammack, Dana W. (2004), 'Toward a theory of new literacies emerging from the Internet and other information and communication technologies', in Robert B. Ruddell & Norman Unrau, Theoretical Models and Processes of Reading. Newark, DE: International Reading Association, pp. 1570-613. Linke, Angelika (2002), 'Das Wort in der feministichen Sprachreflexion: Eine Ubersicht', in D. Alan Cruse, Franz Hundsnurscher, Michael Job & Peter Rolf Lutzeier (eds), Lexikologie/Lexicology: Ein internationales Handbuch zur Natur und Struktur von Wortern und Wortschdtzen. HSK21.1. Berlin: de Gruyter, pp. 121-8. Marcato, Gianna & Thune, Eva-Maria (2002), 'Gender and female visibility in Italian', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 187-217. Martyna, Wendy (1978), 'What does 'He' mean - use of the generic masculine?' Journal of Communication, 28,131-9. Nielsen, Alleen Pace (1977), 'Sexism as shown through the English vocabulary', in Alleen Pace Nielsen et aL (eds), Sexism and Language. Urbana, IL: National Council of Teachers of English, pp. 27-41. Pauwels, Anne (2001), 'Spreading the feminist word: the case of the new courtesy title Ms in Australian English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 137-151. Ritchie Key, Mary (1996), Male/Female Language: With a Comprehensive Bibliography. Boston: Scarecrow Press. Romaine, Suzanne (1999), Communicating Gender. Mahwah, NJ: Erlbaum. Romaine, Suzanne (2001), 'A corpus-based view of gender in British and American English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 153-75. Rosch, Eleanor (1973), 'Natural categories'. Cognitive Psychology, 4, 328-50. Sapir, Edward ([ 1921 ]; 1933), Language: An Introduction to the Study of Speech. New York: Harcourt Brace and World. Silveira, Jeanette (1980), 'Generic masculine words and thinking'. Women's Studies International Quarterly, 3, 165-78. Taylor, John R. (1989), Linguistics Categorization: Prototypes in Linguistic Theory. Oxford: Clarendon Press.
Teubert, Wolfgang (2005), 'My version of corpus linguistics'. International Journal of Corpus Linguistics, 10,1, 1-13. Unsworth, Len (2002), 'Changing dimensions of school literacies'. The Australian Journal of Language and Literacy, 25, 1, 62-77. Unterbeck, Barbara & Rissanen, Matti (eds.) (2000), Gender in Grammar and Cognition: I Approaches to Gender. II Manifestations of Gender. Berlin: de Gruyter. Violi, Patrizia (1997), Significato ed esperienza. Milano: Bompiani. West, Candace & Zimmermann, Don (1991), 'Doing gender', in Judith Lorber & Susan A. Farrell (eds), The Social Construction of Gender. London: Sage, pp. 13-37. Wierzbicka, Anna (1997), Understanding Cultures through their Key Words. New York: Oxford University Press. Whorf, Benjamin Lee (1956), Language, Thought, and Reality. Cambridge, MA: The MIT Press.
2
Czech: friendly to women? Svetla Cmejrkovd
1 Introductio n Languages differ not only in what they can express but also and most particularly in what they must express. Czech can (and in many cases must) communicate not only the gender of a person referred to (in the 3rd person) but also the gender of the speaker (1st person) and that of the addressee (2nd person). A man is addressed (mily ctendn - 'dear reader' m.) or referred to (mladj muz se smdl- 'a young man laughed' m.) with masculine inflection, whereas a woman is addressed (mild ctendfko- 'dear reader' f.) or referred to (mladd zena se smdla- 'a young woman laughed' f.) with feminine inflection; while a man says pfdl bych si 'I would like to' (m.), a woman says pfdla bych si 'I would like to' (f). In many categories, men and women have their own sets of endings. The Czech essayist Pavel Eisner called Czech a thoroughly sexist language long before this topic was ever subjected to linguistic investigation (he mentioned this feature of Czech as long ago as 1946: 377-82). He established a scale of languages, based on the degree of linguistic sexism: With regard to the structure of morphological endings and consequently word forms controlled by gender, we can arrive at absolutely sexless languages like English, then languages with a small degree of sexism like German, languages with a higher degree - these are the Romance languages - and in the end thoroughly sexist languages, and these are represented in Europe by Slavic languages, including Czech. (Eisner 1946: 378, translation mine)
Grammatical gender (masculine, feminine and neuter) is an inherent morphological property of the Czech noun and controls grammatical agreement between the noun and the verb, as well as between the noun and its gendervariable satellite elements, such as adjectives, pronouns and numerals in both attributive and predicative positions. The category of grammatical gender (together with that of animacy within masculines) controls grammatical agreement in both singular (mily ctendfm. vs mild ctendfka f. - 'dear reader') and plural (mill ctendn m.. vs mile ctendfky f. - 'dear readers'). The category of gender (and animacy within masculines) also controls the inflectional
CZECH CMEJWOVA
25
orthography of co-referential nouns and verbs; masculine agreement is featured with the ending -t, whereas feminine agreement is featured with the ending -y. Cestiauton psali cesky m. - ceske autarky psafy cesky L ('Czech authors wrote in Czech'). Although femaleness has salient formal manifestations in Czech, thus producing the impression of the desired gender symmetry, many of the formal and functional manifestations of gender in the area of human reference are, in fact, asymmetric; they can be treated as relics of the so-called 'patriarchal language paradigm' (the term was coined by Eisner in 1946: 366 n.) perhaps universal in all European, if not world languages. In Czech, this applies particularly (1) to the wide usage of the so-called generic masculines and (2) to the grammatical norms accepted in orthography: 1 Generic masculines: Though active processes of word-formation and derivation of new words by means of feminine endings in particular give rise to numerous feminine counterparts to masculine terms, the frequency of masculine terms in text and discourse is much higher than that of their feminine derivatives, and this is due to the so-called generic function of masculine forms (see Introduction, §2.1): it is taken for granted that the meaning of the feminine term autorka ('female author') is covered by the masculine form autor ('male author'), as well as ctendrky ('female readers') by ctendn ('male readers'), and kazdd ('every woman') by kazdy ('every man') etc. This fact, well known from other European languages, has many syntactic consequences in Czech. 2 Orthography. The most conspicuous orthographical consequence of the 'patriarchal language paradigm' is the choice of the ending -* in case of mixed gender groups: it is the animate male presence that decides the choice - if at least one male person is present in the group, only the masculine agreement -i is permitted in Czech orthography: (1) Tncet tn zeny a jeden muz soutezili.i, thirty-three womenf and one nianm competed™ Thirty-three women and one man took part in competition. The choice of feminine agreement in a corresponding verb ending is qualified as a serious grammatical error, and sanctioned as one of the most deplorable. Eisner (1946: 382) wittily demonstrated the norm based on the preference for men and 'neglecting women' in Czech, using a striking example: (2) Jeji Velicenstvo
Her Majestynt
krdlovna anglickd a cisarbvna indickd;te; the Queenf of England and EmpresSf. of India
a ndic Pepa 2andourek vypadli z vozu. and the driverm Pepa 2andourek fellm off the coach.
26
GENDER, LANGUAGE AND NEW LITERACY
Pavel Eisner came up with his 'feminist critique' of language in the mid19408. Based on his observations of gender peculiarities of Czech, he declared his somewhat emotional involvement with this language, rather than conduct a critical analysis of its patriarchal structure. Being a bilingual German speaker of Czech (whose linguistic intuitions were based on his own translation practice), Eisner formulated his indulgence in Czech featuring femaleness with sets of special endings (-a, -a, -la, etc.). Thanks to what he perceives as its 'erotic spell', Czech appears to Eisner as a woman's language (that is also why he prefers to use a common feminine noun cestina 'Czech' and to avoid the term ceskyjazyk 'Czech language' (m.), e.g. the tide Kniha o cestine 'Book on Czech' or a chapter from this book Cestina a zena 'Czech and Woman') (Eisner 1946: 366-88). Some of Eisner's observations may sound old-fashioned and disputable when confronted with today's feminist critique of language and with the strict and detached objectives of so-called gender-fair usage in language. His other observations on Czech are, however, still convincing. He noticed, e.g., that Czech is friendlier and more welcoming to women than other languages, in which the general noun denoting a person is derived from the noun 'man' or associated with maleness. Czech differentiates between the general personal noun clovek and the personal noun with male-specific reference muz ('man') (ibid.: 366). Eisner also discussed gender semantics of the general personal noun clovek in relation to another Czech general personal noun, osoba ('person'): whereas the noun clovekbelongs to grammatical masculines, osoba belongs to grammatical feminines. The fact that Czech relatively widely applies grammatical feminines to designate both male and female persons should also be taken into account when analysing gender 'fair' and gender 'unfair' or 'biased' speech practices in Czech (Cmejrkova 1995,1997, 2003; Danes' 2001; Valdrova 1999, 2001). In the following sections I will trace paradigmatic symmetries/asymmetries in the representation of men and women in Czech and the Microsoft Word Thesaurus, and their syntagmatic consequences in text and discourse. I will: (1) examine functions of grammatical gender in language and discuss appropriate formal and semantic readings (§1); (2) give a survey of lexical gender in Czech and of its treatment in the Thesaurus (§2); (3) outline the derivation of Czech feminines and check this word formation process in die Thesaurus (§3); and (4) scrutinize semantic and pragmatic functions of common gender nouns ('epicenes') and their handling in the Thesaurus (§4). In conclusion, I will discuss a current tendency in the Czech gender system, i.e. the shift towards semantic reading of the category of gender in inflecting languages. Finally, I will attempt to evaluate the Czech version of the Microsoft Ward Thesaurus in regard to the category of gender in Czech.
CZECH CMEJRKOVA
2?
2 Th e meaning of grammatical gender in language: formal vs semantic reading As Muhammad Hasan Ibrahim claimed in his book Grammatical Gender: Its Origin and Development, 'Gender in its origin was an accident of linguistic history' (1973: 30-51). In Proto-Indo-European languages, gender (as opposed to sex) started out as a classification system for nouns with similar phonological properties, particularly with similar endings (Fodor 1959: 21214). This classification system developed into a grammatical system of syntactic concord, causing an assonance-like agreement in elements such as determiners, adjectives, and participles associated with the noun. Over the years these divisions gained a certain amount of semantic motivation, as well as linguistic interpretation. Philosophers, writers and linguists from Plato to Jespersen have endeavoured to show a semantic base for the gender assignment of nouns, whether their referents were animate or inanimate. In the contemporary feminist critique of language, their effort is continued, e.g., by a French feminist critic of language, Luce Irigaray, whose claims that metaphorically whatever has been valorized has been masculine in gender, whatever devalued feminine, have had a strong influence on many gender studies. Irigaray introduces examples such as French k soldi (m.) ('the sun') vs la lune (f.) ('the moon') (Irigaray 1999:120). However, these nouns are differently gendered in other inflected languages: in Czech, for example, slunce ('the sun') is neuter, whereas mestc ('the moon') is masculine and only its poetic equivalent luna is feminine. There are many other linguistic facts that also contradict Luce Irigaray's observations on gender assignments (see also the discussion in Livia 2001: 23-30). In inflecting languages, the gender of nouns referring to inanimates is, to a large extent, motivated by phonological rules; it can be predicted according to them, as most grammarians observe (Corbett 1991: 61) and is arbitrary from the semantic point of view. Thus, for example, the Czech noun zivot ('life') is masculine due to its masculine ending, whereas the equivalent Russian designation zizn'is feminine due to its feminine ending. Whereas in everyday language use the gender assignments of inanimate nouns go unnoticed, in language puns or in poetry and literary discourse these assignments may be foregrounded. A good example is the tide of Pasternak's poem, Sestra moja zizn' ('My Sister Life'), where the feminine assignment of the Russian noun for 'life' gives the title a distinct meaning and presents a difficult task for translators who have to render the meaning in other languages: it is almost impossible to find a good Czech translation, as the noun for 'life' is masculine in Czech, evoking thus an association with brotherhood rather than with sisterhood. Translation studies discuss many other examples of associations connected with individual linguistic items differently gendered across languages and cultures. When foregrounded, the gender differences may challenge the notion of arbitrariness even with inanimate nouns in cases when the
28
GENDER, LANGUAGE AND NEW LITERACY
grammatical classification clashes with the semantic features ascribed to the noun. We can conclude that in most situations grammatical gender assignments of inanimate nouns go unnoticed unless they are focused on, deconstructed, i.e. read semantically. Widi animate referents, on die odier hand, gender assignment has a more conspicuous semantic core. The category of gender is related here to the linguistic representation of men and women. 3 Symmetrica l masculine and feminine terms: lexical gender The one-to-one correspondence between grammatical and natural gender in Czech can be best illustrated by nouns denoting the natural sex of persons, such as muz 'man' and zena 'woman', chlapec, hoch, kluk 'boy' and divka, deuce, holka 'girl', and also by kinship terms, such as otec 'father' and matka 'mother', babicka 'grandmother' and dedecek 'grandfather', syn 'son' and dcera 'daughter', bratr 'brother' and sestra 'sister', etc. The specificity of this group consists in the symmetry of the opposite gender counterparts: these nouns are paired by gender but are not derived from each other; they stand in socalled equipollent oppositions (in Prague School linguistic terminology), i.e. both terms are marked and cannot replace one another. This ancient group of paired gender nouns denoting persons is closed and unproductive in contemporarv languages. The Microsoft Word Thesaurus for Czech acknowledges the specificity of this lexical gender group theoretically articulated in the notion of equipollent opposition, including both masculine and feminine counterparts and assigning them meanings and synonyms. The first crucial pair, muz 'man' and zena 'woman', is dealt with as follows:1 (3a)
(3b)
muz: chlap 'chap', 'fellow' (m.) (muzsky'man'- colloquially- 'chap', 'fellow' (m.)); manzel'husband' (m.) (e/iof"spouse' (m.)); vojin'soldier' (m.) (vojdk 'soldier' (m.)). Antonyms: zena 'woman' (f.), zenskd 'woman' - colloquially - (f.); zena: zensk a 'woman' - colloquially (f.); manzelk a 'wife' (f.) (chot' 'spouse' (f.), stara'old' (f.)).
In comparison to entries in Czech dictionaries, die number of meanings and the repertoire of synonyms in the Word Thesaurus is very limited: in many contexts, mt/zcan be replaced by pan 'Mister' (German Herr) andzenabypam 'lady' (German Frau). However, these possibilities are not taken into account in the Thesaurus. Surprisingly, the entry zena does not offer equivalents such as ddma, madam or milostpanrizdy', which are - in contrast to the neutral zena - stylistically marked and belong either to a more formal (ddma) or to an archaic register (madam, mtlostpani). In any case, these equivalents are characteristic of a polite way of referring to women (ddma) or addressing women (madam, archaic milostpani). It may not be the purpose of die Word Thesaurus to display archaic or 'bookish' expressions; the absence of the entry ddma
CZECH ZMEJRKOVA
29
must, however, be qualified as a mistake. The entry zena shows that it might be useful to distinguish between referential meanings and forms of address, as forms of address may preserve more formal or courteous expressions. The Thesaurus, on the contrary, explains the first meaning of zena using a pejorative equivalent zenskd (which can, in certain conditions, also acquire positive connotations, especially in situations of informal evaluation of feminine qualities): (4) To
by la
fakt
peknd
zenskd.
it,1L wasf really beautifulf womanf This was a really beautiful woman. Analogously, the noun muz in the Word Thesaurus is explained starting with a stylistically lower equivalent, chlap 'chap', and provided with the synonym muzsky, which has stylistic connotations analogous to zenskd and the same potential of both pejorative and complimentary usage: (5)
To
by I
fakt
peknej
chlap/muiskej.
it^ was really handsomem chapm This was a really handsome man. In this respect, there is a full symmetry in the explication of the crucial pair muz - zena. What is strikingly asymmetrical is the exemplification of the second meaning, 'husband' - 'wife'. These symmetrical meanings are illustrated by an equivalent chot' 'spouse' (m. and f., respectively); however, there is another equivalent introduced to illustrate the feminine gender, namely stard 'old' (f.). This is a strange solution: first, stard (in mojestard 'my old', f.) is a very expressive, strongly derogative, possibly even vulgar equivalent, used mostly by men in relaxed speech situations (e.g. in a pub); second, it is not clear why an analogous equivalent story /colloquially starej (in muj starej/stary 'my old', m.) is not introduced to exemplify the meaning of husband. The third meaning of muz, contextually bound, namely vojdk/vojin ('soldier'), has no parallel feminine counterpart in die Thesaurus, though a feminine derivative vojinka/vojanda 'soldier' (f.) exists in Czech, and a symmetrical equivalent could be suggested. Such an interpretation would not be in accord with speech practice, due to historical, cultural and social circumstances. One can easily understand the sentence: (6) Ceta md 30 muzu. platoonm has 30 menm The platoon has 30 men. To start with, I am not sure whether vojin/vojdk 'soldier' (m.) can be considered a synonymous meaning of the lexeme muz 'man', and even less could vojinka/vojanda 'soldier' (f.) be considered a meaning of zena 'woman'.
30
GENDER, LANGUAGE AND NEW LITERACY
Another crucial pair, pan 'master' - pant'lady', reads as follows: (7a)
(7b)
pan 'lord', 'gentleman', 'master': muz 'man' (m.) (clovek 'man' (m.)); velmoz 'magnate' (m.) (feudal 'feudal lord' (m.), slechtic 'nobleman' (m.)); vlastnik 'master' (m.) (majiterowner', 'possessor', 'landlord'; zamestnavatel'employer' (m.)); cinitel 'official', 'public figure', 'functionary', 'important person'; manzel 'husband' (partner'partner', spolecnik 'mate', 'companion'); Bub 'God' (Hospodin 'God almighty', Kristus'Christ',Jezis 'Jesus'). pant 'lady': zen a (vdana ) 'wife' (manzelka 'wife', ddma 'lady', 'madam'); zamestnavatelka 'employer' (f.) (majitelka 'mistress', 'proprietress', 'landlady'); slechticna 'noblewoman' (f.).
The asymmetry in the explication of the feminine noun pant is particularly evident in the absence of the meanings connected with the masculine term cinitel (synonym 'official', 'public figure', 'functionary', 'important person'), though such a meaning could be ascribed to the Czech term pani. Due to its restricted range, the Thesaurus does not offer metaphoric or periphrastic idioms expressing various gender stereotypes, such as pan tvorstva 'lord of creation', pani domu 'lady of the house', lepsi polovicka 'better half, 'good lady', etc., characteristic of traditional gender discourse. The third crucial pair, otec 'father' - matka 'mother' reads as follows: (8a)
(8b)
otec. tatinek 'dad' (tdta, tat'ka, tati, papinek 'dad'); starsi muz 'old man' (in rural setting); pater 'father' (knez 'parson'); Buh 'Father' (bozskd osoba 'Heavenly Father'); zakladate l 'founding father' (pfedek 'forefather', tvurce 'inventor'). matka: mam a 'mother' (maminka, mamicka, maticka, matinka 'mummy'); manzelk a 'wife'; pfiznivkyn e 'patroness' (ochrdnkyne 'protectress', dobrvtitelka'benefactress'); samice 'dam', 'female animal parent'; matice 'foundation', 'league', 'matrix', 'nut'.
Some asymmetries in the explication of gendered pairs follow from biologically determined roles (however, it is not quite clear why matka is ascribed the meaning samice 'dam' or 'female animal parent' while the meaning samec 'sire' or 'male animal parent' is lacking with otec). Other asymmetries are related to the distribution of roles in religious (Svaty otec 'the Holy Father', matka predstavend 'Mother Superior'), social and professional life or to various aspects of evaluation of male and female characteristics. Masculines developed different semantic features than their feminine counterparts did due to transposition processes in metaphoric contexts: thus, fathers are associated with founders (otcovezakladatele 'founding fathers'), mothers with life (matka Zeme' 'mother Earth', matka rodu 'progenitrix') and care for wisdom (matka moudrosti 'mother of wisdom'). The latter association is suggested in the Thesaurus explication by equivalents like 'patroness', 'benefactress' or 'protectress'.
CZECH (MEJRKOVA
31
4 Asymmetrica l masculine and feminine terms: masculines an d derived feminines In inflecting languages, such as Czech, there is a strong tendency towards separate terms for males and females reflected in widely applied processes of derivation of feminine counterparts from masculine nouns (the so-called 'motio' - pfechylavdni, in Czech): doktor - doktorka, profesar - pmfesorka, etc. This tendency is significant particularly in the area of professional personal terms: ticitel- ucitelka 'male teacher' - 'female teacher', poslanec - poslankyne 'male deputy' - 'female deputy', ministr- ministryne'male minister' - 'female minister'. The progressive production of feminine counterparts to masculine terms supplying the system of personal professional nouns is to be interpreted, in my view, as a manifestation of the semantic reading of gendered personal nouns, a tendency to reach symmetry between gendered nouns and the linguistic representation of men and women. Over the decades, the social legitimacy and norms of usage of newly-derived feminines have undoubtedly been subject to change in Czech. Five or six decades ago (Travnicek 1949: 426) two competing variants of a feminine for 'doctor' were discussed in Czech linguistics, namely: (9) Pant doktar Mrs doctorm
miparadila me advisedf
vs
Pani doktorka mi paradila. Mrs doctorf me advisedf
In present-day Czech the latter term has become generally accepted, while the former variant is used rarely and exclusively by older people. Czech speakers have completely accepted such feminine nouns as doktorka 'doctor' (f.), starostka 'mayor' (f.), ministryne 'minister' (f.), poslankyne'deputy' (f.), psycholozka 'psychologist' (f.) >filolozka 'philologist' (f.), etc., all of which were rare a few decades ago. They have become a part of neutral vocabulary. A feminine counterpart can be formed practically from any masculine term (occasionally even from masculine epicenes (see below)). Although the productivity of feminine derivation from originally masculine forms is almost unlimited in Czech, and derivation constantly supplies the repertoire of feminine nouns with new items (e.g. professional titles), the frequency of derived feminines in texts is lower than we would expect. This is due to the fact that in gender-indefinite contexts, masculine forms are considered the norm (the so-called generic masculines). According to structuralist theory, it is the unmarkedness of masculine terms that makes the referential range of masculine terms wider than that of corresponding feminine terms (Jakobson 1932; 1971). While in the area of terms for human referents feminines are always female-specific, masculine terms have both male-specific and gender-neutral reference. The concept of privative, i.e. asymmetrical, gender opposition has been repeatedly questioned (see Dokulil 1958). Examples show that the relationship between the two members of a gender pair is sometimes equipollent, i.e. symmetrical:
32
GENDER, LANGUAGE AND NEW LITERACY
(10) Parody sezucastnilo petucitelek a tn ucitele. in meeting participatedm five teachersf and three teachersm Five teachers (f.) and three teachers (m.) participated in the meeting. Jakobson (1932: 74) acknowledged this when he stated that the unmarked category can under certain conditions express the non-existence or absence of the feature - in our case, femaleness - and it can also express the opposite feature, i.e. maleness. Jakobson adds that the latter is perhaps the most frequent function of the unmarked term. However, he points out that such a contextually bound meaning does not contradict the general and basic meaning of an unmarked category. The contextually bound gender-specific meaning of the unmarked category appears in situations of foregrounding, i.e. in situations where maleness stands in contrast to femaleness, as in the above example. When gender is not foregrounded, the opposition remains hidden and the female reference is only implicit in Czech: (11)
Tato televize se snazi oslovit divdka^, dot mun moznost, aby vyjddnl^ sv&j ndzor. This TV [programme] tries to address the viewerm, giving himm a chance to expressm hism opinion.
This holds for singular as well as plural expressions in Czech: (12)
Rdda, cturozhovorysherein, alenedivimsenekterymsvymkole^urr^, ie nechttyts nauindf^ mluvit. I likef to read interviews with actorsm, however, I do not find it strange if some of my colleagues,,, do not wish to talk to journalists,,,.
This interpretation of the category of linguistic gender was formerly widely accepted in linguistic literature. However, it has recently been challenged: the traditional practice of usage of unmarked masculines to refer to males as well as females has been called prescriptive, and the 'generic masculines' have been called 'false generics' (Hellinger and BuBmann2001:9). Leaving aside the question of prescription (there seems to be no evidence for this hypothesis, at least not in Czech), we can discuss the attribution of 'falsity': one must admit that in some contexts the use of masculine forms may create the impression that men are predominant, or that it is only men who are being referred to and that women are hidden behind men who colonized their territory. It is not easy to say which contexts suggest a genderspecific as opposed to a gender-indefinite reading of a noun. First of all, we would have to distinguish between several types of referential situations, from those in which the referent is an individual who may, however, be unknown (hledany vrah 'the wanted murderer'), to those of generic reference, with a wide field of possible referential situations between the two
CZECH (MEJRKOVA
33
poles. Whether we interpret masculine terms in their generic or their gender-specific meaning depends on many factors, including the meaning of a lexeme; the situation referred to; the author's as well as the interpreter's perspective, etc. If we search for the above-mentioned pairs of gendered nouns (doktordoktorka, profesor - profesorka, ucitel - ucitelka, poslanec - poslankyne, ministr ministryne, herec - herecka, starosta - starostka, psycholog - psychobzka, filolog -filolozka, divdk - divacka, vrah - vrazedkyne, etc.) in the Word Thesaurus, we can find only masculines (with the exception of ministr and psycholog, which are missing). Feminine counterparts, no matter how frequent they may be in everyday language, are not included in the Thesaurus. This is a very strange practice indeed, and it is difficult to understand why feminine counterparts to masculine terms have been omitted. The authors of the Thesaurus probably believed that feminine derivation does not change the core semantics of the masculine form; however, this is not always so. Some of the interrelated gender pairs are only partially synonymous, i.e. they share only a part of their common semantic core. For example, a feminine term may lack some of the meanings of its male counterpart: whereas wisfrcan denote 'master', 'specialist', 'artist', 'head of a workshop', and be used as the tide of an outstanding artist or athlete, mistryne can be used only as the tide of an outstanding female athlete, while mistrovd denotes a female head of a workshop. The Thesaurus displays only the entrv misfr: (13)
mistr. znalec 'expert' (odbornik 'specialist', umelec 'artist', virtuoz 'artist'); femeslni k 'artisan', 'craftsman' (dilovedouci 'foreman', part'dk 'overseer'); ucite l 'teacher' (prorok 'prophet', 'guru'); pfeborni k 'champion' (sampion'champion').
In its lack of feminine derivatives, the Word Thesaurus falls behind the accepted practice of current Czech dictionaries. In modern Czech dictionaries masculine nouns are accompanied by their feminine counterparts (if there are such derivations). Thus, Slovnik spisovneho jazyka ceskeho introduces feminine derivatives in brackets following the masculine term, whereas the more recent Slovnik spisovne cestiny introduces both forms (m. is followed by f., often in a shortened form; for an exhaustive analysis of this dictionary from the point of view of gender linguistics, see Dickins 2001). The dictionary of neologisms (Nova slova v cestine) treats both parallels separately in two dictionary entries. In contrast, similar practice can be found only occasionally in the Word Thesaurus. The pair krdl 'king' - krdlmma 'queen'
Whereas the meaning of krdl 'king' is explained as panovntk 'ruler', its svnonyrn being monarcha 'monarch', the meaning of krdlovna 'queen' is explained as (1) panovnice 'female ruler' and/or chot' krdle 'king's wife'; (2) ddma 'queen' in the game of chess; it is to be noticed that an analogous
34
GENDER, LANGUAGE AND NEW LITERACY
meaning in chess is not mentioned with the item krdL Why this particular pair is introduced, as a whole, in both gender manifestations, is difficult to grasp: is it that queens cannot be ignored even if other female professions, includingpresidenika (= feminine counterpart to 'president'), are missing in the Thesaurus; Despite the fact that masculine terms prevail throughout the whole of the Thesaurus, due to the absence of feminine derivatives, it occasionally shows a preference for feminine terms. Though there is a pair krasavec 'handsome man' - krasavice 'beautiful woman' in Czech, it is only the feminine term that can be found in the Thesaurus; it is explained as krdska 'beauty' or fesanda 'pretty woman'. Why the masculine term is missing in the Thesaurus is not easy to guess. 5 Epicenes , their usage in text and their gender interpretation in the Thesaurus The predominance of masculine terms in text and discourse based on their 'generic' usage (reflected most probably in their predominance in the Thesaurus, where masculine terms implicitly and tacitly also represent their feminine counterparts) is obvious. Before we make overly bitter statements about the relationship between the category of linguistic gender and the linguistic representation of men and women in Czech, and particularly about the 'semantic reading of gendered (so far masculine) personal nouns' in Czech, it is fair to discuss counter-examples that can be found in other parts of vocabulary in inflecting languages. What must be taken into account in regard to the Czech language is the large sphere of both masculine and feminine epicenes (i.e. gendered nouns, either masculines or feminines, designating both male and female persons). Whereas the noun clovek 'person', 'man', widely used to denote a human being in general, belongs to grammatical masculines, as do rodic 'parent', jedinec'individual\jednotlivec 'individual', host 'guest', sirotek 'orphan', kojenec 'nursing infant', sourozenec 'sibling', etc., there are several Czech nouns with general personal or gender-indefinite reference belonging to grammatical feminines (with all the syntactic consequences that follow): not only the widely used osoba 'person', but also osobnost 'person', 'personality', bytost 'being', 'creature', existence 'existence', postava 'character', 'figure', figura 'figure', etc. 5.1 Masculine epicenes The central tenn in the category of common gender is undoubtedly the noun clovek 'man', 'person' - plural tide 'people', which is semantically unmarked and neutral (feminine counterpart clovecice 'female human being' is very rare and stylistically marked, created occasionally to foreground the sex opposition as part of poetic licence). The term clovek applies to men as well as to women. The following example is from an interview between two young women:
CZECH CMEJRKOVA
35
(14) Povazujeteseza cuevedomeho ckrveka? consider you yourself an ambitiousm personm Do you consider yourself an ambitious person? The fact that a general personal noun clovek is masculine in most contexts goes unnoticed. Nevertheless, the fact that this masculine term covers both genders can be foregrounded, e.g. in an ironic statement: (15)
tena je taky clovek. Womanf is also a manm A woman is also a man.
It is also useful to take into account that the noun clovek is mostly used in colloquial contexts, its indefinite meaning being sometimes close to that of the pronoun somebody, anybody, one (analogously to German man): (16)
Clovek strdvi ve skole spoustu casu. One spends at school a lot of time.
The Word Thesaurus gives the following meanings and synonyms: (17)
clovek: bytost 'being' (f.) (jedinec 'individual' (m.)); muz 'man' (osoba 'person' (f.), mluvci 'speaker'); kazdy 'everyone' (kdokoli 'anyone', nekdo 'somebody').
If we look at this explanation, we cannot fail to see that grammatical masculine clovek is explicated or paraphrased by grammatical feminine: (1) bytost 'being' (f.) is a feminine epicene as well as (2) osoba 'person' (f.), used in the Thesaurus as a synonym of muz 'man' (m.), together with a noun mluvci 'speaker' (m. and f.), which can be read as masculine as well as feminine. 5.2 Feminine epicenes
The most neutral among female generics is undoubtedly the noun osoba 'person', frequently used in law (osoby cinne v trestnim nzeni(persons active in criminal proceedings'), psychology (zdinsld osoba 'dependent individual'), social life (doprovdzejici osoba 'accompanying person'), and administration and the workplace (ufedniosoba 'official'). Whereas in written texts the noun osoba 'person' is unmarked and genderindefinite, in colloquial speech, in fairy tales and story telling the noun osoba can raise associations with a female person (analogously, the masculine noun clovek 'man' can raise male associations). This association may sometimes be contextually bound, as in the next example: (18)
Stejnetakjako aJkoholik md nutkavou potfebu pit, fetezove zdvisld osoba md potfebu o takoveho cloveka pecovat a ovlddat ho. Samozrejme, ze zrovna tak
36
GENDER, LANGUAGE AND NEW LITERACY
muze ovsem aJhoholikem byt zena a osobou fetezove zavislou jeji manzel nebo partner. As an alcoholicm has an obsessive urge to drink, a chain-addicted personf has an urge to take care of such a personm (man) and to dominate him. Naturally, in the same manner, an alcoholicm may be a womanf and a chain-addicted personf may be her husbandm or partnerm. The Thesaurus gives the following terms: (19)
osoba 'person': jednotlivec 'individual' (jedinec 'individual', individuum 'individual'); ufedni pfedstavitel 'official'; postava 'figure' (in drama); Bozska osoba 'Heavenly person'; clovek 'man' (muzsky 'man', zenskd'woman').
Surprisingly, not only grammatical masculine, such as otec 'father' (m.), but also grammatical feminine, such as osoba 'person' (f.), can be used to refer to God. Moreover, there is another grammatical feminine associated with attempts to refer to God's existence, namely bytost 'being'. Another noun belonging to the category of grammatical feminines is the qualifying noun osobnost 'personage', 'personality'. Unlike the neutral noun osoba, it transmits a positive evaluation ('a remarkable person' in various respects, particularly in the world of science, culture, social and political life, and also in the psychological sense of 'a type of personality', etc.). In more neutral contexts, it covers representatives of various occupations, backgrounds, settings or political opinions. When taking into account feminine epicenes and their functional range in present-day Czech, we should not forget that historical contexts witness a wide use of such designations as Excelence 'Excellence', Magnificence, V\sosi 'Highness', Milost 'Majesty', 'Grace', Svatost 'Holiness', etc., which are grammatical feminines denoting both male and female beings. The natural sex of the persons referred to is expressed by the possessive pronoun jeho 'his' or jeji 'her'; however, these nouns, which are typically found in historical texts, behave syntagmatically as feminines: (20) Jeho Krdlovskd milost pnsla vs Jeji Krdlovskd milost pnsTa. his King's, majestyf camef her King'sf majestyf camefnf. His Majesty the King arrived. Her Majesty the Queen arrived. Common modern gender nouns belonging to grammatical feminines also include connotatively positive nouns used to characterize persons or rather personalities without regard to their natural gender, such as autorita 'authority', used mostly in politics, science and family life; kapacita 'authority', used mostly in science; celebrita 'celebrity', widespread in social life; and the metaphoric noun hvezda 'star', used in culture, particularly for theatre, film, music
CZECH (MEJRKOVA
37
and sports stars - tenisovd, hereckd, filnurud hvezda, hvezda pop-music 'tennis-, theatre-, film-, pop-star'. These feminines are widely used to designate not only female, but also male persons, e.g. hlava 'head', in hlava cirkue 'head of church'. These nouns are grammatical feminines and they require feminine syntagmatic agreement (which can be preserved, however, only to a certain distance in text, e.g., within a sentence or two). In the Word Thesaurus, these feminine nouns are often explained through masculines, but it is also true that masculines are quite often explained through feminine epicenes: (21)
(22)
autorita 'authority': vaznos t 'respect' (vdha 'respect', moc 'power'); znalec 'expert' (m.) (odbomik 'specialist' (m.), kapacita 'authority' (f.)). kapacita 'capacity': schopnos t 'capacity' (vykonnost 'efficiency'); odbornik 'specialist' (m.) (znalec 'expert' (m.)); autorita 'authority' (f-).
As for the pragmatic overtones and connotations of masculine and feminine epicenes, most of the nouns in the group of epicenes are qualifying expressions and can denote both positive and negative evaluations. Whereas the masculine nouns idol 'idol', genius 'genius', drahousek 'darling', mildcek 'sweetheart', etc. express positive qualification, other masculine nouns, e.g. chytrdk 'cunning', hlupak 'blockhead', blbec 'fool', bldzen 'lunatic', opilec 'drunkard', lenoch 'idler', zunvec 'maniac', zbabelec 'coward', pokrytec 'hypocrite', lakomec 'miser', povysenec 'arrogant person', zlocinec 'criminal', snob 'cultural snob', anonym 'anonymous person' have negative associations. As for feminine epicenes, this group also has both positive and negative connotations: see metaphorical nominations of human qualities both positive, such as pilnd vcelka, vcelicka 'diligent little bee' (f.) , and negative, such as obluda 'monster' (L),pfisera 'fright' (f.), baba 'coward' (f.), backora 'pushover' (f.), bdbovka 'sissy' (f.), and those on the border between positive and negative evaluations (velkd) ryba 'fish' (f.), stika 'pike' (f.), 'a cunning person' (f.), /isfca'fox' (f.),etc. Long lists of both male-specific and female-specific nouns existing not only in the literary language but also in various dialects can be found in Eisner (1946: S6&-77). He also analyses numerous examples of cross-gender reference, i.e. denoting a male person with a feminine noun (such as klepna, fnukna, bdbovka, backora, chuderka) and vice versa (vrabcdk, rardsek). Linguists have made two salient observations about the phenomenon of crossing gender lines (see Yokoyama 1999: 422-3, who analysed the situation in Russian): the use of masculine forms to refer to women (1) can be accounted for by Jakobson's thesis about the more inclusive, unmarked nature of masculine gender in Russian; and (2) carries affectionate connotations, generally produces positive effects, while crossing the gender line in the direction of feminine nouns, with reference to men, produces negative connotations. According to Yokoyama, this is evidence of the fact that maleness is more positively
38
GENDER, LANGUAGE AND NEW LITERACY
evaluated than femaleness (for the situation in Russian, see also Doleschal & Schmid 2001). In Czech, the effects of cross-gender reference do not seem to be as conspicuous as in Russian, and both neutral and connotatively charged (meliorative and pejorative) transpositions can be found in either direction. 6 Conclusio n The pairs of gender nouns and their meanings represent a unit which is subject to dynamic development. This regards not only their occurrence in language and their frequency in texts, but also the interpretation of their semantic and stylistic features, as given in dictionaries. The paired lexemes should be listed in dictionaries as independent entries and the meaning of the feminine forms should be described with or without relation to the respective masculine terms. In this respect, the Microsoft Word Thesaurus is deficient Many lexical items lack an entry: not all kinship terms are included in the Thesaurus, probably because they cannot be replaced by synonyms; it must be considered a mistake that some pairs are represented by a masculine term only, e.g. the entry bratr 'brother' lacks its feminine counterpart sestra 'sister', etc., as I have already mentioned, feminine derivatives are not included in the Thesaurus. Overall, there are more than twice as many masculines as feminines displayed (e.g., of 225 animate nouns with initial letter k, there are 152 masculines and 73 feminines; among these, only 2 pairs of masculine and derivative feminines can be found, namely krdl 'king' krdlovna 'queen' and kmotr 'godfather' - kmotra 'godmother'. The criteria used in the selection of items are not clear, as stated above). Although the category of gender in its origin might have been an accident of linguistic history, the gender system began to be interpreted as a semantic category. In the analytical structure of English, gender is already a purely semantic category (Corbett 1991: 137-43; Livia 2001: 29) and has started to be re-analysed as a semantic category, particularly in the context of the feminist critique of language, raising the challenging question of its correspondence to the linguistic representation of men and women. In inflecting languages, the category of gender may have both formal grammatical and semantic readings, and these readings may interfere with each other. In many inflecting languages, semantic interpretations have come to dominate, and in some contexts may have taken over completely. In a highly inflecting language such as Czech, the process of semantic reinterpretation of the category of gender has many internal structural resources and great potential for language change, as witnessed by the large application of derivation of feminine counterparts to masculine terms and the numerous instances of semantic re-interpretation of gendered epicenes. It would appear that this process of reinterpretation has been accelerated by the impact of English, which forces inflecting languages to make further steps in this direction. It might be only a matter of time before the semantic interpretation takes over in Czech as well, especially with masculines, and Czech starts to attempt to develop into 'a more gender-fair language'. For the present, attention to
CZECH CMEJRKOVA
39
peculiarities of the so-called generic masculines occurs more often in linguistic texts than anywhere else. In linguistic texts, for example, one can find a note like: (Pozn.: kdekoliv v tomto textu uzivdme terminu mluvci, partner, ucastntk dialogu aj., my slime tim osoby pohlavi muzskeho i zenskeho.) '(A note: any use of the terms speakerm/f, partner m , participant in the dialogue^/f, etc. in this text should be understood in the sense of both male and female persons.)' (Hoffmannova etal 1999: 128) Though in many discourse practices the use and frequency of generic masculines go unnoticed, it appears that some contexts have shifted generic masculines from the gender-indefinite to the gender-definite pole of interpretation. In these cases, it is expected that masculine terms might have to be interpreted in a more gender-specific meaning than their unmarked, 'inclusive' meaning would predict (see also Leheckova 2002). The above example, however, also shows that although some authors are aware of the issue of 'gender bias', they avoid the technique of splitting suggested by the English instructions for 'gender-fair' language usage. Notes 1 In the following examples, the meanings are in bold, whereas the items given for each meaning ('synonyms') stand in brackets, in italics. References Cmejrkova, Svetla (1995), 'Zena vjazyce' ('Woman in language'). Slovoaslovesnost, 56, 43-55. Cmejrkova, Svetla (1997), 'Jazyk pro druhe pohlavi" ('Language for the second sex'), in Frantisek Danes a kol, Cesky jazyk na pfelomu tisicileti ('Czech at the turn of the millennium'). Prague: Academia, pp. 146-58. Cmejrkova, Svetla (2003), 'Communicating gender in Czech', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 27-57. Corbett, Greville (1991), Gender. Cambridge: Cambridge University Press. Danes, Frantisek (2001), 'Univerzalia a specifika cestiny v obdobi globalizacnfch promen' ('Universals and specifics of Czech in the period of globalisation change'), in Zderika Hladka and Petr Karlfk (eds), Ces'tina - univerzdlia a specifika. Prague: Lidove noviny, pp. 37-47. Dickins, Tom (2001), 'Gender differentiation and the asymmetrical use of animate nouns in contemporary Czech1. The Slavonic and East European Review, 79, 212-47. Dokulil, Milos (1958), 'K otazce morfologickych protikladu' ('On the question of morphological oppositions'). Slovo a slovesnost, 19, 81-103. Doleschal, Ursula & Schmid, Sonja (2001), 'Doing gender in Russian', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 253-82.
40
GENDER, LANGUAGE AND NEW LITERACY
Eisner, Pavel (1946), Chrdm i turz: Kniha o cestine (Temple and citadel: The book on Czech'). Prague: Podrouzek. Fodor, Istvan (1959), 'The origin of grammatical gender'. Lingua, 8,1-41, 186-214. Hellinger, Marlis & Bufimann, Hadumod (2001), The linguistic representation of women and men', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 1-25. Hoffmannova, Jana; Mullerova, Olga & Zeman, Jiff (1999), Konvenace v cestine ('Conversation in Czech'). Prague: Trizonia. Ibrahim, Muhammad Hasan (1973), Grammatical Gender: Its Origin and Development. The Hague: Mouton. Irigaray, Luce (1999), 'Linguistic sexes and gender', in Deborah Cameron (ed.). The Feminist Critique of Language. London and New York: Routledge, pp. 119-23. Jakobson, Roman (1932), 'Zur Struktur des russischen Verbums', in Charisteria Guilelmo Mathesio quinquagenario a disdpulis et Circuti Unguistici Pragensis sodalibus oblata. Prague: PLK, pp. 74-84. Jakobson, Roman (1971), The gender pattern of Russian', in Roman Jakobson. Selected Writings II: Word and Language. The Hague: Mouton, pp. 184-6. Leheckova, Helena (2002), 'Gramaticka kategorie rodu v typologicky ruznych jazycfch' ('Grammatical category of gender in typologically different languages'), in Alena Krausova, Marketa Slezakova & Zdefika Svobodova (eds), Setkdnis cestinou. Prague: UJC AV CR, pp. 50-6. Livia, Anna (2001), Pronoun Envy: Studies in Language and Gender. Oxford: Oxford University Press. Nova slova v cestine('New words in Czech') (1998), Prague: Academia. Slovnik spisovneho jazyka ceskeho (The dictionary of the contemporary Czech Language') (1960-71). Prague: Academia. Slovnik soucasne cestiny (The dictionary of contemporary Czech') (1994). Prague: Academia. Travnicek, Frantisek (1949), Mluvnice spisovne cestiny ('Grammar of literary Czech'). Prague: Melantrich. Valdrova, Jana (1999), 'Zensky obsah v muzske forme - nektera liskalf generickeho maskulina' ('Female content in masculine form - some stumbling blocks of the generic masculine'), in Marie Cechova & Dobrava Moldanova (eds), Jinakost, cizost v jazyce a literature ('Otherness, foreignness in language and literature'). Usti n. Labem: PF UniverzityJ. E. Purkyne, pp. 105-7. Valdrova, Jana (2001), 'Novinove titulky z hlediska genderu' ('News headlines from the point of view of gender'). Nose fee, 84, 90-6. Yokoyama, Olga. T. (1999), 'Russian genderlects and referential expressions'. Language in Society, 28, 401-29.
3 Treatment of 'woman' and 'man' in the Dutch Word Thesaurus Ingrid van Alphen and Alessandra Corda
1 Introductio n The aim of this chapter is to analyse sense distinctions and synonyms of lemmas related to 'woman' and 'man' in the Dutch Word Thesaurus} This analysis applies only to the Dutch language as it is currently used in The Netherlands, and not in Belgium, Suriname and the Netherlands Antilles. Particular attention will be paid to the following questions: 1 Do the data in the Dutch Word Thesaurus reflect the social roles attributed to women/men in Dutch society? 2 From a more lexicographical perspective: how is the Dutch Word Thesaurus constructed? 3 What changes could be suggested to improve the quality of the Dutch Word Thesaurus} First of all, a brief overview of gender treatment in Dutch will be presented (§2), integrated with a few remarks about the social and professional roles attributed to women and men in Dutch society. Then the crucial pairs will be analysed and discussed (§3). Finally, in the conclusion (§4), answers will be provided to the questions presented above. 2 Grammatica l and lexical gender in Dutch 2.1 Grammatical gender* In Dutch there are three grammatical gender classes: masculine, feminine and neuter. The determinative article for masculine and feminine substantives is de, the determinative article for neuter substantives is het? Almost all nouns referring to 'man' and 'woman' in their social, family and occupational roles belong either to the masculine or the feminine gender class, and are therefore so-called *fe-nouns: de moeder 'the mother', de voder 'the father', dezangeres 'the singer' (f.), dezanger'the singer' (m.). An important exception is meisje 'girl', 'little girl', which is a neuter substantive (het meisje)* In pronominal references, however, the feminine
42
GENDER, LANGUAGE AND NEW LITERACY
pronouns zij 'she' and hoar 'her' are used to refer to meisje, and not the neutral pronouns het 'it' and zijn 'its'.5 The same applies to jongetje 'boy', 'little boy'. With the neuter substantive kind 'child' in both senses of 'young person, f./m.' and 'son/daughter', the pronominal reference depends on the context or communicative situation. If there is no doubt about the reference to a male or female person, masculine or feminine pronouns will be used; otherwise neutral pronominal forms will be used, as we see from the following examples: (1)
(2)
A: Waar is dat^ kind^ dat^ hier op straat speelde, ineens gebleveri? B: Ik weet het niet, ik heb het^ helemaal niet gezien.6 A: Where has that child that was playing in the street suddenly disappeared to? B: I do not know, I did not see it at all. A.: Wat heeft Heleen gedaan vanmiddag? B: Dat^ lieue kind^ heeft met hoar/ vriendin Victorine buiten gespeeld, A.: What has Heleen been doing this afternoon? B: That lovely child has been playing outside with her friend Victorine.
Hoofd 'head' and staatshoofd 'head of state' are also neuter. Since the present Dutch head of state is a female person, Queen Beatrix, feminine anaphoric pronouns are used: (3)
Bij rampen met een nationaal karakter, zoals in de Bijlmermeer en in Enschede, toont hetnL staatshoofd^ vaak belangstelling, om zo uitingtegeven aan haar^ betrokkenheid.7 In the case of national calamities, as in the Bijlmermeer and in Enschede, th e hea d o f stat e often shows interest, to express he r involvement.
With substantives that can refer both to male and female persons, the anaphoric reference is determined by the communicative situation: (4)
De getuige mag binnenkomen: zij kannu hoar verklaring afleggen. The witness may come in: she can now give her statement.8
But when one is speaking in general terms about a specific category of persons, the default reference is masculine: (5)
De gemiddelde docent^ Frans kan dit van zijn^ leerlingen niet eisen. The average teacher of French cannot ask this of his pupils.
However, the use of bodi masculine and feminine pronouns is sometimes compulsory for social, not grammatical reasons. For instance, since the
DUTCH VAN ALPHEN AND CORDA
43
Dutch anti-discrimination law9 requires an employer to ask explicitly for both male and female candidates, in advertisements die forms hij/zij 'he/she', zijn/haar 'his/her' are attested.10 The following example comes from a job advertisement for a software developer:11 (6)
Tijdens zijn^ opleiding heeft hij^/zijfi aangetoond voorkeur te hebben voor het werken met PC's. During his study he/she has shown to prefer working with PCs.
The anaphoric pronouns are, however, often only in the masculine form. But even outside legal and official documents, the use of hij/zij, zijn/haar (or zij/ hij, haar/zijn) is quite common in all kinds of texts directly concerning a particular category of persons (such as pupils, children, users, professionals or even people born under a specific zodiac sign). The plural form, in which there is no gender reference, can also be used to avoid die double reference. Compare: (7a) (7b)
De student^ moet zijn^/haarj cottegekaart bij de computerbalie tonen. The student must show his/her student card at die computer helpdesk. StudenteripL moeten hun^ collegekaart bij de computerbalie tonen. Students must show their student cards at the computer helpdesk.
Summing up, anaphoric reference in Dutch is similar to anaphoric reference in English (see Bettoni: this volume). 2.2 lexical gender In diis section some peculiar characteristics of die expression of lexical gender in Dutch with respect to the examined field will be presented, as support for die analysis of the crucial pairs in die Dutch Word Thesaurus. Among the kinship terms, we have observed above dial kind is used both in the sense of 'young person, f./m.' and in the sense of 'son' (zoon), 'daughter' (dochter), irrespective of age: (8)
Olga is een alleenstaande vrouw van tachtig met een land uit hoar eerstt huwelijk en drie kinderen uit hoar tweede huwelijk. Olga is an 80-year-old single woman with one child from her first marriage and three children from her second marriage.
In The Netherlands, where a very progressive form of registered partnership exists, the civil (hetero- or homosexual) marriage is legally equal to the partnership. Moreover, living together without being married or as registered partners is also a very common situation. The terms vriend '(male) friend' and vriendin '(female) friend' are used both for 'friend' in the generic sense, and to indicate a person with whom one is involved in a
44
GENDER, LANGUAGE AND NEW LITERACY
sentimental relation, irrespective of age. A man can use vriendin to refer to his female partner, even if they are both older than 90; teenagers can also use vriend/vriendin in the sense of 'boyfriend', 'girlfriend'. The terms vriend/ vriendin do not imply necessarily that the partners are living together. Man 'man' and vrouw 'woman' have both the meaning of 'male adult person' and 'female adult person' and the meaning of'husband' and 'wife': (9)
Mag ik je even mijn ma n voorstetten ? Mav I introduce my husband to you?
The term of address meurouw (shortened mw.) applies both to married and unmarried women and corresponds in usage to both 'madam' and 'miss'; the terms for a male person are de heer 'mister' (shortened dhr.) and meneer'mister (only vocative).12 Pauwels (2001) observed that in Australian English 'Ms', now used together with 'miss' and 'Mrs', was mostly used by emancipated and working women, feminists, and lesbians. In Dutch mevrouw (mw.) was proposed in the early 1980s as overall term of address, and this successfully became generalized in a few years. Interestingly, man 'man', singular, andjongens 'guys', plural, are commonly used in informal language to address both genders. A girl can sav to another girl: (10)
Da '5 niet leuk, man! That's not nice, man!
Most feminine occupational terms are derived from the masculine ones.13 While the principle of neutralization has die highest priority in English, in contrast to German, where female visibility is the basic characteristic of gender-fair usage (Hellinger 2001), in Dutch, the neutralization and feminization strategies are still concurrent. Moreover, in The Netherlands there is no official language policy aimed at increasing the linguistic visibility of women through the feminization of tenns indicating professions.14 Productive suffixes for the formation of feminine occupational nouns are -#and -ster (for the nomina agentis in -er): agent 'policeman' - agente, psycholoog 'psychologist' - psychologe, kapper 'hairdresser' - kapster. However, many professions, especially the prestigious ones, and tides have no feminine equivalent: arts 'physician', dokter 'doctor', also as academic title, hoogleraar, professor 'professor', rechter 'judge', meestet (title for a person with an MA in Law), notaris 'notary', psychiater 'psychiatrist'. The majority of political functions have no feminine counterpart either: minister 'minister', president 'president', burgemeester 'mayor', sUiatssecretaris 'state secretary'. In some cases a neutral, non-specific form is used: kamerlid 'member of the parliament', gemeenteraadslid 'member of the local council', bewindspersoon 'member of government/cabinet'. The masculine and feminine forms bewindsman, bewindsvrouw do exist and are also widely used. This is also the case for male or female leaders of a parliamentary party: pactieleider
DUTCH VAN ALPHEN AND CORDA
45
or jractieleidster. Notice that wethouder 'alderman' has a feminine equivalent which is scarcely used: wethoudster. Outside political functions, a neutral form is leerkracht 'member of the teaching staff. There are a few nomina agentis in -oar that indicate professions, such as ambtenaar 'civil servant', kunstenaar 'artist' and leraar 'teacher'. The feminine form lerares is the most widely used, also in compound words: according to a quick Google search (but a more accurate study is needed) many female artists seem to prefer to be called kunstenaar instead of kunstenares. Google again shows that ambtenares is a marked, connotated term, only used ironically outside official documents. The only compound word with leraar for which a feminine form lerares is not attested is hoogleraar 'professor' and it is not likely that diings will change if the numbers of female professors at Dutch academic institutions do not increase significantly. Nowadays a lot of women work as family doctors (huisarts), but this has not yet led yet to die creation of a feminine form for huisarts. As in German, -man '-man' and -vrouw '-woman' are also used in compounds indicating professions15 and are very productive, e.g. brandweerman 'fireman' - brandweervrouw, vuilnisman 'garbage collector' - vuilnisvrouw, raadsman 'adviser' - raadsvrouw, spoiiman 'sportsman' - sportvrouw. Even for vroedvrouw 'midwife', the masculine variant vroedman is attested. Huisman 'houseman' is a widely used counterpart of huisvrouw 'housewife', -man and -vrouw are quite productive and can be used in new compounds: for instance, if in the division of domestic tasks the man usually does the shopping, he can say about himself thuis ben ik de boodschappenman 'at home I am the shopping man'. In the case of chirurg'surgeon', piloot 'pilot' and dirigent 'conductor', the feminine forms chirurge, pilote and dirigente can be derived, but vrouwelijke chirurg 'female surgeon', vrouwelijke piloot 'female pilot' and vrouwelijke dirigent 'female conductor' are also possible. A quick Google check shows that chirurge, pilote and dirigente are more frequent than vrouwelijke chirurg, vrouwelijke piloot and vrouwelijke dirigent, For rector 'headmaster of a secondary school', 'university rector' there are two feminine equivalents, rectrix and rectrice. A female headmaster can choose from three different forms of address: mevrouw de rector, mevrouw de rectrix and mevrouw de rectrice. Up to now, as far as we are able to state, no female university rector has chosen the feminine equivalents; this only happens at secondary school level. Some professions of French origin in -eur, with a feminine form in -euse, are only used in the masculine form: adviseur 'consultant', programmeur 'computer programmer', rechercheur 'detective'. Regisseur 'art director' is, however, also used in the feminine form, like other professions of French origin in -teur with a feminine form in -trice: directeur 'manager' - directrice, inspecteur 'inspector' - inspectrice, redacteur 'editor' - redactrice. Interestingly, it seems that in professions connected with intellectual creativitv (regisseuse, redactrice), the feminine form is more widely used than in professions related to a hierarchical organization (directrice, inspectrice). Most people will therefore associate inspectrice with a low-ranking female school inspector, not with the leading
46
GENDER, LANGUAGE AND NEW LITERACY
figure of a police station. Similarly, a Google search shows that directrice seems to be used more often to indicate a school headmaster, or the head of a care centre or of a cultural or non-profit institution, than to indicate a female manager of a firm.16 As a matter of fact, in traditional male functions, even if a feminine form does exist, a lot of women prefer to use the masculine one, because it is said to be more prestigious. This situation clearly illustrates the ideologically-based 'feminization' vs 'neutralization' problem of choosing between terms which make women visible (e.g. 'pilote'), but which may carry derogatory connotations, and terms which do not distinguish women from men, but which may therefore render women invisible (e.g. 'pilot').17 Professions of English origin are used in the English form and have therefore no feminine equivalent: accountant, consultant, manager. In job advertisements for civil servant posts, the indication m/v or v/m 'm./f.', 'f./m.' is compulsory. Interestingly, in the case of two traditional female jobs, secretaresse 'secretary' and leidster 'day care teacher', the corresponding feminine occupational terms are also used if the employee is a man, both in official advertisements and job contracts and in everyday language.18 Finally, as in other languages, also in Dutch 'double gender' examples can be found: vrouwelijke actrice 'female actress' (234 occurrences in Google), vrouwelijkeschrijfster 'female writer' (29 results), mannelijke acteur 'male actor' (283 results), mannelijke schrijver 'male writer' (34 results).19 It seems that in some cases we can speak of a contrastive usage (in this context there is a clear opposition man-woman), but this phenomenon needs further study. 2.3 Emancipation in Dutch society
Although Holland is generally considered to be a progressive country, both in the field of sexual emancipation and cultural tolerance, social roles for women and men were quite traditional until a few years ago. The reasons can be found in a complex of historical, economic and political factors, which can better be analysed and discussed in other publications. We will only mention the most recent developments. Until 1980, the government policy, mostly inspired by religious and Christian Democratic parties, did not encourage women to keep their jobs after they married.20 For decades, working mothers had to face social criticism and practical problems: differently from Scandinavian countries, day care centres, pre-school and after-school care centres were almost non-existent up to 1980. As a consequence, girls were not encouraged to continue their studies or to achieve a higher educational level. The situation changed very gradually in the last two decades of the twentieth century, when it became clear that the growing costs of the welfare state could not be financed without significant female participation in the labour market. Consequently, the government began to encourage part-time jobs, flexible contracts and reintegration possibilities for women. The dominant family model evolved from man with full-time job and woman
DUTCH VAN ALPHEN AND CORDA
47
as housewife to the present one, in which both parents work. In the lower classes, however, the dominant model is still the older one. Some figures21 can illustrate this development: in 1994, both parents worked in 33 per cent of the families with children, in 2004, in 57 per cent of the families. Compared to other European countries, female work participation in The Netherlands in 2004 was quite high: 65.8 per cent, behind Sweden and Denmark (71.6 per cent), but at the same level as Finland and the UK.22 But most women have small part-time jobs and are not economically independent.23 Indeed, in 2005 The Netherlands was die European country1 with the highest number of persons working in part-time jobs: 46.2 per cent (European mean: 18.5 per cent). Some 75.3 per cent of the women and 22.6 per cent of the men had a part-time job. It is also interesting to note that, compared to other European countries, in The Netherlands the number of female ministers, secretaries of state and members of parliament is relatively high,24 but die figure for top managers in public functions is not high25 and the figure for university professors is very low.26 The places available in pre- and after-school care facilities are now twenty times higher than in 1990. But there is still a negative attitude towards childcare services, especially among people with a low or medium education level, who prefer informal care via family or friends. Half of both women and men still reject formal childcare services for babies. The majority of men and women believe that the mother-child relationship does not suffer if the mother goes out to work: 10 per cent of the women who gave birth to a child in 2003 gave up their jobs, compared to 25 per cent in 1997. However, mothers choose to work less after the birth of their first child, while fathers seldom do so. A problem for further emancipation is represented by the employers: many of them consider part-time working to be a negative feature for the organization, and think that part-time cannot be combined with management functions. Since 1997, women have made up half the students at university level. Female students in higher professional and university education graduate faster and more frequently than their male peers. 3 Analysi s of the crucial pairs As a general preface to this discussion, we should remark that sense distinctions in the Dutch Word Thesaurus are not based on clear linguistic criteria (this is also often the case in monolingual and bilingual dictionaries, see Corda 1998). A closer examination of sense distinctions will in many cases show that a lemma is used in a particular sense only in a limited number of contexts or in fixed expressions, and that some senses are more autonomous and productive than others. The Dutch Word Thesaurus does not differ in this respect from other lexicographical resources, both on paper and digital. A second point to be made is that in the Dutch Word Thesaurus sense distinctions are only clarified through synonyms, not through definitions;
48
GENDER, LANGUAGE AND NEW LITERACY
but since it is impossible to find a synonym for every sense, this choice is often not informative or is even misleading. Finally, no usage or register information is given for the synonyms (and this is also a weak point in many monolingual dictionaries and in the Thesaurus for odier languages). The user is supposed to know that some words are pejoratives or that they are only used in literary or written contexts. Moreover, in the Dutch Word Thesaurus synonyms are always listed in alphabetical order: this means that the best synonym can also be the last one in the list. It is important that these three elements are taken into account in the analysis of the crucial pairs. The Dutch Word Thesaurus is an attempt to create a web of synonyms for the Word user, and this attempt is still far from perfect. The 'traditional' representation of female family and social roles is also partially a consequence of the lexicographical choices at the basis of the Thesaurus. To illustrate this point, the structure of the lemmas vrouw 'woman' and man 'man' will first be examined in detail, and then the proposed sense distinctions and synonyms will be discussed. 3.1 Vrouw and man In the following example (11), sense distinctions and synonvms of vrouw and man in the Word Thesaurus are listed:27 (lla) vrouw 'woman': 1) eega (eega 'spouse', echtgenoot 'husband', echtgenote 'wife', man 'man'/'husband', wederhelft 'other half), 2) maagd (maagd 'virgin', jonkvrvuw 'lady', madonna 'madonna', meisje 'girl'), 3) mei d (meid 'girl\jongedame 'young lady', meisje 'girl'), 4) mevrouw (mevrouw 'madam', dame 'lady'), 5) Moede r 'mother', 6) Prime s (Prinses 'princess', dame'lady'), 7) zij (zz/'she', meisje'girl'). (lib) man 'man': 1) mens 'person, human being', 2) echtgenoot 'husband', 3) kerel (kerel'fellow, guy', vent 'fellow, guy'), 4) koppen 'heads, hands (members of the crew on a ship)', 5) Persoon 'person', 6) boer 'man'. Examining the senses of vtrmw, the following remarks can be made: • The general sense 'female adult' is missing, perhaps because no adequate synonym for the definition could be found. • echtgenoot 'husband' and man 'husband' are not correct, synonyms of sense (1) 'wife': they are synonyms of eega 'spouse' and are also listed under this lemma; • vrouw can never be used in sense (2) maagd 'virgin': this sense is also not listed in dictionaries. Why then is this sense listed here? We could hypothesize that this is the result of the automatic procedure used to establish relations between lemmas and synonyms in the Thesaurus. The same synonvms are also present in the lemma maagd, under the first sense jonkvrouw, 'maid, unmarried young lady', in the lemma jonkvrouw under the
DUTCH VAN ALPHEN AND CORDA
•
•
•
• •
49
second sense maagd and in the lemma madonna. This could indicate that synonymical relations in the Dutch Word Thesaurus are considered to be perfectly symmetrical and automatically applied: however, this hypothesis does not explain why diese synonyms are not found in the lemma meisje 'girl'. It is also possible that the Thesaurus is a result of a semi-automatic procedure (meisje could have been added later by hand to the list of synonyms) or that the algorithms used in the automatic procedure take into account other elements. Sense (3) is unclear and also not listed in dictionaries. Meid is still used in the meaning of 'girl', but only in specific contexts, often in spoken language. The lemma meid has under the sense jongedame 'young lady' meisje and vrouw as synonyms, and meisje and vrouw are also given as synonyms of the lemma jongedame under the sense meid. In the lemma meisje under sense (3) dienstmeisje 'maid (servant)' meid is given as a synonym. Again, the hypothesis is that this sense has been automatically added because vrouw was a synonym for both meid and jongedame. Sense (4) probably refers to the use of vrouwe as address form, that nowadays only survives in fixed expressions like Vrouwe Justitia, 'LadyJustice'. It would probably be more helpful for the users to create a separate lemma Vrouwe. Sense (5) and sense (6) do not exist: vrouw can never be used in the sense of mother or princess. But again, since vrouw is given as a sense (not a synonvm) for moeder and as a synonym for prinses in dame, these pseudo-senses were automatically created. Sense (7) is probably also an automatically added pseudo-sense, since the lemma zij has vrouw as a synonym under the sense meisje. The sense 'queen' (playing cards) is missing.
Looking at the senses of man we notice the following: the difference between sense (1) and sense (5) is unclear: bodi mens and persoon mean 'person'. Moreover, mens is the general term for 'individual of the human sort (f./ m.)',28 and in Dutch man never was nor is productively used in this sense. However, man can have a generic sense in fixed expressions and collocations like iets aan de man brengen 'to sell something' or op de man of 'straightforward, straight from the shoulder'. Also, in the singular form and always preceded by a quantifier man means, depending on the context, 'men' (12) or 'people, persons' (13). For example: (12) (13)
Het laatste halfuur heeft Ajax met slechts tien man gespeeld, Ajax played the last half an hour with only ten men. Het is een grote manifestatie, waar meer dan duizend man op afkomt. It is a big event, which attracts more than a thousand people.
On the other hand, it is not always possible to substitute quantifier + mensen 'persons, people' by quantifier + man. The semantic constraints under which this substitution is possible need further investigation.
50
GENDER, LANGUAGE AND NEW LITERACY
Interestingly, vrouw can also be used in this way, but it is less frequent and never has the sense 'people'. (14)
Een groep van twintig vrouw. A group of twenty women.
In our opinion, there is therefore no reason to include a sense 'person': • In sense (2) no synonyms are given, which is strange, because the synonyms eega and wederhelft listed under the sense 'wife' in the lemma vrouw can also be used as synonyms for this sense 'husband'. In the lemma echtgenoot 'husband', the same synonyms are given as the ones in vrouw under the sense 'wife'. • Sense (3) is also unclear: kerelalso means 'strong man', and this leads us to suppose that this sense refers to the use of man in fixed expressions such as 'behave like a man'. Again, this is not a productive sense of man. • Sense (4) is indicated with koppen, the plural form of kop 'head'; koppen can also be used in the sense of 'crew members' on a ship. • Sense (5), 'person', see above. • Sense (6) refers to use in compound words which indicate a profession, such as sportman 'sportsman', klusjesman 'handyman', and is therefore not a 'sense' of the lemma man. Boer actually means 'farmer, peasant', and it is unlikely that all Thesaurus users will see that in this case boer has been chosen as a 'definition' because it is also used in compound words like melkboer 'milkman', groenlenboer 'greengrocer' and recently pizzaboer 'pizza deliverer'. • The general sense of 'male adult' is missing. In example (15) a suggestion for a new structure of the two lemmas is proposed: (15a) vrouw 'woman': 1) vrouwelij k volwasse n persoon 'female adult', 2) echtgenote (echtgenote 'wife', eega 'spouse', wederhelft 'other half), 3) vrouw op spelkaart 'queen (cards)'. (15b) man 'man': 1) mannelij k volwasse n persoo n 'male adult', 2) echtgenoot (echtgenoot 'husband', eega 'spouse', wederhelft 'other half). Summarizing, the most relevant conclusions are: • The basic senses of female adult and male adult were missing, other senses were either missing or badly explained. • No attempt has been made at symmetry between the two lemmas. As shown in §3.2, this lack of consistency is also to be found in other pairs. Overall, evidence seems to support the hypothesis that the only symmetrical relations in the Dutch Word Thesaurus are the results of automatically created connections between lemmas through senses and synonyms.
DUTCH VAN AIJ>HEN AND CORDA
51
• Since extending synonymical relations in the way described above is quite arbitrary, non-existing senses ('pseudo-senses') have been listed in both lemmas. • The 'pseudo-senses' 'person' and 'individual' added to man suggest that diis lemma is synonymous with the neutral mens 'human being', while this has never been the case in Dutch (apart from the contexts described above), in contrast to English and other languages suffering under 'the myth of neutral "man"' (see Moulton (1977), Spender (1980), Hellinger & BuBmann (2001)). The 'pseudo-senses' added to vrouw 'woman' do not suggest a generic use of this lemma, but, on the contrary, specifically focus on typical female stereotypes: the virgin, the mother, the young girl. This is an androcentric bias in die lexical representation of vrouw vs man. 3.2 Social and family roles
The analysis of the lemmas indicating social and family roles was conducted on 13 pairs of lemmas29 and 6 lemmas with non-specific reference30 (see Table 3.1 at www.lingue.unibo.it/wordthesaurus). The most relevant results for the purpose of this chapter are: 1 Looking at the structure of the 13 pairs of lemmas, for lemmas with female reference, 31 senses and 83 synonyms are listed, for lemmas with male reference, 34 senses and 80 synonyms are listed. From a purely quantitative point of view, there is thus no substantial difference in the amount of information. 2 Synonyms of lemmas with a specific female or male reference are expected to have eidier the same specific female or male reference, or a generic reference. In the 13 lemmas referring to female roles, 4 (in a total of 5 senses) have as synonyms lemmas referring to male roles; in the 13 lemmas referring to male roles, 2 have (in a total of 2 senses) as synonyms lemmas referring to female roles. Vrouw in the sense of 'wife' and echtgenote 'wife' have as synonyms echtgenoot and man (both in die sense of 'husband'). Vriendin in both senses of 'friend' and 'partner' has as a synonym wiend 'friend', 'partner', and minnares '(female) lover' has as synonyms minnaar, amant and vrijer, all '(male) lover'. Echtgenoot 'husband' has as synonym vrouw. Minnaar '(male) lover' has as synonyms minnares and maitresse. In the two pairs echtgenoot - echtgenote, minnaar - minnares the same synonyms occur, but the symmetry is not perfect: the male term has one more sense without synonyms. The symmetry in the synonyms is probably due to the automatic construction of the lemmas. There are thus twice as many cases in which lemmas referring to female roles have as synonyms lemmas referring to male roles, indicating a lack of symmetry in the construction of the lemmas. 3 The number of synonyms and senses with generic reference is almost the
52
GENDER, LANGUAGE AND NEW LITERACY
same for both categories. In the 13 lemmas referring to female roles, 6 lemmas for a total of 7 senses have as a synonym a lemma with generic reference (12 neutral lemmas as synonyms); in the 13 lemmas referring to male roles, 5 for a total of 9 senses have as a synonym a neutral lemma (13 neutral lemmas as synonyms). The lemmas with female reference have 5 senses defined by a generic word, the lemmas with male reference 6. On the whole, there is a slight preference for using words with generic reference in senses and synonyms of lemmas with male reference. 4 Among synonyms of lemmas with generic reference (kind 'child', partner), both lemmas belonging to a pair with reference to specific female and male roles are expected to be listed (next, of course, to lemmas with generic reference). Here we find omissions: neither^'ongm 'boy' nor meisje 'girl' are given as synonyms of kind 'child'. On the other hand, both vriend 'friend' and vriendin '(female) friend' are listed among the synonyms of partner. There is in this case no bias in the treatment of the lemmas with female reference. 5 Obsolete words, words only used in specific contexts or in the written language and lemmas with a low frequency are well represented (both as lemmas and synonyms), while common lemmas are missing. Ouders 'parents', bruid 'bride', bruidegpm 'bridegroom', weduwe 'widow', weduwnaar 'widower' are not represented. The only obsolete term which is missing is vrijster 'spinster'. For instance, the unmarked adolescent 'adolescent' is a technical psychological term and is given as a synonym ofjongen 'boy', while the common words for 'teenager', tiener and puber, are missing. The obsolete juffmuw and yM/jfer'miss' are given as synonyms of dame 'lady'. The adjective ongehuwd 'unmarried' has as synonym celibatair which means 'living without sexual contacts', with reference to religious life: ongetrouwd 'unmarried', which is also listed, is the only possible synonym. 6 Lemmas with female reference like 'woman', 'girl', which do not exclusively express a family or affective relation, are first of all defined on the basis of these stereotypical relations, while this is not the case for the male counterparts. For example, in the case of meisje 'girl' - jongen 'boy' only for meisje the sense 'girlfriend' (verloofde, vriendin) has been defined, but jongen can also be used in the sense of 'boyfriend'. For jongen the senses of kind 'child' and adolescent have been distinguished, and, though these senses also apply to meisje, they have not been mentioned there. Meisje is thus not defined on the basis of a semantic feature like age, but of relational aspects like 'daughter', 'girlfriend' and 'maid (servant)'. Interestingly, under verloofde 'fiance (e)' only meisje is given as a synonym and not jongen: the Thesaurus has in this case again chosen a male and not a neutral perspective. Vrouivelijk 'female, feminine' is only defined through the antonym mannelijk 'male, masculine', while the lemma mannelijk is quite fully developed, with synonyms connected with the stereotypes of strength and virility. Since vrouwelijk has no synonyms but only an antonym, the
DUTCH VAN ALPHEN AND CORDA
53
implication is that vrouwelijk is defined by the absence of the 'male' characteristics. Notice that in the pair vrouw 'woman' - man, 'man', vrouw has been defined as eega 'wife' and moeder 'mother', while man is only defined as echtgenoot 'husband', not as voder 'father'. Moreover, eega is the first of the senses of vrouw. 7 Derived and metaphorical senses are more often listed under words with male reference. See voder 'father' (the senses stamvader, grondlegger 'founder') - moeder 'mother'; zoon 'son' (the sense leerling 'pupil') - dochter 'daughter'; minnaar 'male lover' (the sense aanbidder 'admirer') - minnares 'female lover'; mannelijk 'male, masculine' (the senses flink and stoer 'strong') - vrouwelijk 'female, feminine'. The most relevant general conclusions are: 1 Lemmas with female reference like 'woman', 'girl', which do not exclusively express a family or affective relation, are first of all defined on the basis of these relations - often from an androcentric perspective - and not on the basis of autonomous, intrinsic semantic features (like age or sex). 2 Lemmas with female reference have fewer derived, metaphorical senses. 3 In the synonyms of the pairs there is no symmetry. Words with male reference are more often listed among synonyms of words with female reference than the other way round. 4 There is a slight preference for using words with generic reference in senses and synonyms of lemmas with male reference. 5 The automatic generation of relations between senses and synonvms can contribute to multiply and emphasize the biases in the representation of lemmas with female reference listed above under 1,2,3 and 4. 3.3 Activities, functions, economic and professional roles
The analysis has been conducted on about 150 lemmas indicating professions, activities or public functions.31 As a general remark, professions and functions as a whole are not well represented in the Dutch Word Thesaurus and many common terms are missing. For instance, the pairs huisvrouw 'housewife' - huisman 'houseman', bewindsman- bewindsvrouw 'minister, secretary of state', inspecteur- inspectrice 'inspector', rector- rectrix 'headmaster of a secondary school, university rector', danser - danseres 'dancer' are missing. This is also the case of words related to public or economic functions such as werkgever 'employer', wetgever 'legislator', minister" minister', renter'judge' and for specific professions like schoenmaker 'shoemaker', brandweerman 'fireman', timmerman 'carpenter', kok 'cook'. Also many activities traditionally carried out by women, and for which a male equivalent does not exist, are not represented: among others vroedvrouw
54
GENDER, LANGUAGE AND NEW LITERACY
'midwife', ballerina 'ballerina', schoonheidsspecialiste 'beautician', kraamverzorgster, kraamhulp 'maternity nurse', secretaresse 'secretary', typiste 'typist'. The feminine equivalent of a profession, activity or function is normally not listed in the Thesaurus, with very few exceptions. This is not only the case for traditional male professions such as zakenman 'businessman', pilot 'pilot', politieman 'policeman'. Even for professions which are traditionally seen as female, often only the lemma with male reference is given: boer 'peasant, farmer' but not boerin, kapper'hairdresser' but not kapster. Even in artistic and sport activities, in which a female presence was earlier accepted as normal than in economically oriented activities, the word with female reference is notmentioned: sportman' sportsman' butnotsportvrouw, kunstenaar''artist' but not kunstenares, zanger 'singer' but not zangeres. Terms with generic or male reference related to 'leading', 'directing' and 'managing' are quite well represented, also in their synonymical relations: baas 'boss', manager 'manager', bestuurder 'governor, director', directeur 'director', leider 'leader', chef'boss, leader', meerdere, superieur 'superior', hoofd 'head', aanvoerder 'leader'. Of the few terms in this field with female reference (like inspectrice, directrice, rectrix), only bazin 'mistress' (of an animal) and 'lady of the house' is present. In only one case is a female profession mentioned while the male equivalent is not listed: naaister 'seamstress, needlewoman' (with as synonyms coupeuse 'tailor's cutter' and modinette 'seamstress', both obsolete), a traditional female activity (but one which nowadays has almost disappeared). 3.3.1 Analysis of relevant pairs Only in a few semantic areas, words with female reference are represented next to the male ones, so that relevant pairs can be found: the field of 'monarchy' and the activities related to 'acting', 'teaching', 'nursing', and 'serving (in restaurants and at home)'. With one isolated exception, in these cases, too, the representation of female words is biased in different ways. • Monarchy: the field of 'monarchy' is represented by the pairs koning koningin 'king' - 'queen', keizer- keizerin 'emperor' - 'empress' and prinsprinses 'prince' - 'princess'. Since in The Netherlands only queens have been head of state since 1898, and only princesses have been heiresses to the throne from 1880 up to 1967, it could be supposed that this situation of female dominance would somehow be reflected in the treatment of these lemmas in the Thesaurus, The presence of prim-gemaal 'prince consort' among the senses of prins is the only indication that supports this hypothesis; for the rest, there are several cases of bias in the representation of the female terms. The male terms have more senses and synonyms; moreover, the female terms koningin and keizerin have the male ones as synonyms but onlv keizer (and not koning) has koningin and keizerin as synonyms. Vorst 'monarch' is given as a synonym of the male terms, and also of koningin and keizerin, but the female equivalent vorstin is missing.32
DUTCH VAN ALPHEN AND CORDA
55
• Acting: among the terms for 'acting', two pairs are given: acteur- actrice 'actor' - 'actress' and toneelspeler - toneelspeelster '(theatre) actor' '(theatre) actress'. For three other terms, speler 'player', filmacteur '(film) actor' and figurant 'extra, walk-on', only the term with male reference is mentioned. Lemmas with female reference are not listed as synonyms in terms with generic or male reference. On the contrary, lemmas with male reference are given as synonyms of lemmas with female reference. Moreover, if both lemmas belonging to a pair are represented, words with male reference have more synonyms. • Serving: the terms in the pair ober - serveerster 'waiter' - 'waitress' have exactly the same synonyms. In this case the words widi female reference (serveerster and the obsolete diensterand kelnerin) are treated in the same way as the words with male or generic reference. The synonyms (among which both words with generic and non-generic reference are found) are, as always in the Thesaurus, also listed as lemmas: all the lemmas have the same sense bediende 'employee, servant' and the same synonyms. These relations must have been automatically generated, since the synonymy relation applies for every lemma with the superordinate bediende, but not with all other lemmas (for instance, serveerster is not a synonym of klerk 'clerk'). As noticed also for other languages, the terms related to 'servant (girl)', 'maid', which refer to live-in staff, a situation which was normal decades ago among rich people, are over-represented: dienstbode, dienstmeid, meid, meisje, kindermeisje,** dienstmeisje, gedienstige (obsolete),34 This is also true for the male equivalents: knecht, loopjongen, dienaar, butler, oppasser. Since, as we already noticed, many common professions and activities are missing, it is not clear why this semantic field deserves so much attention,35 given that it is no longer relevant. • Nursing: the area of 'nursing' represents the only exception to the normal treatment of pairs with generic equivalent in the Thesaurus. As a rule, the generic word is always given as a synonym of the word with male reference, and this also applies the other way round. In this case, the only synonyms of the generic lemma verpleegkundige 'nurse, male nurse' are zuster and verpleegster, both with female reference. Verpleegster'nurse' has as synonyms zuster 'nurse, sister' and verpleegkundige, while the male equivalent verpkger has only one synonym, breeder'male nurse'. • Teaching: in the teaching profession women are now over-represented at primary school level,36 while this is not the case at secondary school level." Moreover, in all educational sectors women are under-represented at the higher levels.38 Thus, despite the emancipation policy started in 1993 by the Ministry of Education, female teachers are even nowadays first of all associated with primary education, and this is also reflected in the Thesaurus: in the pair leraar- lerares 'teacher', the only synonyms of lerares are onderwijzeres and schooljuf. Both terms refer to female teachers in primary schools. Among die synonyms of leraar, on the contrary, we find both terms that indicate teaching in primary education (meester, onderwijzer) and
56
GENDER, LANGUAGE AND NEW LITERACY
also more general terms (decent and ondenurijskracht, which apply at all educational levels). The female equivalent docente, however, is not mentioned in the Thesaurus. Leerkracht and ondenurijskracht, both with generic reference, have only male or generic words among their synonvms. Schoolmeester (a male teacher in primary education) has as synonvms docent and leraar, but schooljufjrouw has as synonyms jufjrouw and juf(these terms only indicate a female teacher in primary education), and juf has as synonyms jufjrouw and schooljufjrouw. Only schooljuf has lerares as a synonym (probably automatically generated, since schooljuf is a synonym of lerares}. From the examination of the above data the following conclusions can be drawn: • Female equivalents of professions are usually not given in the Thesaurus, with a few exceptions, and these exceptions are all connected with activities in which female presence has been common for a very long time. • As already seen in the case of lemmas connected with social and family roles, male terms have more senses and synonyms and they are more frequently given as synonyms, both of words with generic and with female reference. • Even in care and domestic activities, in which women have traditionally been employed, and in fields like education, characterized by strong feminization, the treatment of female words is biased in various ways, reinforcing asymmetric positions between women and men. • The fact that professions and activities reflecting the social relations of a century ago are over-represented contributes to emphasizing the traditional, subordinate female role in economy and society. • In the educational sector, despite the growing number of women in secondary and higher education, women's presence in the Thesaurus is limited to primary education.
4 Conclusion : man & machine-made language From both a qualitative and a quantitative39 point of view, the Dutch Word Thesaurus would benefit from some improvements. Generally speaking, the use of synonyms to explain senses (which characterizes all versions of the Word Thesaurus) has many shortcomings, and these shortcomings are reinforced by the use of automatic procedures to generate synonym relations among lemmas. Short definitions, like those in bilingual dictionaries, would be much more helpful. More flexible and informative would be the implementation of a clickable word map, which would not only list synonyms and antonyms, but would also report collocations and examples as well as word frequencies, facilitating associations which are not only based on the synonym-antonym relation.4*
DUTCH VAN AU>HEN AND CORDA
57
In the second place, not all the lemmas in the Thesaurus have synonyms, while, on the other hand, the Thesaurus is not a complete list of all words with synonyms. Why have lemmas without synonyms been listed? The answer is probably to be found in the automatic generation of senses, antonyms and synonyms, which means that if a word is used to define one of the senses of a lemma, or to indicate an antonym or a synonym of a lemma, then that word will automatically be added to the list of the lemmas. Also problematic is the lack of information about usage and register, together with the lexicographic choice to put the synonyms in alphabetical order within every sense, with the result that, even for competent language users, it is sometimes difficult to find the right synonym. Finally, in the semantic fields examined in this article, a large number of obsolete or rare words are present, while current and frequent words are missing. This makes the Thesaurus much less useful for Word users than it might be. The treatment of sense distinctions and synonyms in lemmas related to 'woman' and 'man' does not reflect the family, social and economic position of women in current Dutch society. The representation of words with female reference is biased in different ways, confirming the results of analyses conducted for other languages. For lemmas with female reference usually fewer senses and synonyms are given; in particular, derived and metaphorical senses are less frequently listed. In the family and social field, female terms are first of all defined on the basis of their relation with other terms (or male persons), not on the basis of autonomous semantic features, like age or gender. In synonymy relations there is a clear imbalance between terms with male and generic reference, on the one hand, and terms with female reference, on the other. There are more internal synonym correspondences among lemmas with male and generic reference, while lemmas with female reference do not often occur as synonyms. Almost always (with an isolated exception), if for a profession terms with generic, male and female reference are available, the generic word is given as a synonym of the word with male reference, and this also applies the other way round, while this is not the case for the terms with female reference. Moreover, terms with male reference are more often given as synonyms of female terms than the opposite, and female terms are much less frequently given as synonyms of terms with generic reference. Finally, as we have seen, obsolete words or words connected with obsolete social and professional functions are listed, while common, current words are missing. This is true for words with both male and female reference, but has of course a much stronger negative impact on the words with female reference, because words connected with the traditional female roles are over-represented. Our investigation of the Dutch Thesaurus reveals that this twenty-firstcentury application contains old-fashioned and very stereotypical assumptions about social roles for women and men in Dutch society. The Dutch Thesaurus is an example of the dangers of a 'man-made machine', in which the automatic generation of senses and synonyms enhances the alreadv
58
GENDER, LANGUAGE AND NEW LITERACY
existing biases in the lexical representation of 'woman' and 'man'. We cannot but conclude that the Dutch Thesaurus is wholly permeated by the ideology of MAN (male as norm),41 affected and infected by the idea that male/ masculine is the higher, more prestigious category and that female/feminine is secondary and subordinate. This makes of die Dutch Thesaurus a modern version of the Trojan horse, inside our city walls. Unfortunately, the Thesaurus cannot (yet) be modified or manipulated by the user, while this feature would certainly improve its flexibility. Meanwhile, the spell-check is the only electronic instrument provided by Word that can learn from its mistakes and dial we are allowed to educate. Notes 1 The data are based on the Word 2002 Thesaurus by Polderland Language Speech and Technology © 1996-2000. We also quickly checked a previous version: our impression is that the most recent one has not been substantially improved, only minor changes having been made. 2 See, for a more detailed discussion, Gerritsen (2002b). 3 The deictic pronouns are hij 'he', ze/zij 'she', het 'it', ze/zij 'they' (f./m./nt.), deze 'this one' (f./m. sg.; f./m./nt. pi.), dif'this one' (nt. sg.);
DUTCH VAN ALPHEN AND CORDA
15
16
17 18
19 20 21
22 23 24
25 26
27 28 29 30 31 32
33
59
i.e. van Alphen (1985) and Brouwer (1985); further van Alphen (1996) and Gerritsen (2002a). -heer 'sir', 'mister', 'master' is also used (like -herr in German) in compounds indicating activities, but is not productive any more. The feminine counterpart is -vrouw: see gastheer- gastvrouw 'host' - 'hostess', beschermheer- beschermvrouw 'patron'-'patroness'. Among other evidence, a Google search done on 11 January 2006 shows 819 occurrences of directeur van een bedrijf ('director of a firm/company') vs 18 occurrences of directrice van een bedrijf. See Hellinger (2001). However, along with peuterkidster and kleuterleidster, peuterleider and kleuterleider are also used (peuter indicates a child between ages 2 and 4, kleuter between ages 4 and 6). On the contrary, secretaris indicates a high hierarchical function and is never used as the masculine counterpart of secretaresse. Google search on 1 April 2005. Until 1957 female civil servants would lose their position if they got married. The source of these data is the Emancipatiemonitor 2004, a publication of the Sociaal en Cultureel Planbureau (SCP) and the Centraal Bureau voor de Statistiek (CBS). The source of the European comparison data is Eurostat, http://europa.eu.int/ comm/eurostat/ In 2001,41 percent. Respectively, 31 per cent and 37 per cent; compare UK (33 per cent, 18 per cent), Germany (27 per cent, 32 per cent), France (21 per cent, 12 per cent), Spain (21 per cent, 28 per cent) and Italy (10.5 per cent, 12 per cent). Some 26 per cent: compare UK (30 per cent), Germany (27 per cent), Spain (32 per eent),Italy (19 percent). Some 6 per cent: compare UK (12 per cent), Germany (6 per cent), France (15 per cent), Spain (15 per cent), Italy (12 per cent). See for slightly different comparative data http://europa.eu.int/ comm/research/science-society/women/ wssi/downindi_en.html. In this and in the following examples, the 'senses' (numbered) are in bold, whereas the 'synonyms' given for each meaning stand in brackets, in italics. Interestingly, mens is a de-noun (see §2.1), but can also be used as het-noun (nt.), and in this case it always refers to women. Sixteen pairs were examined: two were not found, in one case only the lemma with male reference was present (vrijgezel, 'bachelor'), but the counterpart vrijster 'spinster' is obsolete. Vrijgezel is still used. Nine lemmas were examined, three were not found. To save space, Table 3.2 available online at www.lingue.unibo.it/wordthesaurus lists only a few crucial pairs and relevant lemmas. Queen Beatrix's husband, Glaus, had the tide prins-gemaal, not koning, 'king'. If Willem Alexander, Beatrix's son, becomes king, his wife, princess Maxima, will hold the title koningin, queen. A female head of state such as Beatrix is called koningin, but acts juridically as the koning, the head of state according to the Constitution. This explains why Claus could not take the title fonmgwhile Maxima will be called koningin. For kindermeisje 'nanny', the common terms with generic reference oppas, babysit 'babysitter' are given as synonyms - but a kindermeisje lives permanently with the family, while an oppasor babysit does not.
60
GENDER, LANGUAGE AND NEW LITERACY
34 Werkster 'cleaning lady' has three senses: 'bee', 'cleaning lady* and 'servant (girl)', 'maid'. It is confusing that the least common sense, 'bee', is given as first. 35 Also the presence of stiermvechter, matador and toreador can hardly be justified as tvpical Dutch professions. 36 Some 68 per cent in 2000, this figure will probably grow to 77 per cent in 2011, according to Vermeulen et oL (2000). 37 According to the Kerncijfers 1998-2002 (71), in 2002, women represented 35.7 per cent of the teachers at secondary level, the average age being 47.1 for men and 41.3 for women. More than 60 per cent of the female teachers work part-time iNota Werken in het Onderwijs 2003, §1.3.). 38 See Thermometer vrouw and management, http://www.minocw.nl/documenten/ emancipatie-doc-2004-thermo_vrouw_en_man.pdf 39 It is impossible to state how many lemmas have actually been listed; however, we made some comparisons with the alphabetical list of the Van Dale Dutch-Spanish bilingual dictionary (Slagter 1992, about 70,000 lemmas), and found that the Thesaurus contained from 50 per cent up to 70 per cent fewer words. 40 See for instance the Visual Thesaurus of English, http://www.visualthesaurus.com/ index.jsp, Wortschatz at the University of Leipzig, http://wortschatz.unileipzig.de/ (also for Dutch), the Edinburgh Word Association Thesaurus, http:// www.eat.rl.ac.uk/, Wordnet, http://www.cogsci.princeton.edu/~wn/ and Lexical FreeNet, http://www.lexfn.com/. 41 Hellinger (2001: 108).
References Alphen. Ingrid C. van (1983), 'Een vrouw een vrouw, een woord een woord'. Tijdschrift voor Vmuwenstudies, 14, 307-15. Alphen, Ingrid C. van (1985), 'Eine Frau — ein Wort: Uber die Gleichbehandlung von Frauen und Mannern und die Konsequenzen fur Berufsbezeichnungen im Niederlandischen', in Marlis Hellinger (ed.), Sprachwandel und feministische Sprachpolitik: Internationale Perspektiven. Opladen: Westdeutscher Verlag, pp. 123-31. Alphen, Ingrid C. van (1996), 'Beroepsnamen en beroepskeuzen v/m. Een sociolinguistische analyse van de beroepskeuzen van 15-jarige meisjes en jongens in Nederland'. in Marlies Lcegwater-Van der Linden (ed.), Taalen beeldvarming; overvrouwen en mannen. Ministerie van Onderwijs, Cultuur en Wetenschappen: Zoetermeer, pp. 42-54. Brouwer, Dede (1985), 'Anders, aber gleich? Uber die Bildung weiblicher Berufsbezeichnungen im Niederlandischen', in Marlis Hellinger (ed.), Sprachwandel und feministische Sprachpolitik: Internationale Perspektiven. Opladen: Westdeutscher Verlag, pp. 132-47. Corda, Alessandra (1998), Treatment of senses and collocations in dictionaries: an issue for lexicology and lexicography', The Structure of the Lexicon in Functional Grammar. Amsterdam and Philadelphia: Benjamins, pp. 195-211. De Caluwe.Johan (1994), 'Is een vrouwelijke ingenieur een ingenieuse?'. Nederlands van Nu, 42,51-4. Eurostat = http://epp.eurostat.cec.eu.int/ Geerts. Guido; Haeseryn, Walter; De Rooij, Jaap & Van den Toorn, Marten C. (eds) (1984), Algemene Nederlandse Spraakkunst. Groningen-Leuven: Wolters-Noordhoff/ Wolters. Gerritsen, Marinel (20()2a), 'Changes in professional terms in the Netherlands:
DUTCH VAN ALPHEN AND CORDA
61
Anglicisation and the neutralisation of gender'. Linguistics in the Netherlands, 18, 101-11. Gerritsen, Marinel (2002b), 'Dutch: towards a more gender-fair usage in Netherlands Dutch', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages. The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 81-108. Hellinger, Marlis (ed.) (1985), Sprachwandel und feministische Sprachpolitik: Imternationale Perspektiven. Opladen: Westdeutscher Verlag. Hellinger, Marlis (2001), 'English - gender in a global language', in Marlis Hellinger & Hadumod BuBmann, Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 105-14. Kerncijfers (1998-2000) = Ministerie van Onderwijs, Cultuur en Wetenschappen (2003), Kerncijfers 1998-2000. Den Haag: Ministerie van OcenW. http:// www.minocw.nl/begrodng/kerncijfers9802/ Moulton,Janice (1977), The myth of the neutral "man"', in MaryVetterling-Braggin, Frederick A. Elliston & Jane English (eds), Feminism and Philosophy. Totowa, NJ: Adams & Co, pp. 124-37. Nota Werken in het Onderwijs 2003 = Ministerie van Onderwijs, Cultuur en Wetenschappen (2003), Nota Werken in het onderwijs 2003. Den Haag: Ministerie van OcenW. http://www.minocw.nl/lerarenbeleid/werkenonderwijs/print.html Pauwels, Anne (2001), 'Spreading the feminist word: the case of the new courtesy title Ms in Australian English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 137-51. Slagter, Peter Jan (ed.) (1992), Handwoordenboek Nederlands-Spaans. Antwerpen: Van Dale Lexicografie. Sociaal en Cultureel Planbureau/Centraal Bureau voor de Statistiek (2004), Emancipatiemonitor 2004. http://www.cbs.nl/nl/publicaties/publicaties/maatschappij/ leefsituatie/emancipatiemonitor-2004-dl-l.pdf. [English summary: http:// www.scp.nl/english/publications/summaries/9037701906.html] Spender, Dale (1980), Man-Made language. London: Routledge & Kegan Paul. Van Langendonck. Willy & Beeken, Jeannine (1995), 'Vervrouwelijking van beroepsnamen?'. Kultuurleven. Maandblad voor cultuur en samenleving, 6, 16-25. Van Santen, Ariane (1996), 'Minister of ministerin'. Taalschrift, 3, 30-2. Vermeulen, Marc; Vink, Rob & Vos, Klaas de (2000), Arbeidsmarktramingen primair onderwijs 2000-2011. Tilburg: IVA.
4 Gender in the English Word Thesaurus Camilla Bettoni
I Introductio n
It is no accident that Microsoft Word Thesaurus was first conceived in English. The USA have been leaders in software design, Microsoft is an American firm, its staff are mosdy native speakers of English, and the majority of their customers, in and beyond the USA, are also English-speaking. Furthermore, the English language has a long and illustrious history of lexicography, and of dictionary and thesaurus compilation - not to mention the fact that it was the first language to see a revolution in lexicography thanks to corpus linguistics.1 It is also no accident that feminist gender-studies first flourished in English-speaking countries. These are places democratic and wealthy enough to afford these studies, and their language is male-oriented enough to offer interesting subject matter. Yet, as the following analysis will show, the treatment of women in the three English versions of the Microsoft Word Thesaurus is in striking contrast to the international role of the English language as a model democratic institution, its glorious tradition of lexicography, and its important achievements in feminist studies and gender reforms. Items referring to women in the Thesaurus are often outdated, sometimes incorrect, occasionally offensive. In this chapter, §2 briefly introduces the gender categories of the English language, and §3 reports on how feminist studies have interpreted them. Then, §4 discusses English as a global and pluricentric language, with its main varieties: American, British and Australian English, while §5 shows how dictionaries and thesauri have dealt with them. In §6, there follows a comment on some crucial Thesaurus lemmas, as found in Word 6 and Word 11. Finally, in §7, some minimal proposals are put forward for changes. If accepted, they may result in a fairer, more modern and, ultimately, more useful tool. 2 Gende r in the English language Compared to most other Indo-European languages, gender in English is now a weak linguistic category. It is no longer an overt grammatical category, and as a semantic one it affects only a restricted class of personal nouns. In a
ENGLISH BETTONI
63
covert way, however, it is persistent and pervasive, and is therefore of great social relevance. Grammatically, English does not require that all nouns be identified as feminine, masculine or neuter. Nor, as a consequence, do adjectives, determinants and other parts of speech conform to the gender of the nouns they refer to and formally agree with them. Semantically, gender in English is a clear category affecting some personal nouns, such as girl, queen vs boy, king. Although these nouns are not formally marked with gender-specific endings, they include among their semantic properties either the [+female] or [-i-male] specification. This in turn determines the formal choice of either the she or he anaphoric pronouns. However, within the whole English lexicon, nouns with a lexicalized gender are few, mainly referring to people. Pronoun choice of she, rather than it, referring to certain common nouns, such as countries, ships or cars, is now old-fashioned. Even most personal nouns, such as child, millionaire, have no lexical gender, and can refer to either female or male referents. When the gender of the referent of these gender-less nouns is known, the choice between their pronouns she or he is linguistic and obligatory. On the other hand, when the gender of the referent is unknown or irrelevant, English allows, at least in principle, a margin of choice. As usual, whenever there are variants, the choice has sociolinguistic relevance and implications. Traditionally, most people, women as well as men, refer in sheterms to nouns such as nurse or whore, and in he-terms to nouns such as surgeon or gangster. Ostensibly, they base this semantics on the statistical probabilities regarding the professions and social roles of their female and male referents. More arguable is the still pervasive practice of using the masculine generic he for nouns such as pedestrian or patient, whose referents are distributed more or less equally among the population. Even more gender-biased is the current masculine pronominalization of nouns such as (schoolteacher or (dishwasher) user, despite the strong statistical probability that their referents are women. Lexically, English can mark the feminine gender of the referent with three main techniques. First, a suffix can be added to the noun referring to a male. The most productive suffix is -ess, as in actress, princess vs actor, prince. Another suffix, much rarer, is -ette, as in brunette, usherettevs brunet, usher. Other gendermarking suffixes include -ine and -we, as in heroine, aviatrix vs hero, aviator. Second, the femaleness of the referent can be marked by adding a whole explicit word, as in female attendant, woman doctor, grr/ reporter vs attendant, doctor, reporter. Third, the added word can be more implicit but no less equivocable, as in pretty lawyer, emotional negotiator vs (brave) lawyer, (tough) negotiator. In these last examples, as we move further away from grammaticalized and lexicalized gender to the levels of discourse and pragmatics, it becomes clear that we also move away from intrinsic features of language as a system towards peculiarities of language usage. Here the choices available to speakers are wider, and the covert conspiracy of the gender-bias of the English language and culture is at the same time more subtle and more apparent
64
GENDER, LANGUAGE AND NEW LITERACY
(see e.g. Johnson 2000: 71-112). For example, in the clause structure, females are said to be cast as agents less often than males; and in conversation with males they are likely to talk less than their interlocutors, and to answer questions rather than asking them. The overall effect is that in gender terms, the male is the controlling code and the female the responding code, where males aim to control things, events, and most crucially people, while females show more responsiveness to things, events, and particularly people (Povnton 1985: 87). All these aspects of both overt and covert gender biases in the English language have been subjected to close scrutiny by feminist linguists. 3 Gende r reform in the English language English is the language that enjoys the most extensive and profound history of linguistic description, and this holds true for the area of gender as well (Hellinger 2001: 106). Since it would be impossible to give here even the shortest review of the main results reached by descriptive studies, first, I will only mention the basic principles which inform suggestions for redressing the gender imbalance of the English language, and then give die example of the title Ms, to illustrate the achievements and failures of gender reforms which have operated on those principles. In the English-speaking world, recommendations have mostly taken the form of guidelines for genderneutral language, issued by public and private organizations who wish positively to promote linguistic equality or negatively to avoid being branded as discriminatory. Most famous are the UNESCO guidelines of 1999. There are two basic principles by which the gender imbalance can be redressed: one is to give women the same visibility that men enjoy, the other is to reduce unnecessary male visibility. In the first case, the strategy adopted is that of feminization or engendering, in the second that of neutralization or degendering. For example, at the lexical level, for nouns of occupations, feminization would solve the asymmetry by using more often the existing feminine forms or by filling gaps in die lexicon if these do not exist. Words like waitress, chairwoman, lady dentist would be recommended for singular female referents, and the formulae waitresses and waiters, chairwomen and chairmen, lady dentists and dentists for plural generic referents. Neutralization, instead, would favour epicene nouns for all referents. A woman, as well as a man, would be referred to as waiter, chairperson, dentist, and women only, as well as men only, and women and men together, as waiters, chairpersons, dentists. Between diese two alternatives, the principle of neutralization has the higher priority in English, unlike in other languages that have grammatical gender (Hellinger 2001: 109; Romaine 2001: 156). This option, of course, runs die risk of making women even more invisible, as epicene terms make most people think of males only. But it avoids the cumbersome use of both feminine and masculine terms for the general meaning, and may have a better chance of being more widely accepted. These two alternatives are not mutually exclusive in a language, and
ENGLISH BETTONI
65
guidelines can offer a variety of solutions that imply either feminization or neutralization. In areas other than the lexicon, feminization may be seen as preferable also for English. Nor are these two options the solutions to all problems. Neither engendering nor degendering language will fix the gender imbalance at the level of discourse and rhetoric, if women are given visibility through linguistic expressions that end up trivializing, stereotyping, or denigrating them. All feminist language reformers agree about avoiding expressions such as career woman and family man, unless career man and family woman become equally widespread and 'normal'. The tide Ms was first promoted in the 1970s as symmetrical to Mr in order to replace both Mrs and Miss, and thus avoid characterizing women according to their (often only presumed) marital status. Three decades later, it has gained some use, but its meaning has turned out differently from the reforming intentions of its promoters. Far from having outmoded the other two terms - especially Mrs, still widely used - Ms is a third choice now available in English. As such, it reflects identity and ideology, both at an individual and at a societal level. Since being married is still seen, even by women, as more prestigious than being unmarried, most married women prefer to be addressed and to identify themselves as Mrs. Ms can therefore be regarded as indicating that a woman is single, separated, divorced, widowed or in a de facto relationship, or that she is a feminist who thinks that her marital status is her own business and does not need to be disclosed publicly. Romaine (2001: 158-9,165-6) for British and American English, Pauwels (2001) for Australian English, and Holmes (2001:119-22) for New Zealand English show that, albeit with some quantitative variation in the different countries, this flagship of feminist reform is more likely to be used by younger, well-educated and urban women themselves than chosen by others in addressing any woman. Overall, despite innumerable anti-discriminatory recommendations covering all aspects of language and the proliferation of guidelines which make them available and often impose them, language reform is slow to affect widespread practice. Non-sexist English is at present consistently employed only by a minority of users, and even by these mostly in writing and when speaking in public. Conversational usage is still in flux, and alternatives between traditionally conservative and feministically advanced, or between feminists' engendered and degendered ones - have become indexical of a variety of meanings. 4 Varietie s of English As a global language (Crystal 1997), English now counts roughly 1,500 million speakers. Some 320 million use it as a native language (ENL), as in Ireland and New Zealand; the rest, well over a billion, use it as a second language (ESL), as in India2 and Nigeria, or as a foreign language (EFL), as in Sweden and Japan. Ever since its spread around the world in the age of exploration five hundred years ago, through emigration and imperialism English has developed numerous geographical varieties which have taken on
66
GENDER, LANGUAGE AND NEW LITERACY
an independent history. However, only in the second half of the twentieth century was it widely recognized as a pluricentric language (Kloss 1978; Clyne 1992). Also the tripartite model of ENL, ESL and EFL, now widely used (Gorlach 1995: 61; McArthur 1998: 42-3), was systematized only in the 1970s (see Strang 1970: 17-18; Quirk et al 1972: 3). There is nothing simple about the nature of the growth of English around the world, and the question of how many varieties should be distinguished cannot be answered conclusively once and for all. Asking it, however, is not purely an academic game, because it forces us to give both a set of social reasons for the distinction and a set of defining linguistic elements for each postulated variety. In fact, uncontroversial language-in-society categories do not always have unambiguous equivalents in (socio) linguistic ones. For example, socially, the distinction between ENL, ESL and EFL is obvious. Yet Gorlach (1995: 61) questions whether there are enough linguistic features which would identify a text as ENL, as against ESL or EFL - once of course we disregard questions of speaker's competence, and we relate the text to a norm. As a global and pluricentric language, English has been shaped by a tension between centrifugal and centripetal forces, because its speakers have a dual identity, reflected through their own national or social variety, and through membership of the wider international speech community. Centrifugal forces give rise to varieties. First, English was only 'Britocentric' for several centuries, then it also became increasingly 'Americocentric' in the twentieth century. In fact, the focus for English nonnativity was even more restricted to the higher-level social usage of south-eastern England on the one hand, and of the north-eastern United States on the other (McArthur 2002: ix). From each of these 'rival' centres, nonnative features would radiate outwards, and become hazier the further one moved, socially and politically as well as geographically, from the core regions, eventually allowing other varieties to gain independence. Some two decades ago, Benson et al (1986:14-15) had no doubt that British and American were still the two predominant varieties, because they have the largest number of native speakers, have the most extensive literatures, and are the varieties normally taught to non-native speakers. In the new century not everybody would agree with excluding Australian English. Who's Centric now? is the title of a collection of papers read at a conference held in Canberra (Moore 2000). There are good reasons for the inclusion of Australian English among the predominant varieties (as well as increasingly fewer reasons for the exclusion of others, such as those of Canada, New Zealand, South Africa). Not only has Australian English been endonormative (see Ammon 1989) for quite some time, but its speakers are also highly conscious and proud of the peculiarities of their language, as witnessed by the impressive achievements of their own lexicographic tradition, culminated in the production of the Macquarie Dictionary. Furthermore, Australian English is now widely taught to learners in the South-East Asian region. Microsoft has done well to include it among the languages in the Tools' section of Word. It is certainly the third
ENGLISH BETTONI
67
predominant variety, albeit in a less strong position than the other two supervarieties. Centripetal forces, on the other hand, give unity to English and point to a common core across borders and oceans. Variously called World English, Global English, Common English, International English, or International Standard English, the number of labels and the vagueness of these terms are due to a certain vagueness of the linguistic phenomena they denote. Overall, today convergence, rather than divergence, seems to dominate the development of the varieties of English used mainly as a native language, especially as far as their lexicon is concerned (see Gorlach 1995:112; Hansen 1997: 63; McArthur 2002: 257-8). This is due partly to intensified international contacts, and partly to the still pervasive normative influence of British English and American English. Furthermore, even these two supervarieties are becoming increasingly blurred (Gorlach 1995: 112; Trudgill & Hannah 2002). On the other hand, divergence seems more typical of the varieties used mainly as a second language. Their drift may be due to nationalistic trends, and to the lack of a common educational policy (Hansen 1997: 62). Yet, for the future, a cautionary note must be accepted (see Gorlach 1995: 45), as the balance between the competing demands of intelligibility and identity is delicate and easily affected by changes such as swings in immigrant policy, different political alliances or new population trends (Crystal 1997:139) .3 How can we then come to terms, practically, with the current - and increasing - range and mobility of so many users of English: native, secondlanguage, foreign users, distributed socially as well as geographically in many regions? No matter where we are, we cannot predict the vocabulary, the accent or the spelling, the grammar, the discourse and pragmatic patterns of the next English text we come across, whether face to face, on television or telephone, in print or on the net. Luckily, some tools are available to help us. Among them, dictionaries and thesauri are important. 5 Englis h language thesauri English is the language that has been studied most lexicographically, and one of the few languages for which computerized corpora are available, which greatly facilitates the analyses of usage in various contexts and styles (Hellinger 2001: 106). No attempt is made here to give even the shortest review of the many achievements of such a long-established scholarly and commercially fruitful tradition. I only wish to comment briefly about how three thesauri treat the gender bias of the English language. First, let us remember what a thesaurus can offer. Unlike a dictionary, which provides an account primarily of the referential meaning of separate words, a thesaurus is a kind of dictionary-in-reverse, where words associated with a particular idea corne together, overflowing the boundaries of their dictionary definitions. It is this sense of the flow of language that
68
GENDER, LANGUAGE AND NEW LITERACY
contributes to the pleasure, and often the humour, of reading a thesaurus. Attached to this sense of flow, as well, is an element of danger.
This danger, according to Tardif (1986: vii), the author of this quotation from the Preface to The Pocket Macquarie Thesaurus, lies in the fact that scarcely any word in English can be replaced without some change in meaning - that is, there are hardly any true synonyms. Let us now see in Table 4.1 what 'flow of language' the compilers of The Cambridge Thesaurus of American English (1994) and those of The Oxford Paperback Thesaurus (1994) have provided for two entries: maternal and paternal Why should, in both thesauri, the 'flow' for maternal be shorter than that for paternal"? Why, in the Cambridge Thesaurus, should a father, but not a mother, be solicitous and concerned'? Why should a mother, but not a father, be loving"? Why, in the Oxford Thesaurus, should maternal overflow with feelings, while paternal do so with interest, and with interest in the boy rather than in a child or a girl? And why - even in the maternal entry - should grandparents be his rather than hers? It would seem that the real 'danger' for women may not be that envisaged by Tardif, so much as that of being denied the possibility of enjoying 'the pleasure, and often the humour, of reading a thesaurus'. Nor would it seem that matters are improving fast, if- compared to the Paperback of 1994 - the 2004 edition of the new 2002 Oxford Thesaurus of English has, in the maternal entry, changed maternal feelings into the baby aroused her maternal instincts; and, in the paternal entry, changed take a paternal interest in the boy into some employers felt paternal concern for their workers. 6 Microsof t Thesaurus analysis We start from a summary of the analysis by Bazzanella el aL (2000) of some 100 gender-related entries of US English in the Thesaurus located in Word 6. Then we proceed to illustrate the main changes introduced for US English by Word 11. Finally, we compare American English with British English and Australian English in Word 11. Our analysis takes into account the same terms of the original study, but only several representative pairs are discussed here. For lack of space, antonyms will not be mentioned. 6.1 US English in Word 6 This is, generally, an unreliable lexical tool, in the sense that it can arbitrarily include and exclude lemmas, and for those included miss important meanings and synonyms or give plainly wrong ones. Within our set of words, for example, it can include bachelor and exclude spinster, and then for bachelor give only the meaning of recipient of a diploma. It can give maternal as the sole meaning for motherly, but give kind (rather than paternal} as the sole meaning for fatherly. It can also give mad scientist as a synonym for inventor (which is misleading), and papal see for chairwoman (which is incomprehensible). Furthermore, the unreliability of the tool extends to the order in which items
ENGLISH BETTON1
69
Table 4.1 Comparison in the treatment of two entries in three thesauri Thesaurus
Comparison of 'maternal' and 'paternal'
maternal adj. motherly, parental, protective, loving paternal adj. fatherly, patrimonial, benevolent, protective, solicitous, concerned Oxford Paperback Thesaurus maternal adj 1 maternalfeelings motherly, protective, see MOTHERLY. 2 his maternal grandparents on one's mother's (1994) side, on the distaff side. paternal adj 1 take a paternal interest in the boy fatherly, fatherlike, patriarchal, protective, concerned, solicitous, kindly, benevolent. 2 his paternal grandfather on the father's side. Oxford Thesaurus of English maternal > adjective 1 the baby aroused her maternal instincts (2004) motherly, maternalistic; protective, caring, nurturing, loving, devoted, affectionate, fond, warm, tender, gentle, kind, kindly, comforting, compassionate. 2 his maternal grandparents o n one's mother's side, on the distaff side. paternal > adjective 1 some employers felt paternal concern for their workers fatherly, fatherlike, patriarchal; protective, vigilant, concerned, solicitous, kindly, warm, friendly, benevolent, compassionate, sympathetic. 2 his paternal grandfather on the father's side, patrilineal, patrimonial.
Cambridge Thesaurus of American English (1994)
are listed. For example, it is hard to find a criterion, other than randomness, for the order of the synonyms for husband: male, groom, spouse, mister, old, hubby, man, consort, bedmate; or for wholesaler, merchant, industrialist, entrepreneur, capitalist, broker, man in a grey flannel suit, manager, retailer, businessman. Finally, this randomness becomes even more confusing where synonyms which are mostly sociolinguistically neutral occasionally mix with highly marked ones, in terms of both slang and obsolescence; see, for example, brick given as a synonym for gentleman, and villein for farmer. The Word 6 Thesaurus can also, more specifically, ignore women altogether or marginalize them. First, some examples on the good side: where the English language has two commonly used opposites for women and men, both are normally listed, as happens for daughter and son, queen and king; also included can be a third gender-neutral lemma which encompasses them both, not only for established triplets such as girl, boy and child, but also for more modern ones such as chairwoman, chairman and chairperson; occasionally it can even happen that the female-related lemma is included and die male excluded, as happens with widow and widower. On the bad side, however, not infrequently the female lemma in a pair is missing: bride, hostess and businesswoman are not listed, while bridegroom, host and businessman are. Women can also be ignored where English has one term for both sexes. For example,
70
GENDER, LANGUAGE AND NEW LITERACY
millionaire can mean only rich man, and consequently have as synonyms only gender-neutral terms such as tycoon and capitalist, or male terms such as indeed rich man or man of means:, and while owner can mean proprietor and thus refer to woman or man, its synonyms include landlord but not landlady. Furthermore, compared to men, women can be marginalized even where they are not downright ignored. Where both lemmas in a pair are listed for search, most often the meanings and synonyms for the female lemma are fewer in number. For example, while mother has only one meaning (parent), father has four, three as nouns (dad, originator, priest) and one as verb (beget)', while wife and. husbandhave both one meaning each (respectively woman and male), the former has seven synonyms (woman, spouse, lady, old lady, missus, little woman, helpmate), the latter nine (male, groom, spouse, mister, old, hubby, man, consort, bedmate); and while lady and lord have three meanings each (respectively woman, woman of good breeding, high-born lady and divinity, ruler, nobleman), lady has five synonyms (woman, female person, adult, matron, dame, all attached to the first meaning), and lord has altogether 22 synonym (divinity, god, Jehovah, messiah, savior, Christ, the supreme being, ruler, master, governor, sovereign, monarch, king, prince, nobleman, aristocrat, count, don, peer, hidalgo, seigneur, magnate, distributed among its three meanings). Finally, as perceptive readers will have already seen from the examples of wife vs husband, and lady vs lord, quality as well as quantity has a negative effect on the representation of women. Why, as a synonym for wife, should a woman be little? Why should all meanings for lady be fairly similar to each other and exclusively human, so as to anchor her safely to earth, while the very first meaning for lord is divinity, so as to place him right away in heaven? Also, why should actress include such a negative synonym as starlet, while no negative synonyms are mentioned for actor'? On the other hand, why should actor include such a positive synonym as artist, and actress exclude it? Yet not all bad news is for women only. Our US English Thesaurus is an unreliable tool which hits mostly women, but does not always miss men. Take the womanlike and manlike pair. If womanlike can mean feminine and consequently be substituted by nine 'feminine' synonyms (feminine, delicate, womanly, etc.), why cannot manlike mean masculine? Instead, according to Word 6, manlike can mean only anthropoid, and consequendy have only six anthropological' synonyms (anthropoid, humanoid, anthropomorphic, etc.). These are legitimate, as English can give gender-neutral meanings to manrelated terms. However, to deprive men of their equally legitimate male and masculine meanings is depriving them of their first and foremost characteristics. Summing up, given the generally conservative, conventional and condescending way in which women are treated, and more specifically, given that to be womanlike is to be delicate, gentle, soft and unmanly, would we not be justified in expecting to find that to be manlike is to be strong, rough, hard and unwomanly, and - why not? - also an artist and a millionaire? Our expectations are not met. Consistency is not a quality of the US English version of the Word 6 Thesaurus.
ENGLISH BETrONI
71
6.2 Changes in US English from Word 6 to Word 11 The latest version of the US English Thesaurus has introduced several important changes, first and generally, in the context in which it appears, second and specifically, in the treatment of its gender-related items. First, the new Thesaurus is presented in a different context. When, under 'Tools', you click 'Thesaurus', a 'Reference Tools' window appears which contains two tools, not just one, as Ward 6 did: under the 'Word or phrase' box, where you type in the searched lemma, you find first the Encarta World English Dictionary, then the Thesaurus. Even if by default the Dictionary does not open to the searched lemma automatically, as the Thesaurus does, it can be made to do so readily. In any case, the Thesaurus appears lower down on the screen, as the second tool. Furthermore, at the bottom of the 'Reference Tools' window, below the Thesaurus, there are two links to the Internet, marked 'Search Encarta Encyclopedia' and 'Search MSN', through which you can conveniendy consult these two tools also. Second, presumably because a dictionary is thus easily available, the information in Word 11 is substantially reduced. Quantitatively, Table 4.2 offers some figures. Bazzanella et al. (2000) list 104 lemmas, 4 of which are not found in the Thesaurus; for the remaining 100 lemmas, there are altogether 136 meanings and 899 synonyms - an average of 1.3 meanings per lemma, and of 6.6 synonyms per meaning. In Word 11, the lemmas found are reduced by a third to 66. Among those excluded, there are whole pairs like Madam and Sir, womanlike and manlike, chairwoman and chairman, witch and wizard, and some terms referring to professional and economic roles like artisan, barber, housekeeper, housewife, salesman, statesman, typist. As well as lemmas, the reduction affects meanings and synonyms. For the 66 lemmas in Word 11, there are altogether 80 meanings and 403 synonyms - an average of 5 synonyms per meaning. Take the first three pairs in Table 4.3, for example: at first sight, the elimination of some of the most arbitrary and unbalanced synonyms seems to allow for the rephrasing of Stalin's remark 'no man, no problem' as 'no item, no problem'. Qualitatively, however, this solution by elimination does not solve all problems. In all fairness to Microsoft, let us start with the first grr/ - boy pair in Table 4.3, where the elimination in Word 11 of the more marked items, together with some other adjustments, redresses the female-male imbalance completely. In the second female- wafepair, however, it does so only partially,
Table 4.2 Quantitative comparison between Word 6 and Word 11
Word 6 Word 11
Lemmas searched
Lemmas found
Meanings
Synonyms Average mean./lem.
Average syn./mean.
104 104
100 66
136 80
1.3 1.2
6.6 5.0
899 403
72
GENDER, LANGUAGE AND NEW LITERACY
Table 4.3 Qualitative comparison between Word 6 and Word 11 Word 6
Word 11
Word 6
Word 11 Word 6
Word 11
Word 6
Word 11 Word 6
Word 11 Word 6
Word 11
girt female child (female child, child, lass, schoolgirl, young woman, maiden, junior miss, demoiselle, filly); sweetheart boy. male child (male child, youth, lad, schoolboy, kid, little guy, young fellow, youngster, junior) girl young woman (n.) (young woman, lass, schoolgirl, daughter, youngster, child, teenager) boy. young man (n.) (young man, lad, schoolboy, son, youngster, child, teenager) female, feminine (feminine, delicate, womanly, effeminate, womanlike. ladylike, gentle, soft, unmanly); woman (woman, lady, girl, dame, matron, madam) male, manly (manly, masculine, virile, macho, staminate, paternal) female, feminin e (adj. ) (feminine, womanly) male, gentleman (n.) (gentleman, guy, chap) lady, woman (woman, female person, adult, matron, dame); woman of good breeding; high-born lady lord: divinity (divinity, god, Jehovah, messiah, savior, Christ, the supreme being); ruler (ruler, master, governor, sovereign, monarch, king, prince); nobleman (nobleman, aristocrat, count, don, peer, hidalgo, seigneur, magnate) lady, woman (n.) (woman, female; lord: noble (n.) (noble, aristocrat, lady, peer of the realm, member of the aristocracy) mother, parent (ma, matriarch, mama, mammy, mom, mommy, mum, mummy) father, dad ; originator (originator, architect, promulgates author, sponsor, creator, publisher, founder, generator); priest; beget mother, look after (v.) (look after, care for, protect, nurse, tend) father, priest (n.) (priest, minister, vicar, member of the clergy) nurse, attendant (attendant, medic, R.N., male nurse, practical nurse, private nurse, registered nurse, floor nurse, night nurse); nursemaid (nursemaid, servant, caretaker, attendant, babysitter, nanny, minder, nurseryman) nurse: care for (v.) (care for, look after, take care of, tend, foster, nurture); harbour (v.) (harbour, cherish, nurture, have) secretary, typist (typist, stenographer, recorder, annalist, word processor, clerk, copyist, administrative assistant, file clerk); cabinet member (cabinet member, government minister, cabinet officer, head of a department, administrator, department chief, secretary of state, secretary of the Navy, Pentagon brass) secretary, desk (n.) (desk, escritoire)
ENGLISH BETTONI
73
and the result remains messy in its asymmetrical treatment of the two lemmas, both lexically and grammatically. In the third lady - lord pair, in all fairness to women, elimination does little for them: while Word 11 now deprives lord of its divine and ruling meanings, it does not raise the lady to the noble and aristocratic realm to which lord seems to belong naturally; here Word 6 did at least mention that lady could mean a woman who is high-born or of good breeding. And what shall we make of the changes undergone by mother, nurse and secretary, as also shown in Table 4.3? In Ward 6, mother and nurse were onlynouns, now they are only verbs. This grammatical change has, in a fashion, redressed their previous imbalance, which for mother y^s striking compared to father (which had three meanings as a noun, one as a verb), and for nurse had to cope with the problem of epicene, female- and male-referring synonvms. But can we call this redress of gender imbalance an improvement for the Thesaurus generally? It would seem not, if we compare mother and father within Word 11, arid find that while the former is only a verb, the latter is only a noun with the sole meaning of priest. Finally, in Word 11, a secretary is now only a desk, and has thus lost both the low-serving and high-commanding meanings that this profession had in Word 6. On the other hand, however erratically, Word 11 also introduces some improvement. Now, for example, spinster and bachelor are perfectly symmetrical as single, unattached, free, as well as respectively bachelor and spinster. Furthermore, however erratically, it can even favour women over men. Now, for example, as a synonym for spouse, wife is listed before husband, and as a lemma it has one more synonym than husband. Summing up, compared to the previous version, the latest version of the Word Thesaurus has in some cases introduced changes arbitrarily, without improving its worse features. In other cases, it has partly redressed the negligent and offensive treatment of women. However, it has done so more by eliminating lemmas, meanings and synonvms than by fixing the problems. What remains is a poorer tool, perhaps a more neutral one in terms of gender bias, but still an unreliable one, which can confuse its users. 6.3 Comparing English varieties in Word 11 Two points should be mentioned straight away. First, all our 66 lemmas found in Word 11 under US English are also listed under UK and AUS English. Hence, there is no variation among the three Englishes in terms of our lemmas. This is not so in other semantic fields: for example, petrolis found in UK but not in US English. Second, there is also no variation between UK English and AUS English, quantitatively or qualitatively, in terms of meanings or synonyms - at least relating to our 66 lemmas. Hence, our analysis deals here only with the US and UK versions. Quantitatively, of the 66 lemmas considered, 47 are identical in US and UK English. Among the 19 that differ, 10 do so by changing meanings, and consequently also synonyms; 6 by adding items in UK English (a new
74
GENDER, LANGUAGE AND NEW LITERACY
meaning in one case, new synonyms in the other 5 cases); for one lemma a synonym is missing, and for the remaining 2 lemmas the difference lies in the order in which their synonyms are listed. Qualitatively, the differences predominandy affect lemmas referring to public roles (chairperson, chief, entrepreneur, financier, legislator, merchant, pilot, professor, secretary) rather than private ones, and more lemmas referring to men (gentleman, husband, male, man, son) rather than women. Table 4.4 illustrates some of these changes. Some changes are improvements. For example, the different meaning and synonyms for son in UK English capture the more prototypical use of the word and give it full symmetry with daughter (which has the meaning of offspring also in the US version). And wife and husband gain conventional respectability by acquiring the prototypical meaning of spouse, and losing that of companion (which, however, remains as a synonym). Other changes are for die worse. For example, diere seems to be no justification in changing the sole meaning of gentleman from man in die USA to bloke in the UK; in either case, the meaning and synonyms remove any sense of nobility, in striking contrast to Word 6, which granted it fully, offering the sole meaning of sir. Most changes, however, seem irrelevant. Randomness does not turn
Table 4.4 Qualitative comparison between US English and UK English son
wife
husband
gentleman
US: young man (n.) (young man, lad, schoolboy, youngster, child, teenager) UK: offsprin g (n.) (offspring, descendant, spawn, daughter) US: companio n (n.) (companion, consort, husband, spouse, partner) UK: spous e (n.) (spouse, companion, husband, mate, significant other, other half) US: companio n (n.) (feminine, womanly) UK: spous e (n.) (spouse, companion, wife, mate, significant other, other half) US: man (n.) (man, guy, chap, male) UK blok e (n.) (bloke, guy, lad, fellow, person)
chief
US: leade r (n.) (leader, ruler, head, person in command, boss, person in charge, captain) UK: leade r (n.) (leader, ruler, head, person in command, boss, captain, person in charge)
financier
US: banke r (n.) (banker, investor, investment banker, backer, merchant banker, sponsor, supporter) UK banke r (n.) (banker, investor, backer, sponsor, supporter, investment banker, merchant banker)
female
US: feminin e (adj.) (feminine, womanly) UK lad y (n.) (lady, woman)
male
US: gentlema n (n.) (gentleman, guy, chap) UK masculin e (adj.) (masculine, manly, mannish)
ENGLISH BETTONI
75
into method in the different ordering of the synonyms for chief in its meaning as leader or for financier. And the symmetrical swop of grammatical category in the meanings of female and male leaves unchanged in UK English the asymmetry already present in US English. Irrelevance and arbitrariness seem the main characteristics also of most additions. With regard to meanings, the lemma pretty gains rather as an adjective (sic!) besides attractive. With regard to synonyms, secretary, in the sole meaning of desk, adds bureau to escritoire; and companion adds fellow traveller to accompanying person, travel companion, escort, attendant. Looking now at the last 7 lemmas among the 19 that are treated differently, we may discern some method that could be seen as distinguishing UK English from US English. Two of these 7 lemmas refer to men, 5 to professions. In the first case, the addition of bloke and fetching as synonyms respectively for man and handsome may seem to favour some British flavour. In the latter case, the differences lie in (1) the meanings of the lemmas entrepreneur, legislator and merchant, which change respectively from US capitalist (n.) to UK tycoon (n.), from representative (n.) to politician (n.), and from mercantile (adj.) to seller (n.); (2) the addition in UK English of two synonyms, fellow and don, for the lemma professor, and (3) the deletion of C.E.O. from the synonym list for the lemma chairperson. Here one could perhaps intuit, for the changes in (1) and (3), a more general meaning in British English, less tied to the corporate business world; on the other hand, for the changes in (2), a richer specification within the academic world. Summing up, first, our lemmas are all treated identically in the two British and Australian versions of the Word 11 Thesaurus. Second, the differences in their treatment between the American version and British/Australian versions are few. In any case, third, most of them are trivial or random, or both. Only a handful can, perhaps, be seen as meaningful. Hence, how can nothing at all on the one hand, and on the other the addition of fetching bloke plus the deletion of mercantile C.E.O. for British English and Australian English justify three separate versions of the Thesaurus:' 7 Conclusio n The status of English as an international language and the dominant position of Microsoft as a producer of software can guarantee the English version of Word Thesaurus a widespread use among a great variety of readers and writers, both native speakers and learners, both women and men. For all of them, the Thesaurus is, above all, an unreliable tool. Is it really appropriate to use the term 'sexist' to describe a word-processing instrument which can assign to secretary die sole meaning of desk, forget to mention a female referent among the synonyms for millionaire, treat mother only as a verb, and father only as a noun with the sole meaning of priest"? Rather than being 'sexist', it is simply a badly conceived instrument. This is confirmed in several other ways. For example, our Thesaurus lists its synonyms without any discernible criterion, semantic, stylistic or alphabetical; it presents itself in three different
76
GENDER, LANGUAGE AND NEW LITERACY
varieties of English, and then fails to distinguish in any way between two of them, and distinguishes mostly in a trivial or random way between the other two. Or, even more tellingly, moving away from our semantic field, it forgets the botanical nature of a 'tree' and gives this lemma the sole meaning of hierarchy, includes petrol in the US version and omits it in the UK and AUS versions, while including gobbledygook in all three. Next, as well as a poor tool in general, Microsoft Thesaurus is indeed also sexist, even if to a lesser extent now in Word 11 than in Word 6. On the good side, the newer version has systematically degendered all nouns referring to occupations and titles that have not been eliminated, and in several cases has also balanced the two lemmas of a pair (e.g., girl- boy). On the bad side, many asymmetries still apply to the treatment of women and men (e.g., lord can also be an aristocrat, but lady is only a woman). Yet, even in these asymmetries, it is unreliability rather than sexism that seems to lighten the heavier imbalance in Word 6 (e.g., now wife can precede husband in a list of synonyms), to the extent that randomness can now attribute to fatherly, but not to motherly, the extra synonym concerned, which conventionally could be expected to characterize more women than men. (Or is there intent, and has the Thesaurus in Word 11 become even more sexist by generously making men now both traditionally aristocratic and modernly concerned'?) In brief, as they stand now, the American, British and Australian versions of Word 11 Thesaurus are unworthy of English as a model for both international gender reform and excellence in lexicographical description. They are also unworthy of Microsoft as a standard setter for quality products. How could matters be improved? The addition of the Encaria World English Dictionary available on the screen immediately above all three versions of the English Thesaurus is particularly felicitous, because it provides a good quality dictionary. This is the direction in which Microsoft should go, that is, replacing the present three versions with one of the many good quality thesauri of World English available on the market (or indeed more than one making them American, British and Australian - if no storage restrictions apply). This choice is not without its pitfalls. First, even for World English, lexicographers tend to select a primary variety, which is still ethnocentrically British or American; Microsoft, however, as an American company, would be justified in choosing American English. Second, even quality thesauri are sexist, as we saw in §4; Microsoft, however, as a software company, would be justified in shifting the blame to professional lexicographers. Short of replacing the whole Thesaurus, I would suggest - as minimal changes - that (1) there be only one English version, with the main regional alternatives labelled as such; (2) the lemmas and meanings be selected according to their highest frequency of occurrence in language use; (3) the listing of synonvms be done alphabetically; and, finally, (4) the whole work be subjected to strict conformitv to some quality guidelines to nondiscriminatory language issued by a well-known international body, or at least to the recommendations found in entries such as 'gender' and 'nonsexist language' suggested by The Cambridge Guide to English Usage (Peters 2004).
ENGLISH BETTONI
77
If the information supplied alphabetically would then be impossible to interpret for anyone who is not already familiar with regional and stylistic differences, compensation could be found technically in the availability of the Encarta Dictionary, and ideologically in the knowledge that at least some criteria have been systematically followed. More specifically with regard to gender issues, all languages bear witness to the presence of unequal power distribution; it would be unreasonable for a thesaurus to ignore it. On the other hand, the largest speech communities which use English as a native language are among those that have tried hardest to reduce inequalities; it would be irresponsible for a thesaurus to ignore the fact that a considerable number of their members object to unequal power distribution and consider discriminatory language usage along gender (as well as other) lines to be offensive. Microsoft is not responsible for sexism, but it is certainly responsible for failing to show awareness and providing available alternatives. Acknowledgements I would like to thank Pam Peters for useful information on lexicographical matters. Notes 1 See. e.g., the COBUILD ('Collins Birmingham University International Language Database') project; for the resulting books and electronic resources, see . 2 As the use of English spreads in countries like India and Singapore, for increasing sectors of their population English is changing from a second to a native language. 3 Peters (2003: 36-7), for example, suggests that second-language users of English, having less of a stake in any of the native-speaker varieties, could be in a better position to weigh the merits of more regular linguistic choices, and thus favour convergence in forging an international standard. References Ammon, Ulrich (1989), Towards a descriptive framework for the status/function/ social position of a language within a country', in Ulrich Ammon (ed.), Status and Function of Languages and Language Varieties. Berlin: de Gruyter, pp. 21-106. Bazzanella, Carla; Gallardo, Catherine; Guil, Pura; Manera, Manuela & Tejada, Paloma (2000), 'Categorizzazioni del femminile e del maschile nelle nuove tecnologie: prime ricerche nel Thesaurus italiano, spagnolo, francese, inglese di Word'. Cuadernos defilologia italiana, 7, 193-245. Benson, Morton; Benson, Evelyn & Ilson, Robert (1986), Lexicographical Description of English. Amsterdam and Philadelphia: Benjamins. Benson, Phil (2001), Ethnocentrism and the English Dictionary. London: Routledge. Clyne, Michael (ed.) (1992), PluricentricLanguages. Berlin: de Gruyter. Crystal, David (1997), English as a Global Language. Cambridge: Cambridge University Press.
78
GENDER, LANGUAGE AND NEW LITERACY
Gorlach, Manfred (1995), More Englishes: New Studies in Varieties of English 1988-1994. Amsterdam and Philadelphia: Benjamins. Hansen, Klaus (1997), 'British English and International English', in Edgar W. Schneider (ed.), Englishes around the World, Vol. 1 General Studies, British Isles, North America, Studies in Honour of Manfred Gorlach. Amsterdam and Philadelphia: Benjamins, pp. 59-70. Hellinger, Marlis (2001), 'English: Gender in a Global Language', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 105-13. Holmes, Janet (2001), 'A corpus-based view of gender in New Zealand English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 115-36. Johnson, Fern L. (2000), Speaking Culturally: Language Diversity in the United States. Thousand Oaks, CA: Sage Publications. Kloss, Heinz (1978), Die Entwicklung neuer germanischer Kultursprachen seit 1800. Dusseldorf: Padagogischer Verlag Schwann. McArthur, Tom (1998), The English Languages. Cambridge: Cambridge University Press. McArthur, Tom (2002), Oxford Guide to World English. Cambridge: Cambridge University Press. Moore, Bruce (ed.) (2000), Who's Centric Now? The Present State of Post-Colonial Englishes. Melbourne: Cambridge University Press. Pauwels, Anne (2001), 'Spreading the feminist word. The case of the new courtesy title Ms in Australian English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 137-51. Peters, Pam (2003), 'What is international English?', in Pamela H. Peters (ed.). From local to Global English: Proceedings of Style Council 2001/2. Sydney: Dictionary Research Centre, Macquarie University, pp. 33-9. Peters, Pam (2004), The Cambridge Guide to English Usage. Cambridge: Cambridge University Press. Poynton, Gate (1985), Language and Gender: Making the Difference. Melbourne: Deakin University. Quirk, Randolph; Greenbaum, Sidney; Leech, Geoffrey & Svartvik, Jan (eds) (1972), A Grammar of Contemporary English. Harlow: Longman. Romaine, Suzanne (2001), 'A corpus-based view of gender in British and American English', in Marlis Hellinger and Hadumod BuBmann (eds), Gender across Languages. The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 153-75. Strang, Barbara M. H. (1970), A History of English. London: Methuen. Tardif, Richard (1986), 'Preface', in Richard Tardif (ed.), The Pocket Macquarie Thesaurus. Milton, Qld: Jacaranda Press, pp. vii-ix. Trudgill, Peter & Hannah, Jean (2002), International English: A Guide to the Varieties of Standard English. 4th edn. London: Arnold.
5 The German Word Thesaurus and sociocultural models Eva-Maria Thune and Simona Leonardi
1 Grammatical , lexical and social gender Before examining the categorization of female and male terms in Word's German Thesaurus, it would be useful to oudine the ways in which gender can be marked in German. First of all, it should be remembered (see Introduction) that three levels can be distinguished: (1) grammatical gender; (2) lexical gender; and (3) social gender. 'Grammatical gender' is a formal category, but it does not refer to the biological difference between 'man' and 'woman'; rather, it is an integral part of the linguistic system, and also includes inanimate entities. All nouns in German belong to one - and only one - of three classes of grammatical gender: masculine (der Mann 'the man', but also der Baum 'the tree'), feminine (die Frau 'the woman', but also dieFlasche 'the bottle') and neuter (das Kind 'the child', but also das Glas 'glass').' 'Lexical gender', on the other hand, regards [+ female] or [+ male] semantic features, and is often inherent to lexemes denoting animate, biologically sexed entities, also perceived as such in cultural and historical terms (Table 5.1). In these lexemes there is normally an observable correspondence between grammatical and lexical gender, exceptions include derived words formed from the diminutive suffixes -chen and -kin, which are morphologically neuter derivational suffixes, thereby producing a derived word that is
Table 5.1 Lexical gender in German Feminine/female-specific
Masculine/male-specific
/rau'woman' Tochter 'daughter' Mutter" mother' Schwester 'sister' Abnntf'nun' Kuh 'cow'
Mann 'man' Sohn 'son' Vater 'father' Bruder 'brother' Monch 'monk' S&r'buir
80
GENDER, LANGUAGE AND NEW LITERACY
grammatically neuter irrespective of the gender of the root word. For this reason Madchen 'girl' (originally a diminutive of the grammatically feminine Magd, 'maiden', 'female servant') is morphologically neuter, even though it refers semantically to a female. Likewise, Mdnnchen 'little man' and also 'male animal', which derives from Mann 'man', is morphologically neuter despite being semantically male. In each class of grammatical gender there are lexemes which do not specify the sex, and so they can refer to both women and men (Table 5.2). 'Social gender' is a non-linguistic category that reflects social and cultural stereotypes regarding the roles occupied by men and women (see Samel 1995). As regards words with a generic reference, there is a prevalence of masculine nouns denoting professions and social roles in the upper tiers of hierarchically structured contexts, while the subordinate roles are often represented by feminine nouns. Examples of this include: Arzte(m.) undKrankenschwestern (f.) —> doctors (m.) and nurses (f.) DerChef(m.) und seine Sekretdrin (f.) —> the boss (m.) and his secretary (f.) 1.1 Gender in German
Reference to a female in German can be specified in a number of ways (see Klann-Delius 2005). As regards agentives, the most morphologically fruitful process is the so-called Mavierungor Motion, namely the use of the suffix -in to form a feminine noun from a male one: (1)
derLehrerthe (male) teacher—* die l^hrerin the (female) teacher
Other suffixes, of French (-ice; -ette) or English (-ess) origin, also perform the same function, but they are no longer particularly productive. In the case of substantivized adjectives, especially participles, the reference to gender is only clear in the singular, when the noun is preceded by an indefinite article or is not preceded by determiners: (2a) (2b)
eine (f.) Angestellte (f.) a (female) clerk ein (m./nt.) Angestettter (m.) a (male) clerk
On the other hand, when the noun is preceded by a definite article, this is the only gender-marking element:
Table 5.2 Non-gender-specific terms in German Feminine
Masculine
Neuter
die Person 'the person' derMensch 'the human being' das Individuum 'the individual' das Haupl 'the head' die l,eiche 'the corpse' derSdugling'the infant'
GERMAN THUNE AND LEONARDI
(3a) (3b)
81
die (f.) Angestettteihe (female) clerk der (m.) Angestettte the (male) clerk
As German has no grammatical gender differentiation in the plural, the plural forms of substantivized adjectives are also lexically non-genderspecific. Another way of specifying gender is through the formation of compound words which have as their head lexemes marked for lexical gender. Traditionally there were many compounds, especially for denoting agentives, with -mann '-man' and -terr'sir', 'mister', 'master', e.g. FeiMnvehrmann'(ma\e) firefighter'; Rotsherr '(male) member of the city council'. Traditionally less common, and initially limited to supposedly less prestigious, stereotypically female professions, are compounds with -frau, e.g. Putzfrau 'cleaning woman'. One of the most evident results of the linguistic policy aimed at ascribing greater linguistic visibility to women is the increase in the number of compounds with -frau for professions or positions that are stereotypically male, starting with compounds including -mann or -herr, for instance Feuerwehrfrau' (female) fire-fighter' has been formed from Feuerwehrmann and Ratsfrau' (female) member of the city council' from Ratsherr. A further strategy is to neutralize gender, in other words, to use genuinely non-specific forms for generic reference rather than the generic masculine. This can be achieved in the first place by using as the head of the compound a term that lexically has non-gender-specific reference, for instance -kraft, as in Lehrkraft 'member of the teaching staff, where this is understood to include Lehrerinnen und Lehrer 'female and male teachers'. Another possibility is to use plural forms, where, as we have said, there is no gender reference, choosing non-marked lexemes. Examples of this are compounds with -leute 'people', for instance, Landsleute 'fellow citizens' (which includes the plural of both Landsmann and Landsfrau), or substantivized participle forms, such as die Studierenden, literally 'the studying (people)' which comprises both die Studenten 'the (male) students' and die Studentinnen 'the (female) students' (see below). Discussion of gender-specific reference has had a considerable influence in the debate about language use in the political and public arena in the German-speaking world (see Bufimann & Hellinger 2003; Hellinger 2004; Cyrus-Eichhoff 2004). In 1980, the first guidelines for reducing gender bias in language use ('Richtlinien zur Vermeidung sexistischen Sprachgebrauchs', see Guentherodt et al.) were published and several others have followed since then in Germany (for example, Hellinger & Bierbach 1993; BBB 2002; BMBF 2003). Discussion of these matters in Austria and Switzerland has been no less fruitful (for Switzerland, see Schweizerische Bundeskanzlei 1996; Schiedt & Kamber 2004; for Austria, see Kargl et al. 1997; 'kurz & bundig' 2000), as indeed it has been in the whole EU area (see EU 2003). The results of this discussion are visible in the conception of reference tools like the Duden, a major normative dictionary of the German language.
82
GENDER, LANGUAGE AND NEW LITERACY
Former editions of the Duden dictionary had been criticized for a lack of feminine lemmata and for using stereotypical examples (see Pusch 1984). In the last editions (see Duden Bedeutungsworterbuch 2002; Duden Deutsches Universalwarterbuch 2003) changes were introduced in three areas (see KunkelRazum 2004: 309 ff.): (1) the number of female terms is now considerably higher than before; (2) the number of examples in which women appear in a non-stereotypical way is considerably higher; and (3) a special chapter has been introduced in volume 9 (Duden Richtiges und gates Deutsch 2001), a son of style manual, which gives concrete examples of the various possibilities which German offers for achieving a balanced reference to gender. A similar although less extensive attempt is to be found in the latest editions of certain foreign-learner dictionaries (e.g. Duden Worterbuch Deutsch als Fremdsprache 2003; see Kunkel-Razum 2004: 311).2 2 Crucia l pairs The analysis was conducted on the thesaurus of Microsoft Word for Mac X.s The German Thesaurus is organized so that for every entry a note in brackets specifies whether the word is a noun ('Subst.'), an adjective ('Adj.'), a verb ('Verb'), or another part of the speech ('Sonst.'). In our examples, the meanings are in bold, whereas the items given for each meaning stand in brackets, in italics. The first crucial pair, 'woman' - 'man', is as follows: (4a)
(4b)
Frau: Ehefra u (Subst.) (Ehefrau, Gemahlin, Gattin, Ehegespons); Dam e (Subst.) (Dame, Evastochter, Frauenzimmer, Weib, Madame, Weibsbild, Herrin, Gebieterin, Matrone, Krone der Schopfung, Hausfrau, Hauswirtin, EdelJrau, Bessere Hdlfte, Das schone Geschlecht, Frau und Gebieterin). Mann: Her r (Subst.) (Hen, Gentleman, Monsieur, Grandseigneur, Mannvolk, Mannsbild, Junggeselle); Eheman n (Subst.) (Ehemann, Gatte, Gemahl, Vermdhlte, Gefdhrte, Brdutigam)', Effendi; Antonyms. .. .
What can be noted above all is the different sequentiality of the two meanings; the first term listed for Frau is the word for 'spouse' (Ehefrau), while for Mann this comes second (Ehemann). This suggests the enduring influence of social stereotypes, namely that die primary semantic sphere associated with 'woman' is the family and above all that of being someone's wife. This is accentuated by the fact that in the list of 'meanings' for Ehefrau there is a total absence of more 'neutral' terms, both those that are now preferred in the administrative register (f. Ehepartnerin - m. Ehepartner) and those denoting cohabitation (f. Lebensgefdhrtin - m. I^ebensgefdhrte, f. Partnerin - m. Partner) . In the case of Ehemann, by contrast, we can at least find among the synonyms the form Gefdhrte, which means generically 'mate' or 'companion', but can also have the meaning 'partner'. The first term given for Mann is Hen, which in Gennan is the non-marked form of address for a male individual (the female form is Frau, see below); its first synonyms are English and
GERMAN THUNE AND LEONARDI
83
French words, probably understood as terms of address (at least in the case of Monsieur), while Mannvolk denotes the man as part of a group of equals (men), and Mannsbild and Junggesette convey the idea of the independent (male) individual. However, if we look at the synonyms for Dame, we find that there are no corresponding forms at all for the last three lexemes suggested as synonyms of Herr, given that the notion of 'woman' as 'individual' is not considered at all. On the other hand, there are numerous references (Evastochter 'daughter of Eve', Krone der Schdpfung1 crowning of creation', das schone Geschlecht 'the fair sex') to terms and idiomatic expressions based on a corporeal femininity. This referencing of metaphoric terms and expressions is another characteristic feature of the German Thesaurus, and, as we will see, is particularly evident for female terms. Finally, the choice of the Turkish word Effendi as a meaning of Mann, especially as it is completely decontextualized, is puzzling to say the least. The word Frau is presented as the opposite of Mann but also of Herr, when the latter is used as a term of address. When Herr has the meaning of 'gentleman', on the other hand, the female equivalent is Dame. (5a)
(5b)
Dame. Madam e (Subst.) (Madame, Dame, Weib, Frau); Kron e de r Schopfung (Subst.) (Krone der Schdpfung, Dame, Weib, Frau, Frauenzimmer); Fra u (Subst.) (Frau, Dame, Evastochter, Frauenzimmer, Weib, Madame, Weibsbild, Herrin, Gebieterin, Matrone, Krone der Schdpfung, Hausfrau, Hauswirtin, Edeljrau, Bessere Hdlfte, Das schone Geschlecht, Frau und Gebieterin; Antonyms ...); Fee (Subst.) ...; Brettspiel (Subst.) (other meaning) Herr. Weltman n (Subst.) (Weltmann, Herr, Kavalier, Ehrenmann); Be sitzer (Subst.) (Besitzer, Disponent ...); Regen t (Subst.) (Regent, Herrscher, Fuhrer .. .); Potenta t (Subst.) (Potentat, Machthaber .. .); Mann vo n Wel t (Subst.) (Mann von Welt, Kavalier, Gentleman, Ehrenmann, Herr, Weltmann); Man n (Subst.) see above; Kavalie r (Subst.) (Kavalier, Weltmann, Gesellschaftslowe ...); Herr uber... (Subst.) (Herr uber, Besitzer, Eigentumer, Eigner, Halter, Herr, Mieter, Nutzniefier, Pdchter); Herr im Hause (Subst.) (Herr im Haus, Hausherr, Besitzer, Eigentumer, Meister, Gebieter, Herr, Grundherr, Antonyms ...); Haupt (Subst.) (Chef, Hauptmann, Fuhrer, Unternehmer, Anfuhrer, VerwaUer, Bandenjuhrer, Verweser, Befehlshaber, Hduptling, Statthalter, I^eithammel, Radelsfuhrer. . .); Halter (Subst.) . . . ; Grandseigneu r (Subst.) . . . ; Gottvate r (Subst.) ...; Gottheit (Subst.) ...; Gott (Subst.) ...; Gewalthaber (Subst.) ...; Gentleman (Subst.) ...; Eigentumer (Subst.) ...; Effendi (Subst.) ...; Dienstherr (Subst.) . . . ; Besitzer (Subst.) . . . ; Befehlshaber (Subst.) ...;AUvater (Subst.) ....
There is a striking lack of proportion between the two terms; there are just four 'meanings' for Dame and as many as 23 for Herr, clearly revealing that no attempt has been made to establish parallelisms between the two lexemes. Moreover, the two groups of words contain lexemes that are exclusive to one
84
GENDER, LANGUAGE AND NEW LITERACY
or the other group; it is worth noting that in the female group the second meaning given is Krone der Schopfung and the fourth is Fee 'fairy' ('beautiful woman'), neither of which have a male equivalent. Of the 'meanings' listed for Herr, only Mann has an equivalent in one of the suggested meanings for Dame (Frau), while the others offer a vast range of denotations: the public role of men in society (Weltmann, Mann von Welt, Kavalier, Grandseigneur, Gentleman); stress on the exercising of authority (Regent, Potentat, Gottvater, Gottheit, AUvater), including divine authority; and power (Gewalthaber, Befehlshaber, Dienstherr), including economic power (Besitzer). Mann is one of the few male terms accompanied by collocations (where 'collocation' is understood in the broad sense of the term, namely the lexical structures in which the headword combines with other words, see Corda 1998:196): Mann von Welt 'man of the world', Herruber.. . 'man in charge of . ..', Herr im Haus 'master of the house'. However, these are very different from the collocations for Frau mentioned above, because they have much more concrete contexts and are used frequently, while the ones for Frau are figurative and have a restricted usage. It is also worth noting that the list of the 'meanings' for Hen includes the term Haupt (see above); this is in fact attributable to both genders, but is not included under Frau or Dame. The items for Haupt listed under Herr are exclusively male. Until the 1980s, the word Frdulein was used as a term of address; a grammatical derivation of Frau formed by adding the neuter diminutive suffix, it is the semantic equivalent of the English Miss and therefore reserved for an unmarried woman. Following debate about the asymmetry in the terms of reference for women and men in German (there was no equivalent term to denote an unmarried man), the title was officially abolished in the Germanspeaking countries, and its use in the spoken language is gradually diminishing. The entry for Frdulein in the Thesaurus is as follows: (6)
Frdulein: Bedienung (Subst.) (Bedienung, Demoiselle, Mddchen, Jungfrau, Jungfer, Maid, Mamsell, Mifi, Weibsbild, Weib, Braut, Blaustrumpf, Kinderfrdulein, Gouvernante, Lehrerin, Junggesellin; Antonym: Frau);...
Its use as a term of address is not considered at all; the sole 'meaning' listed is that of Bedienung, namely 'servant', which has feminine grammatical gender but non-specific lexical gender. The nearest approach to the meaning of Frdulein as term of address is provided by Mamsell and Mifl, two now obsolete terms of French and English origin, which can only be used in German to refer to a female servant. Some of the other 'synonyms' focus on the fact that a Frdulein is not yet a woman (Mddchen, Jungfrau, Jungfer, Maid) (here, as in many other cases, terms belonging to very different registers are all placed on the same level; while Mddchen and Jungfrau, at least in certain meanings, form part of standard modern German, Jungfer now sounds old-fashioned and Maid is either old-fashioned or dialectal). Then there are a number of care-giving roles prototypically associated with females (Kinderfrdulein, Gouvernante, Lehrerin),
GERMAN THUNE AND LEONARD!
85
plus two terms denoting 'unmarried woman', neither of which have positive connotations, Blanstrumpj'(the exact equivalent of blue-stocking) andjunggesellin, deriving from the male termJunggeseUe 'bachelor', but with a much more restricted use. Finally, Weib and the compound word Weibsbild (literally 'picture of a spouse', it actually means 'hag') are in fact two terms denoting 'woman', both of which are derogatory in current usage. The connection with Frdulein is therefore unclear. Turning now to the word Madchen 'girl', this is formed in a similar way to Frdulein, with a diminutive suffix that makes the term grammatically neuter: (7)
Madchen: Fraulei n (Subst.) (Frdulein, Besen, Kind, Bluse* Biene, Fratz, Jungfrau, Ricke, Kdfer, Maid, Pfldnzchen, Puppchen, Antonyms ...).
What is surprising is that the only 'meaning' given for the word is Frdulein, which, as we have said, is being used less and less, and in any case has never referred to childhood, while Madchen covers both childhood and adolescence. The first 'synonym' listed, Besen, is metaphorical (strictly speaking, it means 'broom') and only in one regional variety (Swiss) does it also mean 'girl'. Kind 'child' may be an acceptable way for referring to young people still in their childhood, but it eliminates reference to the female sex. Jungfrau is archaic except for its meaning of 'virgin', and so placing the two terms alongside each other without any mediation simply leads to confusion. Finally, the terms Biene (literally, 'bee'), Fratz 'funnv face', for a human being, Ricke 'doe', Kdfer 'beetle', Pfldnzchen 'little plant', Puppchen 'little doll' are not so much synonyms as metaphorically based diminutive/pet name forms. The male equivalent of Madchen isjunge, which, strictly speaking, is a substantivized adjective, lexicalized, however, as a noun, in this case malespecific: (8)
Junge. Bu b (Subst.) (Bub, Bengel, Bube, Bursche, Jungling, Kerl, Kind, Knabe, Sohri); Flegelhaft e (Adj.) (Flegelhafte, Halbwuchsige); Bluhend e (Adj.) (Bluhende, Frische, Dumme, Grime, Unreife); Blutjung e (Blutjunge, Jugendjrische, Jugendliche, Jungenhafte, Kindische, Kindliche, Knabenhafte, Mddchenhafte, Antonyms ...); Werfe (Verb) (Werfe, Entbinde).
Suggested 'meanings' include a noun (Bub), adjectives (Bluhende'blooming' and Blutjunge 'very young') and a verb (werfe); these morphological differences are stated only in the synonym column (through the abbreviations Subst, Adj. or Verb) but not in the space where the term that is object of the search is inserted. The synonym column thus contains three different parts of speech and whereas the difference between Junge as a noun and junge as an adjective may not be particularly striking or problematic, the other 'meaning' given - werfe (a verb), is the result of interpreting/Mwgeas part of the verb jungen 'to give birth' (used for animals), and it would hardly be surprising if
86
GENDER, LANGUAGE AND NEW LITERACY
this left the user considerably puzzled. Thus, while the Thesaurus does not constrain the user to specify at the beginning of the search the part of speech of the inserted term, it sometimes ends up providing synonyms and meanings belonging to completely different grammatical categories, leaving the user alone to decide which one should correspond to the one searched for. Strictly speaking, the only 'real' equivalent of Junge (noun m.) is Bub; however, this term, which quite apart from anything else is characteristic of regional (southern German) usage, refers to die period of childhood, while Junge extends into adolescence. There is the same lack of clarity in the synonyms, because Knabe covers childhood and early adolescence, while Bengel, Bursche and Jungling relate to adolescence, but now sound rather oldfashioned. As with Mddchen, one of the suggested synonyms is Kind 'child', a non-gender-specific term; however, it is worth noting the presence of the gender-specific Sohn 'son', while the list of 'synonyms' for Mddchen does not include Tochter 'daughter'. In German, adjectives denoting male and female qualities are derived from a base noun with the addition of the suffix -lick, there are in fact two adjectives for 'feminine', fraulich, deriving from Frau, and weiblich, which stems from what is now an obsolete term denoting a woman, Weib. While in current usage the noun Weib is only used with a derogatory meaning, the adjective is the term commonly used for indicating the feminine, also that of grammatical gender: (9a) (9b) (9c)
weiblich: femini n (Adj.) (feminin, mddchenhaft, fraulich, jungfrdulich; Antonyms...) fraulich: femini n (Adj.) (feminin, frauenhaft, weiblich, mutterlich, damenhaft, heiratsfdhig, zart, schwach; Antonyms . ..). mdnnlich: maskulin (Adj.) (maskulin, mannbar, reif, zeugungsfdhig, mannhaft, entwickeU, krdflig, tauglich, heldenhaft, muskulos, tapfer, stdmmig, charakerfest, kernhaft, mutig, beherzt, wittensstark, beharrlich, tuchtig); erwachsen (Antonyms . . .).
What is once again so striking is the wealth of synonyms for the male term and the paucity of synonyms for the female one; the listed items are entirely equivalent, because both weiblich and fraulich are interchangeable with feminin; this is also true for the first item given to mdnnlich (maskulin), but the second item (erwachsen 'adult') is significantly different. The suggested synonyms for the three lexemes amount to a compendium of gender stereotypes: the adjectives referring to women either focus on a particular stage of life (adolescence, in mddchenhaft 'girlish') or features traditionally associated with women and which emphasize the need for protection (schwach 'weak', zart 'fragile'). It is also worth noting the presence of the adjective mutterlich, which derives from Mutter 'mother' and denotes the care-giving features of the woman seen prototypically as a mother, and the absence of the corresponding male term amongst the synonyms of mdnnlich. It is also not clear why jungfrdulich 'virgin' (adj.) is listed under weiblich. By contrast, a wide
GERMAN THUNE AND LEONARDl
87
range of adjectives are listed as synonyms of mdnnlich, all of which have positive connotations; they emphasize the achievement of maturity in various fields: physical (krdftig, tauglich, tuchtig 'strong-mature', muskulos, stdmmig 'solid'), character-related (reif, tauglich, tuchtig 'mature', heldenhaft 'heroic', charakterfest 'strong in character', kernhaft 'having backbone', wwrig'courageous', beherzt 'hearty', willensstark 'strong-willed', behardich 'thoroughgoing'), and sexual (zeugungsfahig, reif, mannbar'fertile'). Another crucial pair is Mutter - Voter, 'mother' - 'father': (lOa) Mutter. Schraubenmutte r (Subst) (Schraubenmutter, Chefin); Mama (Subst.) (Mama, Mami, Mutti, Mutterlein, Mutterchen, Hausmutterchen, Gebdrerin, Kindbetteriri); Fra u (Subst.) (Frau, Ehegattin, Gattin, GemahUn, Hausfrau; Antonyms ...). (lOb) Vater. Priester (Subst) (Priester, Pfarrer, Pastor); Anverwandte (Subst.) (Anverwandte, Erndhrer, Familienoberhaupt, Antonyms ...). The first 'meaning' given for Mutter, Schraubenmutter, is in actual fact marginal: it does not denote a woman at all, but rather a 'nut' (with which to tighten a bolt); at this point the synonym Chefin 'female boss' is incomprehensible. The first effective 'meaning' given for Mutter is Mama, a pet term that in German is used exclusively in a family context, generally as a term of address. The same applies for the first two synonyms (Mami and Mutti, both 'mummy'); the terms Mutterlein and Mutterchen, however, are unlike many diminutives in that in current usage they carry no implication of youth; indeed, they denote a woman who is getting on in years ('grandmother'). The same is true of the compound word Hausmutterchen, which also underlines that the range of action is restricted to the domestic sphere (Haus-). The last two 'synonyms' (Gebdrerin 'child-bearing woman', Kindbetterin 'woman who has just given birth') are, strictly speaking, 'technical' terms in that they relate to the immediate post-delivery phase of motherhood. The synonyms listed after Frau are limited exclusively to the woman as 'spouse', almost as if this were a necessary condition for being a mother (Ehegattin, Gattin, Gemahlin and, in part, Hausfrau}. Hausfrau, which, strictly speaking, means 'housewife', denotes an activity stereotypically associated with women; given the absence of any other terms indicating activities, the inference is that Hausfrau and Mutter are equivalent. As far as Voter is concerned, it is surprising to find that the first 'meaning' listed is Priester, that is 'father' in the religious sense of the word, for which the corresponding synonyms are other religious terms, Pfarrer 'parson' and Pastor'pastor'.5 The second 'meaning' is Anverwandte (which in the indefinite masculine singular should be Anverwandter), i.e. 'relative'; this more neutral term does not feature in the entry for Mutter. With the synonyms for Anverwandte(r) no attempt is made to find a neutral expression, because there is Erndhrer 'nourisher' (m.), with a clear semantic inference to the role of the 'father' as a provider of nourishment for the family. Then there is FamiUenoberhaupt, in other words, the paterfamilias; the head of this noun is
88
GENDER, LANGUAGE AND NEW LITERACY
Haupt, which, as we saw above, can be used for both male and female reference. The fact that it is only proposed for the male term contributes to reinforce stereotypes. Perhaps even more illuminating is a comparison of the adjectives deriving from Mutter and Voter. (lla) mutterlich: fursorglich (Adj.) (fursorglich, fraulich, aufopfernd, venuandt, besorgt, gutig, hingebungsvoll, liebevoU, ruhrend, selbstlos, uneigennutzig; Antonym: vdterlich). (lib) vdterlich: fursorglich (Adj.) (fursorglich, wohlwollend, liebevoU). At first sight, it seems that an attempt has been made at symmetry, because the suggested 'meaning' is the same (fursorglich 'attentive', 'care-giving'). However, there are differences between the two genders in the 'synonvms' (or fursorglich. Curiously, there are more terms for the feminine lexeme, but one cannot fail to note that they all relate to semantically similar character features:6 affectionate solicitude (gutig, liebevoU); care-giving and protectiveness towards others (besorgt), which carries over into a spirit of self-sacrifice (hingebungsvoll, aufopfernd) and self-denial (selbstlos and uneigennutzig). In the entry for vdterlich, just two 'synonyms' are listed for fursorglich, one of which, liebevoU, is also listed for the feminine term, while wohlwollend [sic] is exclusive to vdterlich; iwhlwottend always denotes 'benevolence' towards others, but it does not concern two subjects on an equal footing, but has an 'upper-to-lower' social dimension to it and is thus asymmetrical. In view of die wealth of 'synonyms' for mutterlich, all of which have positive connotations and which reflect the qualities stereotypically associated with the woman as mother, it can be stated that in the Thesaurus the real feminine equivalent of mdnnlich is neither fraulich nor weiblich but rather mutterlich, thereby reaffirming the correspondence between 'woman' and 'mother', especially when only positive features are highlighted. It is also worth noting that the word Eltern, a plural noun denoting the parental couple, was 'not found'. 3 Socia l roles A profound lexical renewal has taken place in German in the past 30 years. This has been promoted by a linguistic policy directed towards achieving greater symmetry in terms of gender reference, increasing the visibility of the female sphere and doing away with linguistically fossilized cultural stereotypes (see Linke 2002). However, few (if any) traces of this process emerged from an examination of the thesaurus. Female forms created from a masculine noun by the suffix -in (see above) are not, generally speaking, present; for instance, there is KeUner ('waiter') but not KeUnerin ('waitress'); likewise, there is Professor'(male) professor', but Prvfessorin '(female) professor' is absent. The same applies for compounds with -mann/-frau: there is Kaufmann 'salesman' but not Kauffrau 'sales-
GERMAN THUNE AND LEONARDI
89
woman', and not even the generic plural Kaufleute. One can find the masculine noun Student, but there is neither the corresponding feminine form Studentin, nor Studierende(r), a substantivized participle, which is the most lexically generic form (and currently widely used and recommended). It is also worth emphasizing the instances where the feminine form is listed, because in fact die suggested items reveal a powerful asymmetry. One glaring example is the pair Lehrerin- Lehrer 'female teacher' - 'male teacher': (12a) Lehrerin: Erzieherin (Subst.) (Erzieherin, Kindermddchen, Lehrer, Gouvernante, Amme). (12b) Lehrer. Lehrher r (Subst.) (Lehrherr, Lehrmeister)', Lehreri n (Subst.) (Lehrerin, Lehrer); Erzieher (Subst.) (Erzieher, Pddagoge, Mentor, Pauker, Dozent, Professor, Rektor, Antonym: Schiller) ...
The terms listed under Erzieherin 'governess' (ignoring the reference to the masculine Lehrer) relate to activities concerning the very first stages of childhood, which have more to do with care-giving than with education as such (Kindermddchen 'baby-sitter',7 Gouvernante 'governess', Amme 'wet nurse'). For the male term, the first suggested synonym (Lehrherr 'apprentice master') refers in actual fact to the ambit of professional training and apprenticeship (Lehre)', this area is not represented in the feminine form at all. The term Erzieher has cross-references that relate not only to education in the strict sense of the term (Pddagoge, Mentor, Pauker), but also to nouns regarding university teaching, such as Dozent and Professor (terms which, at least in German usage, are not equivalent to Lehrer, which is a primary or secondary school teacher). Astonishingly, the word Rektor, which denotes the most senior of all university posts, is also listed. The above observations are further reinforced if one examines the terms for Lehrkraft, a compound noun which uses the non-gender-specific formative -kraft (see above). This is in consequence non-gender-specific, but actually only male-specific terms are listed: (13)
Lehrkraft Lehre r (Subst.) (Erzieher, Pauker, Schulmeister, Lehrmeister, Pddagoge).
The same asymmetry can be noted by observing the suggestions for Erzieherin - Erzieher, where the only given meaning of the female term is Kindermddchen, while for the male one it is Pddagoge. As for the 'synonyms' of the female term, not only is care-giving underlined (e.g. Kinderwdrterin, Kinderjrau, both 'nanny'), rather than teaching, but there are also lexemes that are explicitly associated with the domestic, non-public sphere (Hauslehrerin, Privatlehrerin). As far as the economic and professional hierarchy is concerned, one can observe that female-specific forms for the highest positions are very slow to establish themselves. Besides those cases where such terms are simply missing (e.g. there is Anjuhrer 'male leader' but not Anfuhrerin 'female leader'), it is worth paying close attention to pairs such as Chefin - Chef. For
90
GENDER, LANGUAGE AND NEW LITERACY
the feminine form the first meaning given is that of Mutter 'mother' (with puzzling synonyms such as Schraubenmutter 'female screw' and again Chefiri). What this implies is that the most usual domain where a woman can exercise her authority is the domestic one, and even there exclusively as a mother. Only as a second meaning do we find the term Leiterin 'chief, 'head' (with the synonyms Chefin and Direktorin). As for the masculine term Chef, only one meaning is given, that of Anfuhrer (for which, as stated above, no feminine form is to be found in the Thesaurus), but which is in turn provided with 16 synonyms, most of them relating to the highest positions in the business and administrative hierarchy. Another significant instance is Direktorin - fiirektor, where for the female term just one meaning is provided, that of Leiterin; this seemingly positive acceptation, however, quickly yields place to what is in fact an extraordinary restriction of the semantics: as a synonym of Leiterin, the system gives Direktrice, which means 'senior female employee in a fashion store'. The male form, on the other hand, also has only one 'meaning', that of Chef'(male) boss', but has numerous 'synonyms' that cover a broad range of possibilities relating to the exercise of authority, thus underpinning the interpretation to be given to Chef. A similar asymmetry can also be noted further down the scale: Angestellte/r '(female)/(male) clerk', a substantivized past participle, has various terms given for the female and the male; the first meaning listed for the female term is Kammerzofe, which besides being decidedly old-fashioned, clearly refers to a subordinate position as a '(personal) maid'. In this respect the analysis of Personal is also highly revealing; the terms listed under Hauspersonal build up a picture relating to what is now an obsolete type of employment, which simply confirms common stereotypes regarding domestic work. For instance, there are female terms like Dienstmddchen 'maid' (literally a compound of Dienst 'service' + Mddchen 'girl'), now no longer recommended given the presence of the asymmetrical use of Mddchen, Putzfrau 'cleaning woman', Hausgehilfin 'home help', Waschfrau 'washerwoman', Zimmermddchen 'chambermaid', Kindermddchen 'baby sitter', Halbtagsmadchen 'morning help' (f.), Magd 'maid', etc. The male terms, on the other hand, are words like Chauffeur '(male) driver' and Dienstbote 'male servant'. 4 Conclusio n Given that the Thesaurus purports to be a text-editing instrument, there are some serious shortcomings. First of all, there is an utter lack of clarity regarding the morphological categories, especially with regard to substantivized adjectives. For instance, Anverwandte is listed as a 'meaning' of Voter (see example lOb); however, its use as a substitute is only possible when Vateris preceded by a definite determiner (der Voter—* der Anverwandte). If, however, Vateris to be understood as indefinite, the use of Anverwandte would produce an incorrect phrase (mein Voter-» *mein Anverwandte; the correct form should be mein Anverwandter). The same holds for the lack of gender specification
GERMAN THUNE AND LEONARDI
91
both in the meanings and in the synonyms; clearly, a given term can only be used as a substitute if its gender matches that of the noun phrase as a whole; but in many cases precisely this gender information is missing and the user is effectively left on her/his own to cope on the basis of whatever grammatical notions she/he happens to have. Haupt (nt.), for example, in most cases cannot be mechanically inserted as a synonym of Herr (m.) without producing an incorrect result, e.g. Der Herr der Gemeinde 'the master of the community' -» *der Haupt der Gemeinde (the correct form is das Haupt der Gemeinde). At a semantic level, the fact that neither the 'meanings' nor the 'synonyms' are differentiated in the slightest poses serious problems when it comes to choosing the appropriate word. Indeed, it is never specified whetfier the option is 'neutral' or whether it is particularly marked in diatopic, diastratic and even diachronic terms. Maid (see example 7) is only used to refer to a young woman in southern dialectal varieties, and even here it is a marked form. Jungfrau (see example 7), in the sense of 'girl', is no longer part of current usage apart from when it is used ironically. Fratz (see example 7) implies affectionate closeness to the 'girl', and is part of a colloquial-familiar register. It can also be noted, especially for the female-specific lexemes, that the 'meanings' and 'synonyms' include metaphorical forms belonging to obsolete cultural models, now little used except ironically. The problem is accentuated by the fact that this characteristic, which is both a question of semantics and of usage, is not indicated. The inherent dangers of this type of suggestion is well illustrated in example 1: (Frau), where, listed under the 'meaning' of Dame, there are such 'synonyms' as Krone der Schopfung and Evastochter. Quite clearly, there are very few contexts in which Frau can be replaced by expressions of this kind. The Thesaurus therefore requires the utmost care on the part of the user, who has to distinguish between the various possibilities without being given any help in choosing between lexemes belonging to different registers. The use of the Thesaurus therefore presupposes a very high level of linguistic competence if the user is to avoid making glaring errors or coming up with antiquated or dialectal terms that are frankly ridiculous. The lack of symmetry between male and female forms is systematic: one area in which this can be seen is the different semantic connotations attributed to the female term in the crucial pairs examined above, which reinforce the usual stereotypes of the 'woman' as 'mother' and 'wife' (see examples 10 and 11), the person who looks after the household and others, and in general 'serves' others. 'Intellectual' (see example 12) or authority- and command-related features (see example 9) are never ascribed to the woman. Often the possibility of a female-specific version of a lexeme, above all of agentives, is not taken into consideration at all. Hadumod BuBmann & Marlis Hellinger (2003) have underlined how in the German-speaking countries there has been a marked 'feminization' of the language, which has contributed to giving greater visibility to women in society, both through the use of
92
GENDER, LANGUAGE AND NEW LITERACY
the so-called Movierung, the female-specific form produced by the suffix -in (e.g. Minister 'minister* —> Ministeriri), and through the increase in compound forms with -Jrau (Kauffrau). When the Thesaurus is searched for the female forms listed by BuBmann and Hellinger, they simply turn out not to have been included (notfound). Similarly, generic plural terms formed by the lexeme -leute 'people' (e.g. Kaufleute 'tradespeople', which includes both Kauffrauen and Kaufmdnner), are not generally found either. Finally, all too often there is an absence of substantivized participle forms (present and past), the use of which is currendv recommended, above all because they are semantically neutral in the plural and therefore especially suited to generic reference; see, for instance, the form die Studierenden 'students', which includes both Studenten 'male students' and Studentinnen 'female students'. In conclusion, there is an astonishing absence in the thesaurus of all the forms that give visibility to women, which according to the study by Hadumod BuBmann & Marlis Hellinger (2003) are now well established in German. Certainly no less surprising is the fact that a similar failure to take account of current issues regarding women's role in society is to be found in other electronic devices, such as the Google search-engine: it has been pointed out (Arndt 2005) that if one initiates a search on www.google.de using the feminine forms of (prestigious) professional and social roles stereotypically associated with men, for example Biologin '(female) biologist' or Wissenschaftlerin '(female) scientist', one is immediately confronted with the question Meinten Sie Biologe 'Did you mean Biologe?' (or Wissenschaftler), i.e. the male forms. Significantly, this is the same type of message one gets if one makes a simple spelling mistake and this obviously gives the impression that female forms are in some way less correct than the male ones. It may well be true that most biologists are men, and consequently die websites containing the term Biologe (m.) are much more numerous than those containing Biologin (f.) (ca. 285,000 sites vs 115,000). But this logic does not hold the other way round, since a search for Kindergdrtner '(male) kindergarten teacher' does not prompt the message Did you mean Kindergdrtnerin?, although the hits for Kindergdrtnerin'(female) kindergarten teacher' are definitely more numerous (ca. 99,600) than those containing the term Kindergdrtner (ca. 43,100), The masculine form thus appears always to be registered as the default form, whereas the derived feminine forms are considered proper only for stereotypically female roles. All this is all the more surprising given recent developments (see above) found in the latest dictionaries and vocabularies. The so-called new media, far from giving due weight to the most recent issues regarding the linguistic representation of women and men and their categorization, seem on the contrary to perpetuate socio-cultural models that in actual society are well on the way to becoming obsolete. This trend appears particularly unwelcome, in view of the fact that these devices are very widely used, predominantly by
GERMAN THUNE AND LEONARDI
93
people who are not familiar with the recommendations set out by the various guidelines for the avoidance of gender bias in language. On this basis, the concrete risk is that these superficial and badly conceived devices will end up seriously compromising the efforts that are currently being made to secure greater visibility for women in language. Notes 1 2
3 4 5
6
7
At present the attribution of grammatical gender is often semantically unclear (see Lazzeroni 1993). This seems particularly important to us, since socio-cultural information is inevitably conveyed by the choice of examples in dictionaries, and in the past these uniformly showed a very one-dimensional image of women in Germany (see Raffeiner&Thune 1999). This version of the Thesaurus is essentially no different from the previous versions for Mac or those for Windows, which have also been checked. The connection between Bluse ('blouse') and Mddchen is unclear. This provides further confirmation of the imprecision of the German Thesaurus, because Voter generally denotes a Catholic priest; Pastor cannot thus be a synonym, because this term is used for a Protestant minister. With the exception offraulich, the above-mentioned adjectival derivative of Frau, which reaffirms the equation between Mutt^rand Frau, and of verwandt, 'relative', whose link with muttertich, above all through fursorglich, is not entirely clear. This form, which contains the noun M&dchen 'girl', is now avoided and generally replaced by the more generic Kinderfrau.
References Arndt, Susanne (2005), 'Fehler im System', http://www.brigitte.de/frau/ gesellschaft/ frauen_google / BBB (2002), = Sprachliche Gleichbehandlung von Frauen und Mdnnern: Hinweise, Anwendungsmoglichkeiten und Beispiele. BBB-Merkblatt des Bundesverawaltungsamtes. BMBF (2003), = Dos Bundesministerium fur Bildung und Forschung <www.bmbf.de>. BuBmann, Hadumod & Hellinger, Marlis (2003), 'German: Engendering female visibility in German', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 141-74. Corda, Alessandra (1998), Treatment of senses and collocations in dictionaries: an issue for lexicology and lexicography', in The Structure of the Lexicon in Functional Grammar. Amsterdam and Philadelphia: Benjamins, pp. 195-211. Cyrus-Eichhoff, Karin M. (ed.) (2004), Adam, Eva und die Sprache: Beitrdge zur Geschlechterforschung. Mannheim: Dudenverlag. Duden Bedeutungsworterbuch (2002), (= DerDuden in zwolfBanden. Vol. 10). Mannheim: Dudenverlag. Duden Deutsches Universalworterbuch (2003). Mannheim: Dudenverlag. Duden Richtiges und gates Deutsch (2001), (= Der Duden in zwolf Bdnden. Vol. 9). Mannheim: Dudenverlag. Duden Warterbuch Deutsch ats Fremdsprache (2003), Hg. in Zusammenarbeit rnit dem Hueber Verlag. Mannheim: Dudenverlag.
94
GENDER, LANGUAGE AND NEW LITERACY
EU (2003) = European Parliament 1999-2004. Session document. Final A5-0053/ 2003. Guentherodt, Ingrid; Hellinger, Marlis; Pusch, Luise F. & Tromel-Plotz, Senta (1980), 'Richtlinien zur Vermeidung sexistischen Sprachgebrauchs'. Linguistische Berichte 69,15-21. Hellinger, Marlis (2004), 'Empfehlungen fur einen geschlechtergerechten Sprachgebrauch im Deutschen', in Karin M. Eichhoff-Cyrus (ed.), Adam, Eva und dieSprache: Beitrage zur Geschlechterforschung. Mannheim: Dudenverlag, pp. 275-91. Hellinger, Marlis & Bierbach, Ch ristine (1993), Eine Sprache fur beide Geschlechter: Richttinien fur einen nicht-sexistischen Sprachgebrauch. Bonn: UNESCO. Kargl, Maria; Wetschanow, Karin; Wodak, Ruth & Perle, Nela (1997), Kreatives Formulieren. Anleitungen zu geschlechtergerechtem Sprachgebrauch. Universitat Wien und Bundesministerium fur Unterricht und kulturelle Angelegenheiten. Vienna: Bundeskanzleramt. Klann-Delius, Gisela (2005), Sprache und Geschlecht: EineEinfiihrung. Stuttgart: Metzler. Kunkel-Razum, Kathrin (2004), 'Die Frauen und der Duden - der Duden und die Frauen', in Karin M. Eichhoff-Cyrus (ed.), Adam, Eva und die Sprache: Beitrage zur Geschlechterforschung. Mannheim: Dudenverlag, pp. 308-15. 'kurz und bundig' (2000), Vorschldge zum geschlechtergerechten Formulieren. Universitat Klagenfurt und Frauenreferat des Landes Karnten. Lazzeroni, Romano (1993), 'II genere indoeuropeo. Una categoria naturale?', in Maurizio Bettini (ed.), Genere e ruoli nette culture antiche. Bari: Laterza. Linke, Angelika (2002), 'Das Wort in der feministichen Sprachreflexion: Eine Ubersicht', in D. Alan Cruse, Franz Hundsnurscher, Michael Job & Peter Rolf Lutzeier (eds), Lexikologie / Lexicology. Ein internationales Handbuch zur Natur und Struktur von Wortern und Wortschdtzen. HSK 21.1. Berlin: de Gruyter, pp. 121-8. Pusch, Luise F. (1984), DasDeutsch als Mdnnersprache. Frankfurt/M.: Suhrkarnp. Raffeiner, Evelyn & Thune, Eva-Maria (1999), '"Seit er gestorben ist, wollte sie nicht mehr lesen" - Sprache und Geschlecht in deutschen Grammadken', in Eva-Maria Thune (ed.), Deutsch lehren und lernen in Italien. Padua: Unipress, pp.189-204. Samel, Ingrid (1995), Einfiihrung in die feministische Sprachitrissenschaft. Berlin: Erich Schmidt. Schiedt, Margret & Kamber, Isabel (2004), 'Sprachliche Gleichbehandlung in der Schweizer Gesetzgebung: Das Parlament macht's moglich, die Verwaltung tut es', in Karin M. Eichhoff-Cyrus (ed.), Adam, Eva und die Sprache: Beitrage zur Geschlechterforschung. Mannheim: Dudenverlag, pp. 332-48. Schweizerische Bundeskanzlei (1996), Leitfaden zur sprachlichen Gleichbehandlung im Deutschen. Bern: Schweizerische Bundeskanzlei.
6 Gender construction and the Word Thesaurus for Greek Marianna Katsoyannou and Dionysis Goutsos
In our contribution we first examine the system of gender in Modern Greek, placing emphasis on the grammatical means of the language, as well as their use and participation in the construction of social gender. We thus distinguish between morphological, lexical and social gender and refer to the reflection of social roles in Greek gender. We then turn to the main question, namely the way gender construction is reflected in the Word Thesaurus for Greek, which seems to follow the inadequate lexicographic practice of non-electronic dictionaries. 1 Gende r in Modern Greek LI Morphological gender
Modern Greek has a gender-marking case system, which can best be described as a typical case of co-operation between meaning and form in the assignment of gender (see Anastasiadi-Symeonidi and Chila-Markopoulou 2003: 16). Thus, gender in nouns is marked through a morpheme, representing a fusion of the grammatical categories of number and case. Modern Greek is one of the European languages in which the grammatical category of gender makes its presence felt in large parts of the language system. To a certain extent, one can argue that this is a conservative feature of the language, carrying on an Indo-European inheritance. Thus, Modern Greek still recognizes three morphological genders: masculine, feminine and neuter. In addition, grammatical gender is found in a large number of word categories: articles, nouns, adjectives, as well as most pronouns (in the third person). Each one of the three genders has more than one inflectional paradigm, which are not equally represented either in the vocabulary or in their frequency of occurrence in discourse.1 Although most grammars agree in distinguishing three morphological classes2 (masculine, feminine, neuter), there is significant overlap in case marking to the extent that inflectional paradigms cannot be systematically related to semantic distinctions such as male/female, animate/inanimate or human/non-human. We can define
96
GENDER, LANGUAGE AND NEW LITERACY
gender in Greek as a purely grammatical category, arbitrarily (and obligatorily) assigned to the lexical units of the language: thus, all nouns are marked as masculine, feminine or neuter. To complete the picture, we should mention those cases of gender variation found in a small part of the vocabulary. There are in fact morphologically related nouns where gender corresponds to a greater or lesser distinction in meaning, e.g. napwpia (f.) 'testimony' vsfiaprupio (nt.) 'torture', 7iepi'0o/.o<; (m.) 'enclosure' vs nepiftoh (nt.) 'orchard, garden', etc. (all examples are from Triantafyllidis 1991: 219-22). There are also a few cases where the same noun can be used both in masculine and in feminine, e.g. dfifiot; 'sand', 6i
GREEK KATSOYANNOU AND GOUTSOS
97
with the noun they specify, thus indicating grammatical rather than natural gender. Grammatical gender is also assigned to proper nouns which are normally preceded by the definite article: r\ Mapia 'the Mary', o UeipOQ 'the Peter'. As regards the contribution of semantics to gender assignment, most nouns and nominalized adjectives bearing the features [+ animate] and [+ human] can be assigned two or even three genders, e.g. TtcoA^T;/? (m. noun) 'salesman' - nco^tjrpia (f. noun) 'saleswoman', TO^AO? - wq>Arj - tixpA.6 (m. - f. - nt. adjective) 'blind'. The same is true for third person pronouns marked for gender, e.g. aorog - amrj - amo 'this'. In each case the basic type is thought to be the masculine one, on the basis of which the other types are produced, according to rules for each inflectional paradigm. Nouns bearing die same features, i.e. [+ animate] and [+ human], which do not allow for gender distinctions, e.g. TO naiSi 'the child; the guy', or nouns in which gender is differentiated by lexical means, e.g. ayopi (nt.) 'boy' - Kopiwi (nt.) 'girl', are always neuter (see §1.2). On the other hand, nouns which do not bear these semantic features are not necessarily neuter, but can belong to any of the three genders. As a result of this, the neuter has a special role to play: while the correspondence between referential (natural) and grammatical gender (i.e. male/female and masculine/feminine) could lead to a bipolar distinction, in fact, the contrast is based on a three-pronged distinction, where the neuter is employed to fulfil several functions in gender construction. 1.2 Lexical gender In addition, referential gender is shown in a great variety of- mainly animate - nouns, including professional, national and proper names, as well as kinship terms. In some cases, gender is included in the meaning of the lexeme itself, e.g. the nouns ayopi 'boy' and Kopirai 'girl' are morphologically neuter, but semantically marked as male and female, correspondingly. The same is true for a number of words which belong to two semantic categories: 1 [+ animate], [+ human]: apart from the pair avcpaq 'man' - yovaitca 'woman', this category also includes kinship and familiarity terms such as narepac; 'father' - jjrjrepa 'mother', yioq 'son' - Kopr\ 'daughter', nannovQ 'grandfather' - yiayid 'grandmother', ya^npoq 'groom' - vv(pr} 'bride'. We can also find here an occasional group noun, neuter and in the plural, covering both members of the pair, e.g. ra naiSid 'the children' (meaning yioc; + Kopq 'son + daughter' or yio<; + yw<; 'son + son' or Koprj + Kopt] 'daughter + daughter'). However, in the most common case the masculine member is used as a group noun, e.g. in the sentence rjpOav 01 Ocioi f^oo 'mv uncles have come', 01 OEIOI (m. plural) can mean either 'uncle + uncle(s)' or 'uncle(s) + aurit(s)'. This use presupposes the inclusion of at least one male member in the group: if it is 'aunt + aunt(s)', the expression should be tjpOav 01 Oeici; poo 'my aunts have come'.
98
GENDER, LANGUAGE AND NEW LITERACY
2 [+ animate], [- human]: this category includes animals. In this case there are no group nouns but some alternative possibilities: (a) The animal may be named by a unique word, belonging to any of the three genders: (pa'A.aiva (f.) 'whale', OETOQ (m.) 'eagle', (pidi (nt.) 'snake'. Thus gender is not related to sex and is just considered as a particular lexical feature of the noun (see also Ralli 2003: 75). (b) The pair male/female may be represented by a pair of words composed of neuter/feminine instead of masculine/feminine, e.g. fiodi (nt.) 'ox' - ayekada (f.) 'cow', Kpiapi (nt.) 'ram' - KaraiKa (f.) 'goat', where the male is always represented by the neuter. This mainly applies to domestic animals or animals that are most familiar to humans. In conclusion, there is a complex correspondence between referential and grammatical gender in Greek, though preference for the masculine seems to be dominant in most cases. 1.3 Gender conventions Although Greek is equipped with the necessary morphological devices for a full correspondence between grammatical and referential gender, there is a clear imbalance in the use of these devices. Thus, in a phrase like ra \iir\ML fiovvd 'the high mountains', all constituents are marked as neuter, whereas in 01 (f/rjXoi koyoi 'the high hills' they are marked as masculine, following the gender, number and case of the head in both cases. On the other hand, things are different where the head includes more than one noun belonging to a different gender. According to Greek grammars (e.g. Tzartzanos 1991, vol. 1: 47-8), there are two possibilities: for inanimate nouns the neuter is used, whereas the masculine 'predominates' in the case of animate nouns. For instance, in a sentence such as: (1)
r\ Koo^iva KCU TO y/vyeio kafuiouv aav Kaivovpyia the kitchen and the refrigerator shine like brand new
The noun KOD^iva 'kitchen' is feminine, y/oyeio 'refrigerator' is neuter and Kaivovpyia 'new' is marked as neuter plural; similarly in: (2)
o naTepat; KCLI rj fjqiipa [iov tjrav eOsAovreQ aroix; O^maKovq Aycbvez my father and mother were volunteers in the Olympic Games
Here eOsMvtEQ 'volunteers' is marked as masculine plural (see also Frangoudaki 1988; Archakis & Kondyli 2002: 158-9). The clear conclusion to be drawn from these cases is that the feminine is the only gender that can never be used in cases of combined heads,4 remaining thus marginal in gender assignment. This imbalance is even more prominent in generic and group uses of
GREEK KATSOYANNOU AND GOUTSOS
99
nouns and pronouns. Apart from the cases of lexical gender noted above, we can here mention phrases addressed to a mixed audience. In cases such as: (3) (4)
av KdnoioQ evdiaqteperai . . . if someone is interested . . . noioQ unopti va //£ 7iAtjpO(f>opriaei yia . . . who can inform me about . . .
indefinite markers (KGLTIOKX;, noioc,} are in the masculine gender. The same happens in phrases belonging to a general statement: (5)
6X01 01 /za^/Tfi'g 0a e&Taawvv rtjv idia iizpa All students will be examined on the same da.
In (5), naOtjiit; 'students' is marked as masculine plural. In these cases the use of masculine is obligatory even if all students are female. As grammar textbooks put it, 'the male gender is generally the strongest personal gender' (Triantafyllidis 1991: 216 and Tzartzanos 1991 , vol. 1: 47) . It is significant that occasional deviations from this convention, coming, e.g., from university lecturers who address their class in the feminine, are generally taken as instances of joking. There are also cases where the use of the masculine gender with a generic meaning is obligatory as e.g. in [+ animate] , [+ human] nouns, which are not semantically marked as male or female but are used to refer to both: (6)
0A0jm °lm. noMrec;^ eivai iaoin anevavn aw vo^io All citizens are equal before the law.
In example (6), TIOMIEQ 'citizens' is marked as m. plural. The role of the masculine gender is reinforced by the phenomenon of gender, number and case agreement: in example (6), the masculine gender (referring to both men and women) is present in four out of the eight words of the sentence: 6/wi 'all', 01 art., no^het; 'citizens', iaoi 'equal'. Nouns like avOpwnoq 'human being', yovioq 'parent', %pr\air\<^ 'user' belong to the same category (see Makri-Tsilipakou 1989). Another indication of the imbalance in the use of morphological gender is the meaning and use of diminutives in proper names like KcoaraK^ (m., referring to men) vs KanvdKi (nt., referring to women). In these cases the diminutive for men is also in the masculine gender, whereas the diminutive for women changes into neuter. Similarly, the use of diminutives for common nouns has the effect of making the word derogatory for women (e.g. yovaiKcaci 'insignificant little woman') but not for men (e.g. avipoKi 'little man'), especially in cases where connotations work differently for men and women (see Makri-Tsilipakou 2003). This correlates with the preferred adjectives which collocate with the nouns for man and woman. A number of studies using material from various dictionaries, mainly from the Greek
100
GENDER, LANGUAGE AND NEW LITERACY
Thesaurus entitled Antilexicon (Vostantzoglou 1962), mention that the abundance of positive terms for the masculine is enhanced by the mass of derogatory terms for the feminine. The same can be observed in the case of the adjectives avdpiKoq 'male/masculine' and yovauceiot; 'female/feminine', derived from the corresponding nouns in phrases such as avdpiKe<;/yvvaiKeiez Sovfaiet; 'male/female business' or avdpiKB^/yovaiKBie^ KOD^EVTCK; 'male/ female words', which carry distinctly different connotations (see MakriTsilipakou 1996: 6; Archakis & Kondyli 2002: 159-60; Pavlidou 2002: 51-2: Pavlidou*/ a/L 2004). 1.4 Social gender and social roles
The uses of language described above must be seen as integrated within a wider socio-political frame, in which gender construction and roles are related to a particular division of labour and protection of a social status quo. In the case of Greek society, a fundamental parameter of this social code is the notion of honour, which shapes (or at least is expected to shape) notions of gender. Thus, while Greek women are not traditionally thought of as the 'weaker' or the 'fair sex', they are given a number of negative features that delimit their social role.5 Thus social hierarchy can be detected, for instance, in sayings, proverbs, fairy tales and stereotypical expressions (see §3 below) , which build on and reproduce a number of sexist stereotypes placing woman in an inferior position. One implication of this social hierarchy on the linguistic level concerns the use of surnames, which is unique among European languages. Like all nouns, surnames are formally marked for gender, just as in Slavic languages. However, in contrast to the latter, this gender is always masculine and appears in the nominative case for males and in the genitive for females. For instance, we have Dionysis Goutsos (m. nomin.) but Marianna Goutsow (m. gen.), Dionysis Katsoyannos (m. nomin.) but Marianna Katsoyannow (m. gen.). Nikiforidou (1991) suggests that the main use of the genitive in the system of Greek is to indicate ownership (see also Tsamadou 1984). As a result, the fact that the official identity in the case of women is morphologically related to her father's or husband's identity suggests that it follows the identity of her 'owner' (see also Pavlidou 2002: 44). We can add here that morphology acts in this way as a means of counteracting the recent positive effects of legislation, according to which women retain their original surname after marriage. One last category concerns the names of professions, an issue which is inextricably linked with language change and reform, relating to Greek diglossia. First, many professional nouns, as happens in many other languages, have only a masculine form, since these professions have traditionally been related to men. Thus, a number of professions like vavirfQ 'sailor yapac, 'fisherman', uSpau/.iKO(; 'plumber', have no feminine form. In this case, the problem is how language can follow social progress, that is how to create new forms that will make women visible in new professional spheres.
GREEK KATSOYANNOU AND GOUTSOS
101
What is peculiar in Greek is that, whereas different forms can often be created for the feminine gender, these forms are socially stigmatized, since they come from 'low' social dialects and are used by uneducated people. The deliberate use of such forms for women is derogatory and is usually thought to be ironic. As a result, in the place of feminine word-forms like /touAfiUT/va 'woman MP', yiarpiva 'female doctor', or npoedpiva 'female president, chairwoman', the corresponding masculine forms are used (/touAfiOTf/g, yzorpog, npoedpoq). These may equally refer to both men and women but are not totally devoid of a [+ male] semantic feature, since their use, without any further clarification, is automatically assumed to relate to men. In addition, the masculine article can also be used with those nouns even where women are involved, for instance in a sentence like
(7)
o^ npoedpo*;^ wo avMoyov fwc; eivai rjf Kupiaf X the (m.) president (m.) of our society is the (f.) Mrs (f.) X
although the feminine version can also occur, i.e.
(8)
ijf npoedpos^ roo avM.6yoo //a? eivai rjf Kvpiaf X the (f.) president (m.) of our society is the (f.) Mrs (f.) X
This problem relates to Greek diglossia and its inheritance and, at the same time, it is an instance of a linguistic sexism, which suggests that the feminine forms of certain nouns are 'words against world order' (Frangoudaki 1989: 44). Although the feminist movement has never met with popularity in Greece, language reform was achieved with the establishment of demotic Greek as the official language of Greece in 1976, following the demise of katharevousa.6 The new official language was symbolically associated with all progressive movements, but linguistic changes with respect to gender have not been particularly successful, partly because of the fact that demotic Greek was not considered prestigious enough to support new language practices, including new vocabulary formation. Going back to professional nouns, when a masculine name follows the older katharevousa morphology, the formation of a corresponding feminine version may follow either katharevousa or demotic rules. In the former case, however, the resulting type is difficult to integrate into the modern language, whereas in the latter case, the failure to use the masculine version is interpreted as ignorance and the new word is stigmatized as belonging to an uneducated idiom. In consequence, masculine forms of katharevousa origin, like npoedpot; 'president', are used for both genders, while demotic forms like npoedpiva. 'female president', are used to express derogatory views on women. This is a special application of the diglossia principle, according to which the use of a low variety in an inappropriate context may be interpreted as derogatory. The 'small difference', in this case the variant of a derivational
102
GENDER, LANGUAGE AND NEW LITERACY
ending, is sociolinguistically realized as the ridiculing of the object of reference, i.e. women. 2 Gende r in the Greek Word Thesaurus 2.1 Crucial pairs Our analysis concerns the Thesaurus of Microsoft Word 2000, which is also the most recent version, since no updating has been made for the XP version. It must be pointed out from the outset that the quality of the Greek Thesaurus is considerably lower than those of other widely used languages (e.g. the English Thesaurus) and this undoubtedly has to do with the limited tradition of lexicographical material of this kind in Greek. It is significant that the only major published thesaurus for Greek first appeared in the late 1940s and has remained unaltered since then (Vostantzoglou 1962). In the case of the electronic thesaurus in question several important defects can be observed: 1 Overabundance of synonyms, e.g. 24 synonyms are suggested for yovaiKa 'woman' and 27 for avipaq 'man'; similarly, for ofiopqHx; 'handsome, pretty' we find 33 synonyms and 27 antonyms. 2 Frequent mixing of word categories, e.g. for dvrpcu; 'man' (noun), one of the meanings given is aradepot; 'stable' (adjective), for which two synonyms are given: araOepog 'stable' (adjective) andfipdxoq 'rock' (noun). 3 Introduction of 'related words' or 'expressions', which include words or extended phrases semantically related - at least to some extent - with the main word, e.g. for dvipat; 'man' the following related expressions are given: av0pcomvo ov 'human being', A.OJIKO ov 'rational being', votffiov £ok> 'intelligent being', anoyovog TOD Addp. 'Adam's offspring'. This is completely unpredictable and does not seem to follow any clear principles. 4 In many cases, antonyms pick up different aspects of the word meaning but are always given all together in a separate group; for instance, the antonyms of Koni^a. 'girl' include navrpe^evrj yvvaiKa 'married woman', ypid 'old woman', ywt; 'son', ayopi 'boy', which are all presented together. 5 Most importantly, the synonyms given originate from a wide diversity of registers, levels of formality, field domains, etc. The problem is particularly serious, since Greek vocabulary draws on a plethora of sources from its particularly rich diachronic variation, relating to its long-standing history of diglossia. In the case of the Thesaurus, alternative forms are listed without any indications for appropriate use or any sense of usefulness for the user. For example, some of the synonyms given for KoneAa 'girl' are the dialectal forms tcrounpa and Tffouna, the old-fashioned/obsolete Kopdai, KopaaiSa, Kopaaid and via and the archaic vedviSa and Kopdaio. It could thus be argued that problems with the representation of gender in the Greek Thesaurus reflect the poor overall quality of the tool as a whole.
GREEK KATSOYANNOU AND GOUTSOS
103
What is particularly significant, however, is that defects or 'mistakes' do not equally affect both genders but seem to be less frequent in the case of the masculine or male pole of the antithesis. This imbalance is clearly shown in the arrangement of the crucial pair dvrpat; - yvvaiKa 'man' - 'woman', which is blatantly divergent in both the selection of meanings and their sequencing. Whereas for dvrpat; 'man', the meanings selected are 'man', 'soldier', 'of male age', 'spouse', 'stable', 'manly', 'noteworthy' (in this order), the meanings provided for yovaiKa 'woman' are 'woman', 'spouse' (second in order), 'maid', 'effeminate'. The two words thus turn out to represent an opposition between positive and negative qualities and in particular firmness, courage and magnitude as male qualities and servility and effeminacy as female ones. The individual terms given as synonyms also reflect this gross and rather embarrassing stereotyping, by including offensive or obsolete words. Antonyms are also revealing, since those given for man are 'woman', 'child', 'timid', 'coward', whereas those for woman are 'man', 'spouse', 'male', 'virile' (or 'hot'!), 'brave fellow'. Similarly, the related expression 'better half is only given as a synonym for 'woman'. It is clear that the selection of synonyms is related to the situation described above with reference to the structure of society and social roles attributed to women. These are rarely real synonyms that can be used in place of the word in question, but are rather attributes that are traditionally ascribed to each social gender. A surprising reversal of positive and negative qualities is found in the pair apaevmoq - Orj^DKoq 'male/masculine' - 'female/feminine'. In the case of 'male', the meanings given are 'man' and 'infertile', whereas for 'female' the meanings are 'womanly' and 'fertile' ('fertile, creative, inventive'). Although this seems to attribute positive qualities to the female side, there is nevertheless a hint of a patronizing attitude in the overall choice of acknowledging fertility as the only positive feature of women, and this is undoubtedly related to their social role. In relation to the latter, we can note the presence of the entry OTJ^VKO, a rather derogatory neuter noun, infrequently used to connote cunningness, for which two meanings are given: E^DTIVTI yuvaixa (e£vnvr} yvvaiKa 'clever woman') and KOpT] (synonym: Koptj, Kopiwi 'daughter', 'girl'). Apart from the obsolete association of the word with the generic second sense, it is interesting to observe that no synonyms are given for the corresponding apaeviKO 'male'. The same lack of balance is shown in the pairs ayopi - Kopitai 'boy' - 'girl' and yiot; - Koprj 'son' - 'daughter'. For ayopi three senses are given, corresponding to 'male offspring', 'non-adult' and 'lover' (for 'boyfriend'), with two synonyms each: apoeviKO nai8i (synonyms: apoe.viK6 naiSi, yiog), avfjXiKoq (synonyms: avyhKot;, veav/ag), ayarcnjiicoc; (synonyms: ayanrjuKOQ, spcofievot;). In contrast, three senses are given for the two terms for 'girl' with 17 and 20 synonyms each. These senses include the alternative term Koprj ('girl') with synonyms meaning 'virgin', 'toddler', 'young female', as well as the senses of 'daughter', 'lover' (for 'girlfriend') and 'female servant' or 'cleaning lady'. The picture is similar for the last pair, 'son' - 'daughter',
104
GENDER, LANGUAGE AND NEW LITERACY
since no synonyms at all are given for the male term, whereas the female term has 6 senses with 19 synonyms in total. Apart from the usual senses noted above (including 'female servant'), special emphasis is given to the sense avmcavipi ) 'unmarried female', for which the following synonyms are given: avtmavrprj, dyafiot; 'unmarried', avavdpoQ 'without a man' (obsolete), napOevos, napOiva 'virgin'. It would thus seem that the increased emphasis on female terms in these pairs goes hand in hand with the highlighting of stereotypical and outdated roles assumed for women such as virgin or cleaning lady. As noted above for the case of Vostantzoglou (1962), the plethora of terms for women is only accounted for by the preponderance of derogatory items. 2.2 Social roles
Turning now to the social role of male and female in the Greek Word Thesaurus, we can indicate a number of terms relating to activities, titles, economic and professional roles as well as evaluative terms that suggest a similar imbalance in the representation of men and women. First of all, whereas synonyms may be given for the male form of nouns or adjectives referring to professions etc. (e.g. ayportji; 'farmer', aB^r\c, 'sportsman'), no synonyms are given for the corresponding female forms. This is especially noticeable in roles which are traditionally related to women, e.g. SdffKa^OQ 'teacher', fiaaihdz 'king', voiKOKvpqt; 'house-husband'. The only exception found in the Thesaurus is also telling, since it concerns onripertj:; - unfjperpia 'male servant' - 'female servant', for which, moreover, antonyms are only given for the male form (12 forms), whereas many derogative forms such as SooAaxi, SOO^IKO, donkhaa (diminutives of Sooto 'servant', 'slave') are given in the female form. Similarly, all synonyms are given in the male form, which is reasonable for male terms but actually results in a Thesaurus loaded exclusively with male forms of nouns and adjectives and thus concealing the highly inflected character of the language - to the detriment of female forms. In relation to this, the absence of many professional and other roles and tides from the Thesaurus also contributes to the suppression of the female presence in the social sphere. For instance, there are no entries for ypanuaiEou; 'secretary', tjOonoioQ 'actor' - 'actress', Kouiubtpia 'hairdresser', voaoKOfta 'nurse', vtjmay(oy6<; 'kindergarten teacher', SaKwJioypdqxx; 'typist', 7ia)A.rjrpia 'saleswoman', oiKodeanoiva 'hostess' - let alone epyoSorpia 'female employer', enompia 'female supervisor', etc. Finally, an interesting case is that of the evaluative adjective opopvoQ 'pretty, handsome', for which synonyms are given again for the male form only. This follows the well-known rule, according to which the masculine form is used for the adjective entry; what is particularly interesting, however, is that in the particular synonyms (34 in total) and antonyms provided (27 in total), we can find particular derogative terms reserved for women only (e.g. Kapoacd^a, fiai'nov, navovK/.a, cr/oUa). Whereas similar gender-specific
GREEK KATSOYANNOU AND GOUTSOS
105
derogatory terms are also given for men (KOoaffi/iodoQ, Oepahrjt;, aawpo£, nrjOiKot;, novpyot;) , what is particularly surprising is that female-specific words should appear within a male entry and that this should happen in listing antonyms for derogatory terms. In addition, two related expressions are given (itnoDKid KCU au%copio 'to be eaten in a swallow', 'a mouthful', va rtjv nieiQ aro nortjpi 'to be drunk in a glass'), which again are female-specific and particularly patronizing and sexist. 3 Conclusio n In conclusion, it seems that the poor quality of the Greek Thesaurus affects the representation of the two genders. We believe that the restricted value of the Thesaurus is particularly unfavourable to the female gender because of the following features: 1 imbalance in both the selection of meanings and their sequencing; 2 marked prominence of male forms with respective absence of female forms; 3 systematically negative or stereotypical associations of synonyms given for women; 4 inclusion of derogative terms, even where it is clearly uncalled for. To be fair to the Word Thesaurus, these problematic points do not seem to differ from existing lexicographic practice, for which similar observations have been made (Tsokalidou 1996; Pavlidou 2002), though no systematic study exists (with the exception of the recent Pavlidou et al 2004). Our conclusions point to the fact that the design of the Greek Thesaurus does not differ from old-fashioned dictionaries of Greek, which both reflect current linguistic practices and are heavily biased towards a sexist view of language and the world. Notes 1 For statistical evidence, see Katsimali & Kavoukopoulos (1996). 2 See, however, Ralli (2002) and (2003) for an alternative classification. 3 See Ralli (2003: 63), according to which it is not always clear whether gender is a feature of the stem or the inflectional suffix. 4 This view concurs with the conclusions of Chila-Markopoulou (2003), where more details are given. 5 For a survey of related work see Papataxiarchis & Paradellis (1998). 6 Until 1976, the linguistic situation in Greece represented a prototypical case of diglossia, in Ferguson's terms. Katharevousa was an artificial form of the language, whose use was obligatory in writing as well as in formal situations.
106
GENDER, LANGUAGE AND NEW LITERACY
References Anastasiadi-Symeonidi, Anna & Chila-Markopoulou, Despina (2003), 'Iuyxpoviice<; icctt 5iaxpoviice<; Taaei<; ato yevo<; tt|<; eM.nviicr|<;', in A. Anastasiadi-Symeonidi et aL, To ysvoq. Athens: Patakis, pp. 13-56. Anastasiadi-Symeonidi, Anna; Ralli, Angeliki & Chila-Markopoulou, Despina (eds) (2003), To yevoi;. Athens: Patakis. Archakis, Argyris & Kondyli, Marianna (2002), Eiaaycaytj orrjv KoivojvioykoaaoAoyia. Athens: Nisos. Chila-Markopoulou, Despina (2003), Tevo<; icai auncpeovia ait] vea eXtaiviicfj', in A, Anastasiadi-Symeonidi et aL, To yivoq. Athens: Patakis, pp. 132-67. Frangoudaki, Anna (1988), TXxbaaa X,av0dvouoa-2'. DiniS, 82-5. Frangoudaki, Anna (1989), TXwcaa Xav6dvouaa-3 f| ytati 8ev urcdpxouv pouXeutpia; rcapd uovo xopei>Tpie<;'. Dini 4,42-4. Katsimali, Giorgia & Kavoukopoulos, Fotis (1996), ZrjTij^ara veoeAkjviKijc; yxcxrao?: AiSaxriK^ npoaeyywrj. Rethymno: University of Crete Press. Makri-Tsilipakou, Marianthi (1989), 'The gender of anthropos: an exercise in false generics'. Proceedings of the the 3rd Symposium on the Description and/or Comparison of English and Greek, School of English, Department of Theoretical and Applied Linguistics, Faculty of Philosophy, Aristotle University of Thessaloniki. Thessaloniki, pp. 61-83. Makri-Tsilipakou, Marianthi (1996), Ti dAAa^e Xoutov;' Studies in Greek Linguistics: Proceedings of the 16th Annual Meeting of the Department of Linguistics, Faculty of Philosophy, Aristotle University of Thessaloniki. Thessaloniki, pp. 435-46. Makri-Tsilipakou, Marianthi (2003), 'Greek diminutive use problematized: gender, culture and common sense'. Discourse & Society 14, 699-726. Nikiforidou, Kiki (1991), 'The meanings of the genitive: a case study in semantic structure and semantic change'. Cognitive Linguistics, 2, 149-205. Papataxiarchis, Evthymios & Paradellis, Theodores (eds) (1998), Tojoiorrftec; KOI
7
Gender on-line in the Italian Word Thesaurus Manuela Manera and Carlo, Bazzanella
1 Linguisti c gender construction in Italian Research on gender in Italian has a significant tradition, from philosophical, sociological and linguistic1 perspectives, though the current situation is composite and not really satisfactory from a woman's point of view.2 We will limit ourselves here to providing: 1 a general outline of the linguistic devices used to construct gender in Italian; 2 a minimal sketch of social gender and socio-economic roles in Italy; 3 an analysis of the changes in meanings and synonyms for nouns related to 'woman' and 'man' from a previous version (see Bazzanella et al 2000) to the 2000 version of the Italian Word Thesaurus. LI Grammatical gender In Italian, like other Romance languages (see French, Portuguese, Spanish in this volume, Romanian which has conserved the Latin neuter, see Haase 2000: 221) and several other world languages, gender is an integral part of the language system: it is grammatically marked and transparent,* both in the singular and the plural,4 pertaining to many nouns, several pronouns, both definite and indefinite articles, adjectives, participles,5 and certain numerals. Feminine (e.g. un-a6 (f.) 'a (article)', gatt-a (f.) 'cat', sdraiat-a (f.) 'lying') and masculine (un (m.) 'a (article)', gatt-o (m.) 'cat', sdraiat-o (m.) 'lying') genders are distinguished in Italian: (1) (2)
Unaf beUaf gattaf e sdraiataf vicino al camino. A lovely cat is lying by the fireplace. Un^ bel^ gatto^ e sdraiato^ vicino al camino. A lovely cat is lying by the fireplace.
Grammatical gender may correspond to biological differences (as in the case of gotta- gatto), thus being 'naturally' motivated (see below, lexical gender),7
108
GENDER, LANGUAGE AND NEW LITERACY
or may be arbitrarily attributed, in the case of inanimate entities, e.g. il cucchiaio (m.) 'the spoon', lafarchetta (f.) 'the fork'. With regard to agreement,8 which in Italian involves all the elements of the nominal syntagm (i.e. the noun itself, articles, adjectives, and determiners) , the masculine generic is widely used: 'singular and plural masculine nouns are used to represent either exclusively male referents or both male and female ones; for example, i cittadini (m.) "citizens" can signify both "male citizens" and "male and female citizens"' (Marcato & Thune 2002: 201 ).9 While non-gender-specific terms, such as persona (m., f.) 'person', allow for a double reading, both feminine and masculine (i.e. ungendered person reference), grammatical agreement is independent of die semantic bias, e.g. persona requires a feminine suffix (e stat-Of 'has been'): (3)
La persona (m., f.) cut ho parlato e statOf molto gentile. The person to whom I spoke was very kind.
Some pronouns10 which are gender underdetermined, such as chi 'who', cui 'to whom', may be disambiguated by the preceding reference or by a successive anaphora: (4) (5)
n bambino^ con cui ho giocato ha 7 anni. The child with whom I played is 7 years old. La persona con cui ho parlato e stata molto gentile; e un impiegato (m.) dawero competente. The person to whom I spoke was very kind; he is a very competent clerk.
Diminutives," which are commonly used in Italian, vary according to sex if the basic noun has the two gendered forms (e.g. monett-o (m.) vs monell-a (f.), diminutive: monell-inovs monett-ina), but maintain their grammatical gender: e.g. birb-etta (f.) 'little scallywag', even if referring to a male child. 1.2 Lexical gender Three cases of lexical gender can be pointed out: First, there is a closed class of nouns where referential gender is expressed by different lexical roots, which have the corresponding grammatical gender and may, in addition, be followed by gender-specific endings ... e.g. female-feminine (sg./pl.) madre/madri 'mother/mothers' vs. male-masculine (sg./pl.) padre/padri 'father/fathers'. Secondly, there is a nominal class, sometimes called mobile gender nouns, where feminine and masculine terms share the same lexical root, which is then specified by means of the suffixes -a or -o in the singular, and -e or -t in the plural, according to whether the referent is female or male ... e.g. female-feminine (sg./pl.) figlia/ figlie 'daughter/daughters' vs. male-masculine (sg./pl.) figlio/figli''son/sons'.
ITALIAN MANERA AND BAZZANELLA
109
Thirdly, in the cases in which the nominal root does not permit genderspecification by the suffixes -a and -o/-e, satellite elements, e.g. determiners or adjectives, have to be used in order to achieve the overt marking of referential gender. This class contains nouns such as ... giudice/giudici (sg./pl.) 'judge/ judges', and nouns derived from present participial forms, such as insegnante/insegnanti (sg./pl.) 'teacher/ teachers'. (Marcato and Thune 2002: 190-1)
Agentives are mainly formed by derivation, e.g. la (f.) contadina (i)/il (m.) contadino (m.) 'the peasant farmer' from the adjective contadin-o/-a 'peasantlike' (see ibid.: 192). Gender suffixes used in derivation are -a/-o, -a/-e (e.g. la (f.) pastor-a (f.)/il (m.) pastar-e (m.) 'the shepherd'), -trice/-tare (Vat-trice (f.)/ Vat-tore (m.) 'the actress'/'the actor'); a female-specific suffix is -essa (sometimes affected by derogatory connotations, see Sabatini 1986: 116 and Cortelazzo 1995), as in professar-essa/profess-ore. With specific regard to kinship terms, many of them have the same lexical roots and mark feminine and masculine forms with distinct morphemes: -a (f.), -o (m.); e.g. cugina 'cousin' (f.), cwgino 'cousin' (m.); zia 'aunt', zio 'uncle'. Some are undetermined, such as nipote which, furthermore, corresponds to two different kinds of blood relationship: both 'niece'/'nephew' and 'grandchild' ('grandson'/'granddaughter'). Compounding can also be used to achieve gender-specification: a personal feminine noun (e.g. la (f.) 'the' donna (f.) 'woman') is combined with a masculine nominal modifier (e.g. magistrate), i.e. la (f.) donna (f.) magistrate (m.) 'the woman magistrate', in a sequence donna 'woman'+ occupational term, or in the opposite sequence: occupational term + donna 'woman' (e.g. il (m.) giudice (m./f.) donna (f.) 'the woman judge (lit. the judge woman)' (see Serianni 1989: 120; Marcato & Thune 2002: 195-4). Historical and cultural influence is particularly evident in occupational terms,12 and, though several efforts have been made to change this attitude in Italian (see e.g. Muraro 1981; Sabatini 1986; 1987), they have but partially13 succeeded. 'Politically correct' nominal splitting (i.e. explicit, double marked use in shortened fonn, separated by a slash, such as amici/che, studenti/esse, to refer to both men and women) is still rare, collocations such as signore e signari 'ladies and gentlemen' aside. In fact, linguistic policy cannot be separated from the general socio-economic political picture: as in any other language, in Italian, too, linguistic categorization is strictly intertwined with social construction (see e.g. Piccone & Saraceno 1996) and ideology; for example, linguistic (non-)visibility mirrors socio-economic political visibility, and even reduces it. 1.3 Social gender and socio-economic roles in Italy The general sociolinguistic situation is particularly complex in Italy, where both social and geographic differences (between North/South, large cities/
110
GENDER, LANGUAGE AND NEW LITERACY
towns/villages in the country or in the mountains; see e.g. Berruto 1987) are still relevant.14 With specific regard to gender issues, the current social situation15 can be defined as mixed and contradictory between, on the one hand, some positive achievements resulting from the feminist movement in the 1970s (see e.g. the divorce referendum, the legalization of abortion, paternity leave, the creation of Commissioni pan opportunita 'Commissions for Equal Opportunities' in various public bodies), and, on the other hand, conservative attitudes16 which seem to have sprung up again in the past few years. This negative trend is seen not only in the pervasive use of a woman's body as an object of desire in advertisements (Prendimi, sono tua. .. Take me, I'm yours'; Chiamami Peroni, sard la tua birra 'Call me Peroni, I will be your beer'; Tua executive. Su misura per i tuoi affari 'Your executive. Made-to-measure for your business', where the measurements 100'-19euro-99euro are graphically referred to the female model's body), but also in sectors of the media and in the still partially persisting opinion which is reluctant to see a change in women's subordinate role. Opinions aside, recent data show a very low percentage in the number of women1' in managerial roles; e.g. in CNR (Consiglio Nazionale dette Ricerche 'National Council of Research') of a staff of 4,284 researchers, 1,575 are women, while of 377 managers, only 37 are women (see La Repubblica, 19 October 2004). According to the data of SIC (Societa italiana di chirurgia 'Italian Surgeons' Society'), only 1 per cent of women hold top positions in that field (see La Repubblica, 19 October 2004). 2 Analysin g gender in the Italian Wor(l)d Thesaurus 2.1 The crucial pairs in 2000
Besides the expected expansion of man (meant as human being) to include also women, the following conclusions were drawn from the analysis of the meaning distinctions and synonyms of nouns related to 'woman' and 'man' in the 1998 version of the Italian Ward Thesaurus (see Bazzanella et al 2000: 195-202). Male entries are more detailed, accurate and wide-ranging, both in meanings and in synonyms; see e.g.:18 (6a) (6b) (7a) (7b)
bambina: fanciulla (bimbo, ragazza); figlia . bambina. bimbo (fanciullo, marmocchio, /wzrgo/o); figlio (rampotto, figliolo, creaturd); beb e (infante, lattante, neanato, pupa, baby, maschio). compagna:. arnica (donna, amante, sposa, ragazza). compagna amico; ragazzo (uomo, amante, partner, amore, marito, sposo); compare.
In some cases, the semantic fields are not treated equally, e.g. while signora 'lady', 'Mrs' is related to moglie 'wife', signore 'gentleman', 'Mr' is not related
ITALIAN MANERA AN D BAZZANELLA
111
to marito 'husband' (on the other hand, sovrano 'king' is provided as a synonym for signore): (8a) (8b)
signora: donna (dama); mogli e (consorte, coniuge); padrona. signore. uomo ; padron e (proprietario); sovran o (principe, gentiluomo (galantuomo).
re);
In other cases, the female counterpart of male entries is missing, e.g. there is re 'king', and not regina 'queen'; principe 'prince' and not principessa 'princess'. Also the distribution of feminine and male lexemes varies, implying variation in visibility: (9a) (9b)
arnica: compagna; amante (donna). arnica compagn o (intimo); alleato ; sostenitore (fautore, simpatizzante); amante (innamorato); conoscente; confidente.
The asymmetry is particularly evident when a single synonym is given, as in the following example: (1 Oa) domestica: coif (donna). (lOb) domestica: collaboratore. In general, the universality and power of'man' are boosted, while the woman is reduced to subordinate roles: (lla) donna: femmina ; signora (dama); mogli e (consorte, sposa); domestica ; compagna (arnica); ragazza ; coif. (lib) uomcr. esser e (umano) ; maschio ; individu o (signore); amant e (compagno) ; sposo (marito). In a nutshell, a clear asymmetry in the treatment of the pairs related to 'man' and 'woman' (the latter being often invisible, and in any case subordinate) is apparent in many ways. 2.2 Corpus in 2004 The number of lexemes checked in the Word Thesaurus 2002, Office XP, Windows XP, Microsoft (see MicrosoftSvord2002; Thesaurus©!994-2000 ExpertSystemModena) is 537. They can roughly be grouped into the following 11 categories: gender (55), age (52), relationship (80), family (46), activities and social roles (157), professional titles (43), fictitious referents (11), qualities (27), sex and maternity (31), animals (30), other (4). The analysis concerns: • pairs with different lexical roots which symmetrically distinguish male and
112
• • • • •
GENDER, LANGUAGE AND NEW LITERACY
female (e.g. uomo - donna 'man' - 'woman', padre - madre 'father' 'mother'; see first class in §1.2); the mobile gender nouns, based on the same lexical roots (cugino - cugina 'cousin' (m./f.), (More - attrice 'actor' - 'actress'; see second class in §1.2); gender-indefinite terms (e.g. giudice 'judge'; see third class in §1.2);19 masculine generics (see above); words related to a specific sex (e.g. gravidanza 'pregnancy', pene 'penis'); words relevant to gender issues (e.g. sesso 'sex', verginitd 'virginity').
By widening the corpus with regard to the number of items and the inclusion of antonyms, we have tried to sharpen the analysis of this technical tool and to identify the picture of gender that is mirrored/(re)constructed by its choice and lexicographical structuring of items. 2.3 Symmetry and asymmetry
2.3.1 Correspondences and absences In general, there is greater symmetry between male and female forms in the current Thesaurus than in the preceding version: e.g. amico - arnica 'friend' (m./f.) are completely symmetrical20 (see §2.1. and the on-line tables at www.lingue.unibo.it/wordthesaurus). Complete symmetry is achieved, in some cases, by the presence of mobile gender nouns, as well as the already mentioned amico- arnica, orfano- orfana 'orphan' (m./f.), anziano- anziana 'elderlv person' (m./f.).21 However, mobile gender nouns do not always have symmetrical entries, e.g. commesso lists: impiegato 'clerk', 'employee' (incaricato 'person in charge', dipendente 'employee', apprendista 'apprentice', lavorante 'worker', addetto vendite 'sales person', venditore 'seller'), rappresentante 'sales representative' (agente 'agent'), and as antonyms padrone 'proprietor', principale 'chief, titolare 'owner'; while commessa is not understood, as is usual, as 'employee' (f.), but only as ordinazion e (ordine, ordinativo, commissione) 'order, errand'. Another example of partial asymmetry is the pair vecchio 'old man' - vecchia 'old woman', which are qualitatively and quantitatively symmetrical except for their last meaning: in the male form patriarca 'patriarch' appears, in the feminine madre 'mother'. Symmetries can also be only apparent, hence misleading, e.g. among the svnonyms for the meanings offanciuUo 'young boy' andfanciuUa 'young girl', signorino '(young) master' and sigriorina 'Miss' may appear symmetrical, but signorino is now obsolete and conveys different meaning and usage. Some feminine forms which were missing in 1998 have been inserted, e.g. deputata 'member of Parliament' (f.), soldatessa 'soldier' (f.), both awocata and awocatessa 'lawver' (f.). However, while regina 'queen' (see §2.1) has arrived, sindaca 'mayor' (f.) is still missing. Other asymmetries, gaps and absences ('flaws' aside, see §2.3.3) are more significant with regard to the gender issues. Let us present die results of our research in some detail. In 10 cases out of
ITALIAN MANERA AND BAZZANELLA
113
118 which should/could be symmetric, die feminine lexeme is absent, e.g. there is patriarca 'patriarch' but not matriarca 'matriarch'. In odier cases, die feminine lexeme of a pair only occurs as synonym of anodier item, but is not an entry in itself, e.g. zarina 'czarina' appears under imperatrice 'empress' and regina 'queen', puericultrice 'nurse' under bambinaia 'baby-sitter', and girlfriend** under innamarata 'woman in love'. The same holds for femminismo 'feminism', which appears only as an antonym of maschilismo 'male chauvinism'. These cases of omission seem to depend on a lexicographic choice, while die absence of others, related to common and widespread words, seem completely unintentional, e.g. die following entries are missing: fratetto 'brodier' (while sorella 'sister' is present), maschilista 'male chauvinist', femminista 'feminist', patriarcato 'patriarchy', matriarcato 'matriarchy', transessuale'transsexual', lucciola 'fire-fly' (which also has a metaphorical meaning, i.e. 'prostitute'), infermiere- infermiera 'nurse' (m./f.), volontario- voUmtaria 'voluntary (assistant)' (m./f.) (both are registered only as adjectives with die meaning of intenzionale'intentional'), lupo- /w/wz'wolf (m./f.). The asymmetry is more substantial in many cases, e.g. in maschio 'male'/ femmina 'female' the feminine meanings focus on relationship and weakness,"3 while die male ones focus on individuality and strengdi. The first meaning of femmina is donna, to mean, as die synonyms underline, arnica, amante, compagna, moglie, sposa, partner; die second meaning is persona timida 'shy person' (i.e. persona debole 'weak person', persona pavida 'timid person'). The woman seems to be considered only in relation to die man, who, on die contrary, is seen as an individual (see die Introduction); maschio (as noun) includes bambino 'child' (m.) and uomo aitante as meanings, with die respective synonyms ragazzo, uomo for die former, and uomo robusto for die latter. Another persistent source of asymmetry derives, paradoxically, from die gender-indefinite terms (see §1.2, third class), particularly die ones which should be exploited by following a politically correct usage, but which, on die contrary, turn out to be only, or mainly, masculine, e.g. single provides die unique meaning of scapolo 'bachelor',24 celibe 'not married' (m.), and ammogliato 'married' (m.) as antonym; giovane (as noun) lists, beside die correct persona fra I'adolescenza e la maturitd 'person between adolescence and maturity', die meanings scapolo 'bachelor', giovanotto 'young man', celibe 'not married' (m.), widiout die feminine correspondences. 2.3.2 Quantity and distribution Generally, die male lexemes offer more meanings, i.e. more quantitative information, than the corresponding feminine ones, though diere has been an increase in symmetry in die number of meanings, synonyms, and lexemes referring to women/men as compared to die preceding version, e.g. die pair compagno/a is symmetrical now (see Table 7.1). On the contrary, in die preceding version compagna was only arnica (donna
114
GENDER, LANGUAGE AND NEW LITERACY
Table 7.1 The pair compagno- compagna in the Word Thesaurus 2002 compagno Meanings amico (s) convivente (s) Commilitone (s) collaboratore (s) confratello (s) Complice (s) comunista (s) pan (s)
Antonyms
compagna Synonyms amico, collega, socio, compare, sodale convivente, partner, consorte, amante commilitone, camerata collaboratore confratello complice, correo comunista pari, paragonabile, simile, corrispondente nemico (s), awersario (s), concorrente (s), estraneo (s)
Meanings SIMM. SIMM. SIMM. SIMM. SIMM. SIMM. SIMM. SIMM. SIMM.
Synonyms
SIMM! SIMM. SIMM. SIMM. SIMM. SIMM. SIMM. SIMM. SIMM.
'woman', amante 'lover', sposa 'bride', ragazza 'girl'). However, the quantitative asymmetry persists, and concerns different levels. With regard to the morphological class, one finds cases such as umano 'human' (m.) and umana 'human' (f.); while the former appears as both an adjective and a noun, and shows the meaning uomo 'man', the latter appears only as an adjective. More commonly, the quantitative asymmetry concerns the number of entries and the social gender (see §1.3), as in padre - madre 'father' 'mother": (12a) padre: genitor e (capo della famiglia?b babbo, papa, procreatore); progenitore (capostipite, antenato, avo); uom o venerabil e (uomo venerato*6); autore27 (inventore, scopritore, fondatore, iniziatore, maestro, guida, caposcuola, promotore, fautore, benefattore, protettore, mecenate);sacerdote (prete, monaco, frate); sant o patrono ; Di o (Padre Eterno, Creatore). Contrari (antonyms) (figlio, discendente, successore, seguace, discepolo, epigono, diavolo, demonio, Satand). (12b) madre: genitric e (mamma}', femmina ; suora (monaca, superiora, priora, badessa); origin e (causa, matrice, fonte, sorgente); matrice ; co n figli ; principale (d'origine). This pair has worsened from the point of gender hierarchy: in the preceding version capofamiglia 'head of family' was not inserted under padre 'father'; under genitore 'parent' both procreatore 'procreator' (generatore 'creator') and parente 'parent' (madre 'mother', padre 'father') were registered, while the latter has now been omitted. The same asymmetric proportion can be found for padrone 'master' - padrona 'mistress': while padrone is possessors, principale, signore, persona riverita; armatore, and, as an adjective, conoscitore 'one who knows', 'expert' (m.), the correspondent padrona 'owner' (f.) is only conoscitrice 'one who knows', 'expert' (f.).
ITALIAN MANERA AMD BAZZANELLA
115
Not surprisingly, asymmetry is particularly evident in professional titles and activities, where the same profession includes different levels in the male entry (e.g. cuoco 'cook' (m.): cuciniere; chef, gastronomer, while cuoca 'cook' (f.) is only cuciniera), and the prestigious activities are reserved for men: e.g. businessman is a meaning only for the male lexeme operatore 'tradesman', which is symmetrical to operatrice 'tradeswoman'). In general,28 in the semantic field of domestic help, the feminine lexeme is richer than the corresponding male one (given the traditional difference of roles, see below donna di servizio 'house maid'), but redundant. In some cases, as in cameriere 'waiter', the trend is reversed: the male lexeme registers also, among its meanings, maggiordomo 'butler' (lacche 'lackey') and ragazzo 'boy', thus referring to other facets and functions with different prestige, and mixing different usage domains (see §2.3.3), i.e. diachronic (see lacche, which is antiquated) and geographic variables (see ragazzo, which is used only in certain Italian regions). Strangely enough, even when almost complete symmetry is obtained by an apparent automatic reversal, some terms are deleted, as in awocato- awocata 'lawyer' (m./f.), where giurisperito'jurisconsult' and causidico'pettifogger' do not have feminine correspondents: (13a) awocata. legate (s) (legate, giureconsulto, difensore, procuratore, consulente); protettor e (s) (protettore, patrocinatore); azzeccagarbugli 29 (s) (azzeccagarbugli, mozzaorecchi, leguleio (sic)); giurisperito (s) (giurisperito}; causidico (s) (causidico). (13b) awocata/awocatessa: legate (s) (legale, giureconsulta, difensora (sic), procuratrice, consulente); protettrice (s) (protettrice, patrocinatrice); azzeccagarbugli (s) (azzeccagarbugli, mozzaorecchi, leguleia (sic)). If we consider the pair signore 'Mr' - signora 'Mrs', at first sight they look similar, but under signora the religious correspondent of Dio 'God' (which is the sixth male meaning) is missing, whereas Nostra Signora is quite common in Italian to refer antonomasially to the Madonna (see Zingarelli 2003, which gives the collocation an autonomous entry). Furthermore, signora does not include the entry donna riccd® 'rich woman', while under signore, among the synonyms of uomo ricco, 'rich man', one finds uomo distinto 'distinguished (gentle)man', uomo elegante 'elegant man'. A case in another domain where the feminine lexeme is richer is pupa 'baby', 'doll', which has several, but redundant, meanings, producing a negative effect,31 since sexual appeal is stressed.32 The asymmetry derives also from the different sequentiality and ranking within the main entry, and from the distribution of meanings and synonyms; a term which should be registered as a synonym is given as a meaning, gaining extra visibility and introducing an imbalance between the female/male pair: e.g. ragazzo 'boy' lists among the meanings both fidanzato 'fiance' and boyfriend, while ragazza 'girl' gives only fidanzata 'fiancee' (and see below the pair par excellence: donna- uomo 'woman/man').
116
GENDER, LANGUAGE AND NEW LITERACY
2.3.3. Some limits Though the Word Thesaurus provides users with an undoubtedly useful device, some limits should be pointed out: • It does not signal varieties of language different from the standard usage, neither with regard to diachrony (e.g. fantesca 'maid', vecchiardo 'old man', putto 'putto'are no longer used), nor with regard to register (e.g. matusa 'aged people', sbarbina 'girl' are typical of young people's language), nor with regard to the regional diffusion (e.g. picciotto 'boy' is Sicilian). • It does not signal ironic, metaphorical (e.g. lucriola3*) or derogatory uses. • It does not provide the user with contextual cues in general: here we need only point to the oddity of using mozzaorecchi '[literally:] cut-ears' as a synonym of awocato/a 'lawyer' (m./f.). Beside these limits, there are some problems which seem to derive from an apparent automatic reversal from man to woman, e.g.: • Among the meanings of fanciulla 'girl', sbarbateUo 'greenhorn' (!) (m.), scugnizzo 'urchin'34 are listed. • In domestica 'maid', among the synonyms of servitrice 'servant' (f.) one finds famiglio 'man-servant', which is only referred to a male, and is completely obsolete. • Under genitrice 'parent' (f.): the first meaning is generators 'parent' (m.). • Under mantenuta 'kept woman' one finds gigolo. • Under sfruttatrice 'exploitress', symmetrical to the male sjruttatore 'exploiter' as synonyms of protettrice 'protectress', symmetrical of protettore 'protector', we find ruffiana 'madam', pappone 'pimp' (m.), lenone 'pander' (m.). There are few cases of possible automatic reversals, or some sort of influence, in the opposite direction (from woman to man), e.g., the entry gravido 'pregnant', 'loaded' (m.), which enumerates among its meanings incinta 'pregnant' (f.), and playboy, which strangely enough corresponds to divoratrice di uomini 'maneater'. Other flaws35 cannot be understood as a possible reversal from masculine to feminine form, e.g.: • Under maschile 'male' (adj.) among the synonyms of da uomo 'as a man', one finds maritale (i.e. related to husband, or to marriage). • Under accompagnatore 'companion' and accompagnatrice 'female companion' among the synonyms of guida 'leader' one reads hostess; and under accompagnatrice 'female companion' among the synonyms of guida one finds steward. • Among the synonyms of papa 'father' one also finds papa 'Pope'.56
ITALIAN MANEKA AND BAZZANELLA
117
3 Conclusio n To conclude, let us quote the crucial pair uomo- donna 'man' - 'woman': (14a) uomo: essere uman o (s) (essere umano, essere ragionevole, creatura pensante, creatura)', gener e umano (s) (genere umano, specie umana)', indi viduo (s) (individuo, cittadino); Cristian o (s) (cristiano, fi^io di Adamo [see below: figlia d'Eva]); maschi o (s) (maschio)', signor e (s) (signore, tale, tipo, tizio [see below: signora]); il prossimo (s) (ilproximo, umanita, gente); sesso forte (s) (sesso forte, compagno, marito, sposo, coniuge, consorte, partner, amante, convivente, fidanzato, innamorato, moroso, lui); adulto (s) (adulto); operai o (s) (operaio, tecnico, dipendente, lavoratore, lavorante, addetto, incaricato); soldat o (s) (soldato, militare, milite, uomo armato, marinaio); uno (s [sic]) (uno, quakuno). Antonyms: animale (s), donna (s), femmina (s), gentil sesso (s), sesso debole (s), bambino (s). (14b) donna: femmin a (s) (femmina); genti l sesso (s) (gentil sesso, bel sesso, sesso debole); figlia d*Eva (s) [see above: cristiano] (figlia d'Eva); signora (s) [see above: signore] (signora, signorina); mogti e (s) (moglie, sposa, coniuge, consorte, convivente, concubina, mantenuta); amant e (s) (amante, compagna, amichetta, arnica, amore, fidanzata, innamorata, flirt, partner, betta, fiamma, morosa, ragazza, filarino, lei); nubile (s) (nubile, zitetta): tale (s) (tale, tizia, una); ganza (s) (ganza, pollastra); donna di servizio (s) (donna di servizio, domestica, cameriera, cottaboratrice famittare, coif, governante, fantesca); padron a (s) (padrona); dam a (s) (dama, regina). Antonyms: maschio (s), uomo (s), sesso forte (s), signore (s), padrone (s). adolescente (s), fanciulla (s), teen ager (s), giovinetta (s), sbarbina (s).
As we can see, in the most recent version of the Italian Word Thesaurus there is still a recurrent asymmetry with regard to the lexeme donna 'woman'.3' Besides being figlia d'Eva 'Eva's daughter', while uomo 'man' is not 'son of Adam', and not cristiana 'Christian' (£.), while man is cristiano 'Christian' (m.), among the synonyms of moglie 'wife' there is not only concubine 'concubine', but also mantenuta 'kept woman'. While donna is still characterized by gentil sesso 'gentle sex', bel sesso 'beautiful sex', sesso debole 'weak sex', all the relational male lexemes (compagno 'companion', marito 'husband', sposo 'bridegroom', coniuge 'spouse', consorte 'consort', partner, amante 'lover', convivente 'live-in-partner'.ykfcmzato 'fiance', innamorato '(man) in love', 'lover' (m.), moroso 'boyfriend', lui 'he') are listed under the meaning sesso forte 'strong sex'. With regard to the number of meanings, synonyms, and antonyms, the quantitative asymmetry is markedly on the side of the woman. Furthermore, while the man is a 'human being', 'humanity', an 'individual', etc., the main semantic fields of the woman are sexual and professional, i.e. the subordinate professions, related to lower status, as donna di servizio 'housemaid' or simply donna 'woman1. The asymmetry which still characterizes the Italian Thesaurus partially corresponds to the covert lexical asymmetries which have been pointed out both
118
GENDER, LANGUAGE AND NEW LITERACY
in standard Italian38 and in dialects.39 What has been called 'semantic loading' (Marcato & Thune 2002: 199) appears evident also in idiomatic expressions (e.g. sei solo una donna 'you are only a woman'), and in proverbs, such as Donna al volante, pericolo costante 'A woman at the wheel, a constant danger' (at odds with the insurance data). Of course, die Thesaurus structure itself is problematic. By being unavoidably a-contextual it reinforces stereotypical categorization and hinders die consideration of several important parameters. Among these the most significant are the specific parameters of relationship and power,40 not only for the topic of this volume in general, but for the Italian situation in particular, which is marked by unsolved issues and contradictory tendencies. Acknowledgements Sections 1 and 3 were written by Carla Bazzanella, section 2 was written by Manuela Manera. This chapter has profited considerably from discussions widi Camilla Bettoni, Eva-Maria Thune and Simona Leonardi. Notes 1 See e.g. Berretta (1983), Sbisa (1984), Sabatini (1986,1987), Violi (1986), Cortese & Podesta (1987), Marcato (1988; 1995a), Bazzanella & Fornara (1995), Marcato & Thune (2002), Orletti (2001). 2 A recent UN report (Comitatoper I'eliminazione detta discriminazione contro le donne, January 2005) highlights the fact that women in Italy are mainly perceived as sexual objects and subjects responsible for child-raising. Italian women have scant access to managerial roles (in the Parliament their presence is limited to 10 per cent, thus putting Italy in 79th place in the world ranking), their wages are lower (up to 35 per cent less), and their jobs often temporary (see La Repubblica, 3 March 2005). 3 According to Dressier & Thornton:
4 5 6 7 8
Gender is always transparent with articles and other prenominal modifiers, unless there is final vowel elision, as in the singulars I'erede'lhe heir' (m. or f.), indef. un erede (m.), phonologically identical to un'erede(f.), quest'erede'this heir' (written: questo/a erede). Gender is also transparent in all other modifiers with the exception of indeclinable adjectives (blu 'blue'), adjectives in -e, pi. -i (e.g. felice'happy'), some pronouns (c/tf'what, which'). (1996:8) Unlike Latin, Italian is not inflected for case: see Vincent (1988). This high redundancy in marking gender (as well as number) turns out to be useful in the cases where nouns are not gender-determined (see below). -a is the typical feminine suffix. As Doleschal (2000: 160) states: 'I conceive of the ending -a as signalizing both gender and sex in Italian.' As Violi (1986: 41) underlines, the gender semantic distinction cannot be reduced to primarily formal aspects, given the profound symbolism within language. According to Hockett (1958) and Corbett (1991), grammadcal agreement is the
ITALIAN MANERA AND BAZ7ANELLA
119
determining criterion of gender; for different proposals see Dahl (2000), Doleschal (2000),Vogel (2000). 9 See also Violi (1986), Sabatini (1987), Marcantonio & Pretto (1988: 315-22). 10 'Within the Italian pronominal system, only personal, demonstrative and possessive pronouns (third person singular and plural) have gender-variable forms' (Marcato & Thune 2002: 197). 11 The use of diminutives has been traditionally attributed to women's language (see, among others, Dressier & Merlini Barbaresi (1994), De Marco (1995)), but it is not limited to them. 12 Involvement of Italian women in the job market is the lowest among the 25 EU countries (ISTAT 2005:386). The difference between regions is highly significant: from 65 per cent in Emilia Romagna to a bare 40 per cent in Sicily, Apulia and Campania. 13 The use of signora 'Ms, Mrs' instead of a professional title (e.g. professoressa (f.) 'Professor', dottoressa (f.) 'Doctor'; see Sabatini (1987: 111), Marcato and Thune (2002: 202)) is still widespread. 14 Of course, the sociolinguistic parameters are also significant in other countries and should be taken into account with regard to gender (see e.g. HoudebineGravaud (2003)'s analysis of France, and several contributions in this volume). 15 We shouldn't forget that women in Italy were granted the right to vote only in 1946, 40 years after women in Finland (where the President of the Republic and half of the Government ministers are currently women). 16 See the substitution of a woman's own surname with that of her husband's (art. 143bis of the Civil Code); and the proposed amendments to the Civil Code with regard to husband's and children's surname (disegno di legge 1739, 2002) are still under examination. In contrast, in France, since January 2005 children are free to choose either the mother's or the father's surname. 17 The number of women between 25 and 54 years of age employed in Italy today is 42 per cent (compared to the OCSE average of 76.4 per cent), thus ranking among the lowest in the world, just above Turkey, South Korea and Mexico; see La Stampa, 30 August 2004. 18 In this and in the following examples the meanings are in bold, whereas the items given for each meaning stand in brackets, in italics. 19 Some words borrowed from English, e.g. broker (which is correctly registered under the feminine lexeme mediatrice 'mediator (f.)'), are undifferentiated with regard to gender, others are referentially clear, such as boyfriend, playgirl, etc. 20 In the tables below 'SIMM.' marks a totally symmetry in the two entries, while 's' means 'substantive noun'. 21 Even a completely unlikely succeditrice, registered as not common by DISC (1997), is found. 22 Compare the lexical richness of boyfriend, which provides the following meanings: ragazzo (fidanzato, fidanzatino, compagno, amico, moroso, innamorato, filarino, flirt., amichetto, cavatiere, accompagnatore). 23 See also under mogtie 'wife', which enumerates meta 'half (the sole term without male correspondence) among the synonyms of the meaning sposa 'bride'. 24 With regard to zitella, it gives as antonyms moglie and sposa, but also madre 'mother'. 25 Under Italian law (1. 19 May 1975, n. 151), the capofamigha 'head of family' is not the father: both wife and husband are encharged with the same rights and obligations with regard to family management.
120
GENDER, LANGUAGE AND NEW LITERACY
26 On the contrary, madre 'mother' does not appear either as donna venerabile 'venerable woman' or donna venerata 'venerated woman'. 27 Consequently, only autore 'author' (m.) (vs autrice 'author' (f.)) registers padre 'father', given after the meanings creatore 'creator' and scrittore 'writer' (m.) (which are symmetrical in autrice). 28 But seruitore - seruitrice 'servant' (m./f.), e.g., are symmetrical. 29 Not found as lexeme. 30 This entry is given both by De Mauro (2000) and Zingarelli (2003). 31 Another case of negative asymmetry in the feminine lexeme is maga 'sorceress', which besides the common entries gives the following meanings: Strega 'witch', donna abile'skilled woman', seduttrice'seducer' (f.). 32 Other feminine lexemes (such as pottastra 'chick', 'simple girl', 'lassie' and bambola 'doll') emphasize sexuality and the lack of intelligence/autonomy. This insistence on sexuality is limited to the feminine entries: e.g. one of the meanings offanciulla (but not offanciullo) is verging 'virgin'. 33 E.g. DISC (1997) registers it 34 Scugnizzo is regionally marked. 35 A simple mistake: under amore 'love' among the synonyms of persona amata 'loved person' one reads lui, lui 'he', 'he' instead of lui, lei 'he', 'she'. 36 The lexeme papa, compared with mamma 'mother' is also interesting with regard to quantitative asymmetry: papa, genitore (padre, babbo, pa\ papa [sic]), Antonyms (figlio). mamma: madre (genitrice), femmina . Furthermore, varieties of registers are offered only for the male lexeme, while in mamma the sexual side is highlighted (femmina). 37 'Woman': 12 meanings + 35 synonyms + 11 antonyms = 58 vs 'man': 12 + 34 + 6 = 52. 38 See, e.g., oltraggiare 'to offend outrageously and abusively' referred to men, which involves a sexual connotation if referred to women, see Marcato (1995a: 259-60). 39 With regard to dialects, see e.g. Sicilian (see Riolo 1995), and Venetian (see Marcato & Thune 2002: 200). 40 Following Bazzanella & Fornara (1995)'s data, the social function and the interactional role of participants in a corpus based on job interviews heavily affect the use of so-called 'hedges' and specific discourse markers (which were traditionally attributed to women's language, see Lakoff 1975). E.g. senta, ascolti 'listen' are meaningfully correlated with the interviewer, both man (36 out of 37) and woman (14 of 16).
References Bazzanella, Carla & Fornara, Orsola (1995), 'Segnali discorsivi e linguaggio femminile: evidenze da un corpus', in Gianna Marcato (ed.), Donna e linguaggio. Padua: Cleup, pp. 73-86. Bazzanella, Carla; Gallardo, Catherine; Guil, Pura; Manera, Manuela & Tejada, Paloma (2000), 'Categorizzazioni del femminile e del maschile nelle nuove tecnologie: prime ricerche nel Thesaurus italiano, spagnolo, francese, inglese di Word'. Cuadernos defilologia italiana, 7. 193-245. Berretta, Monica (1983), 'Per una retorica popolare del linguaggio femminile, owero: la lingua delle donne come costruzione sociale', in Franca Orletti (ed.). Comunicare nella vita quotidiana. Bologna: II Mulino, pp. 215-40.
ITALIAN MANERA AND BAZZANELLA
121
Berruto, Gaetano (1987), SocioKnguistica dell'itaKano contemporaneo. Rome: La Nuova Italia. Corbett, Greville (1991), Gender. Cambridge: Cambridge University Press. Cortelazzo, Manlio (1995), 'Perche non si vuole la presidentessa?', in Gianna Marcato (ed.), Donna e Knguaggio. Padua: Cleup, pp. 49-52. Cortese, Giuseppina & Podesta, Stella (1987), 'Strategic di interazione verbale: le donne nel parlato radiofonico', in Cesare G. Cecioni & Gabriella Del Lungo Camiciotti (eds), Lingua letteraria e lingua del media neU'italiano contemporaneo. Florence: Le Monnier, pp. 102-39. Dahl, Osten (2000), 'Animacy and the notion of semantic gender', in Barbara Unterbeck & Matti Rissanen (eds), Gender in Grammar and Cognition. Berlin: Mouton de Gruyter, pp. 99-115. De Marco, Anna (1995), 'L'influenza del sesso nell'uso dei diminutivi in italiano', in Gianna Marcato (ed.), Donna e Knguaggio. Padua: Cleup, pp. 87-98. De Mauro, Tullio (2000), II dizionario deUa lingua italiana. Turin: Paravia. DISC (1997), Dizionario della lingua italiana, Francesco Sabatini & Vittorio Coletti (eds). Milan: Rizzoli-Larousse. Doleschal, Ursula (2000), 'Gender assignment revisited', in Barbara Unterbeck & Matti Rissanen, Gender in Grammar and Cognition. Berlin: Mouton de Gruyter, pp. 117-66. Dressier, Wolfgang & Merlini Barbaresi, Lavinia (1994), Morphopragmatics: Diminutives and Intensifiers in Italian, German, and Other Languages. Berlin: Mouton de Gruyter. Dressier, Wolfgang U. & Thornton, Anna Maria (1996), 'Italian nominal inflection'. Wiener linguistische Gazette, 57-9,1-26. Haase, Martin (2000), 'Reorganization of a gender system: the Central Italian neuters', in Barbara Unterbeck & Matti Rissanen (eds), Gender in Grammar and Cognition. Berlin: Mouton de Gruyter, pp. 221-36. Hockett, Charles F. (1958), A Course in Modern Linguistics. New York: Macmillan. Houdebine-Gravaud, Anne-Marie (2003), 'Trente ans de recherche sur la difference sexuelle, ou le langage des femmes et la sexuation dans la langue, le discours, les images'. Langage et societe, 106, 33-61. ISTAT (Istituto Statistico Nazionale) (2005), Rapporto Annuale. La situazione del Paese nel 2004. Rome: ISTAT. Koniuszaniec, Gabriela & Blaszkowska, Hanka (2003), 'Language and gender in Polish', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 259-85. Lakoff, Robin (1975), Language and Woman's Place. New York: Harper & Row. Marcantonio, Angela & Pretto, Anna Maria (1988), 'II nome', in Lorenzo Renzi, Giampaolo Salvi & Anna Cardinaletti (eds), Grande grammatica italiana di consultazione, Vol. I. Bologna: II Mulino, pp. 315-32. Marcato, Gianna (1988), 'Italienisch: Sprache und Geschlechter. Lingua e sesso', in Lexicon der Romanistischen Linguistik (LRL), Vol. IV. Tubingen: Niemeyer, pp. 236-46. Marcato, Gianna (1995a), 'La gentildonna di Guascogna', in Gianna Marcato (ed.), Donna e Knguaggio. Padua: Cleup, pp. 249-65. Marcato, Gianna (1995b) (ed.), Donna e Knguaggio. Padua: Cleup. Marcato, Gianna & Thune, Eva-Maria (2002), 'Gender and female visibility in Italian', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 187-217.
122
GENDER, LANGUAGE AND NEW LITERACY
Michard, Claire (2003), 'La notion de sexe en francais: attribut nature! ou marque de la classe de sexe appropriee?'. Langageet societe, 106, 63-80. Muraro, Luisa (1981), Maglia e uncinetto: racconto linguistico politico sull'inimicizia tra metafora e metonimia. Milan: Feltrinelli. Orletti, Franca (ed.) (2001), Identitd di genere nella lingua, nella cultura, netta societa. Rome: Armando.
Piccone, Stella S. & Saraceno, Chiara (eds) (1996), Genere: La costruzione sociale del femminile e del maschile. Bologna: II Mulino. Riolo, Salvatore (1995), 'La misoginia nelle parlate italiane', in Gianna Marcato (ed.), Donna e linguaggio. Padua: Cleup, pp. 377-87. Sabatini, Alma (1986), Raccomandazioni per un uso non sessista detta lingua italiana. Rome: Presidenza del Consiglio dei Ministri. Sabatini, Alma (1987), Ilsessismo nella lingua italiana. Rome: Presidenza del Consiglio dei Ministri. Sbisa, Marina (1984), La mamma di carta. Per una critica dello stereotipo materno. Milan: Emme Edizioni.
Serianni, Luca (1989), Grammatica italiana: Italiano comune e lingua letteraria (con la collaborazione di Alberto Castelvecchi). Turin: UTET. Vincent, Nigel (1988), 'Italian', in Martin Harris & Nigel Vincent (eds), The Romance Languages. London: Routledge, pp. 279-313. Violi, Patrizia (1986), L'infinito singolare: Considerazioni suUa differenza sessuale nel linguaggio. Verona: Essedue. Vogel, Petra M. (2000), 'Nominal abstracts and gender in modern German: a quantitative approach towards the function of gender', in Barbara Unterbeck & Matti Rissanen (eds), Gender in Grammar and Cognition. Berlin: Mouton de Gruyter, pp. 461-93. Zingarelli, Nicola (2003), Vocabolario detta lingua italiana. Bologna: Zanichelli.
8 The lexicographical representation of gender in the Microsoft Word Thesaurus: Polish Johanna Miecznikowski
1. Introductio n Compared to other computer-mediated dictionaries, which form highly complex hypertextual structures,1 the Microsoft Word Thesaurus is a simple lexicographical tool. Providing a list of synonyms, its primary aim is to help the user find adequate lexical items to express an intended concept, and avoid repetition. Simple as it is, the tool might be of use for producers of Polish texts in a domain of central interest in a gender linguistics perspective, i.e. the domain of personal reference. Finding an adequate nominal expression to refer to persons can be a difficult task in Polish, especially when it comes to reference to females. Polish is a language with a rich system of morphologically marked grammatical gender and a large set of nouns possessing lexical gender. The extent to which these morphological and semantic resources are used to make female gender explicit, however, does not only depend on the referent's sex, but also on the type of noun, the syntactic context, textual constraints (see Dalewska-Gren 1994) and the communicative importance of gender as a category. Moreover, as in all gender languages, nouns with lexical gender lexicalize dominant gender stereotypes, but are also subject to reinterpretation processes, depending on individual speakers' representations of gender roles and on on-going social change. Does the selection of forms and meanings proposed by the Thesaurus help the text producer make his or her choice when looking for an adequate noun to refer to an intended female referent? How does this tool represent gender-related semantic differences and connotations of personal nouns? These questions will be addressed in the main section of this chapter (§4), after a short overview of the Polish gender system (§2) and some remarks on the social context in which gender-related phenomena have been discussed in Polish linguistics (§3).
124
GENDER. LANGUAGE AND NEW LITERACY
2 Gende r in Polish 2.1 Grammatical gender
Polish nouns belong to three main gender classes: masculine, feminine and neuter. Within the masculine gender, three subclasses must be distinguished. The resulting gender system consists of five classes, defined on syntactic grounds2 and corresponding partly to morphological inflection classes:3 feminine (e.g. rgka 'hand'), neuter (e.g. stance 'sun'), masculine personal (e.g. chfopiec 'boy'), masculine animate (e.g. kot 'cat') and masculine inanimate (e.g. dom 'house'). In die singular, the forms of masculine personal and masculine animate nouns coincide. In the plural - more precisely, in the nominative and accusative plural, since in die remaining cases gender distinctions are not expressed at all - feminine, neuter, masculine animate and masculine inanimate nouns fall into one category, which thus enters into opposition widi the category of masculine personal nouns. Thanks to agreement relations with nouns, grammatical gender is marked on a large range of word classes in Polish: demonstrative, possessive and relative pronouns, adjectives, quantifiers, numerals, and die verb. The latter agrees with die subject in the past tense. In the singular, it takes a masculine, feminine or neuter form; in die plural, it shows the characteristic asymmetry of the Polish gender svstem: plural forms in all three persons are either !>masculine personal] or [-masculine personal]. Agreement is normally determined by the grammatical gender of the noun; conjoined feminine and masculine nouns require masculine agreement. 2.2 Lexical gender Lexical gender, as a semantic category, is relevant for animate nouns. When analysing the Thesaurus, we will be interested above all in personal nouns. In Polish, most personal nouns are lexically either masculine or feminine. Lexical and grammatical gender usually coincide, widi a few exceptions (e.g. augmentatives like chiopisko (nt.) 'big man', babsztyl (m.) '(terrible) woman'. A small number of personal nouns have double gender, with partially differing inflectional paradigms (e.g. kaleka (m./f.) 'cripple', 'handicapped person'). Another set of personal nouns, as well as indefinite pronouns, are riot semantically marked for gender (e.g. osoba, (f.) 'person', ktos (m.) 'someone', collectives such as panslwo, (m.) 'Mr and Ms'/'ladies and gentlemen'). On die level of word-formation, lexical gender is either expressed by the stem (e.g. kinship terms such as ciocia (f.) 'aunt', wuj (m.) 'uncle') or marked by derivational suffixes. In die latter case, feminine forms are usually derived from masculine ones. Common derivational suffixes are, among others, -ka (e.g. aktor (m.) - aktorka (f.) 'actor'), -ini/-yni (e.g. daiorca (m.) - dozorczyni (f.), 'caretaker', 'guard'), -ica (e.g. uczen (m.) - uczennica (f.) 'pupil'), -a (e.g. robotnik (m.) - robotnica (f.) 'worker') or -awa (e.g. brat (m.) 'brother' -
POLISH MIECZNIKOWSKI
125
bratawa (f.) 'brother's wife'). Some masculine nouns do not allow derivation (e.g. mecenas (m.) 'lawyer'); the extension of this class is not clearly defined, however (see Miemietz 1993). 2.3 Different types of generic masculines and their relation to lexical gender
In the domain of personal reference, the masculine (personal) category is to be considered the unmarked term of the grammatical masculine - feminine opposition in Polish. As we have seen above, the unmarked status of the masculine appears in word formation ('"formal" (un)markedness', see Linke 2002: 124) or in the agreement patterns of conjoined nouns. Another important indicator of this masculine-feminine asymmetry in Polish is the possibility of using masculine nouns generically ('"distributed" (un)markedness', see ibid.). This possibility implies that a noun's semantic feature [+masculine] may weaken to such an extent that it may be applied, as a hypernym, to female referents. As in other languages, however, this does not prevent a male bias in the interpretation of generic masculines (see Jaworski 1986: 25-48, Linke 2002: 122-3), due to the systematic ambiguity of such expressions between a sex-indifferent and a male-only reading. The contexts in which masculine nouns may be used are not equal for all types of nouns (see Dalewska-Greh 1994) and may be ordered according to a definiteness hierarchy. Most nouns with primary lexical gender, and a restricted set of nouns with derived feminine correspondents, cannot refer to females at all (e.g. syn (m.) 'son'). Another class of masculine nouns may be used in the plural with reference to mixed groups (e.g. polacymnomvA 'the Polish') or non-referentially in the singular (e.g. Kaidy polak ma prawo do opieki lekarza 'everym Polem has the right to medical care'). Finallv, a subclass of masculine nouns can be used also to refer to female individuals and groups consisting only of females. These are nouns that have either no feminine counterpart (e.g. mecenas (m.) 'lawyer') or else have a derived feminine counterpart - mostly in -ka - that language users may want to avoid under certain circumstances because of negative connotations or meaning restrictions. With regard to these nouns, the possibility to partly neutralize the feature [+masculine (personal)] has been extended to predicative uses: (la) (1 b)
ona jest dobrym profesorem she's a goodmjgin$tr professorm sginstr * ona jest dobrym polakiem
*she's a goodmjg.instr Polem.sg,nstr Furthermore, these nouns can even be used referentially for female referents, under the condition that they remain uninflected4 - the deletion of specifically masculine endings mirrors the semantic reinterpretation of the noun. This referential use requires feminine agreement:
126
(2)
GENDER, LANGUAGE AND NEW LITERACY
nasza profesor przyszia °urf nom.sg professor0 came^..
This agreement compensates to a certain degree for the lack of gender marking on the noun (see e.g. Levin-Steinmann 2003: 156-7). Apart from agreeing elements, the referent's sex is often expressed explicitly by means ofpani 'Ms', kobieta 'woman' or a proper name, elements with which the noun enters into a relationship of 'tight apposition' ('enge Apposition', see Wiese 1968: 99): (3)
rozmawiatam z paniq profesor / z profesor Sawickq I talked to (Ms.) professor / to professor Sawicka.
The indedinabilitas of the noun and its agreement patterns in this type of referential use indicates that the 'feminine' form is lexicalized to a certain degree. However, the frequent explicit expression of feminine gender by other elements as well as masculine agreement in predicative uses point to the fact that lexicalization is not complete, and that reference to females in the case of these nouns is still strongly related to the other types of generic masculines mentioned earlier.5 3 Gende r linguistics and the women's movement in Poland In their state-of-the-art of feminist linguistics in Poland, Koniuszaniec & Blaszkowska (2003: 276) observe that '[t]he issue of language and gender has so far failed to attract attention among Polish linguists', showing that the statement made by A. Jaworski as early as 1986, according to which 'in Polish, this subject [the study of sexism in language] has so far received almost no attention' (p. 2), is still valid.6 The complex grammatical gender system in Polish and its relation to sex have been studied for a long time, but not necessarily in a critical perspective. Systematic studies of gender stereotyping in contemporary language use are lacking, and analyses comparing the communicative behaviour of women and men are rare.7 What is particularly significant in the context of the present article, finally, is that scholars of Polish have rarely adopted a gender-related reflexive perspective on linguistic scientific practice itself, observable in its written form in academic publications, lexicography, or grammar and textbook writing (but see Jaworski 1986, Chapters 4 and 5). This situation is partly due to the fact that the feminist movement - which in other countries has contributed to promote studies on gender-related phenomena - has been comparatively weak in Poland in the past four decades (in this period, equality of women and men was officially declared to have been realized). Since 1989, various feminist organizations have developed, and have sought to establish a dialogue with the American and European women's rights movement. Nevertheless, feminist ideas continue to play only a minor role in the wav most Polish women construe their identity.8
POLISH MIECZNIKOWSKI
127
On the other hand, the Polish women's movement has, of course, not confined its attention to purely linguistic matters. Apart from its engagement in a theoretical discussion of women's identity and their role in society, it has participated actively in the debate about abortion legislation in the 1990s and has tackled political and practical issues such as domestic violence, the election of women to the Polish and to the European parliaments, or legal assistance for women. In contrast, sexist language use and its prevention have not been given any special attention (Handke 1994c). Similar priorities have been established by the Polish government with respect to the initiatives in favour of gender mainstreaming that it undertook when preparing Poland's accession to the European Union.9 In its recent report on the situation of women in Poland, the Polish government states that gender-targeted measures primarily concern legislation and the labour market.10 4 The Thesaurus in Microsoft Word 2000 In the broader context sketched above, we would expect neither wellestablished dictionaries published in a traditional format, nor the Polish Word Thesaurus, to be significantly influenced by considerations of political correctness with respect to the lexicographical representation of gender. But this does not mean that there is no point in asking whether innovative lexicographical solutions are to be found in the Thesaurus, something that is far from implausible given its recency, its technical make-up and its intended uses. Is the category of grammatical and lexical gender dealt with in a transparent way? Does the Thesaurus accurately represent the differentiated referential properties of various types of personal nouns in Polish? I will look for answers to these questions in three steps. First, I will describe structural features of the Thesaurus that influence the lexicographical representation of gender (§4.1). In the next section, I will discuss in detail one crucial pair of personal nouns: kobieta (f.) 'woman' and mezczyzna (m.) 'man' (§4.2). The last section is dedicated to the way professional titles are represented in the Thesaurus (§4.3). 4.1 Personal nouns: four types of lexical entries
A lexical entry in the Polish Thesaurus of Word 2000 contains a list of 'meanings', each specified as to part of speech (noun, verb, adjective or other), further meaning-specific 'synonyms' as well as eventual 'similar terms', mostly hypernyms, which are associated with the entry as a whole rather than with specific meanings. Personal nouns are represented in the Thesaurus by four main types of entries: 1 Masculine and feminine forms are both listed and have distinct entries,
128
GENDER, LANGUAGE AND NEW LITERACY
meanings and synonyms matching the headword's lexical gender. An example is the pair matka (f.) 'mother' - ojciec (m.) 'father':11 (4a) (4b)
matka: rodzicielk a (f.) '(female) parent' (mama (f.) 'mom', mamu«a(f.) 'mummy' (diminutive of mama)). ojciec. rodzi c (m.) '(male) parent' (rodziciel (m.) '(male) parent', patriarcha (m.) 'patriarch', protoplasta (m.) 'ancestor'); wynalazca (m.) 'inventor'.
2 Masculine and feminine forms are both listed in the Thesaurus and share some or all of their meanings; the shared meanings and corresponding synonyms consist of both masculine and feminine forms. An example is wnuczka (f.) 'granddaughter' - wnuk (m.) 'grandson', which as their first meaning refer to each other and to the common synonym prawnuk (m.) 'great grandson', and have a second meaning, syn (m.) 'son', associated with a mixed set of masculine and feminine synonyms. 3 Masculine and feminine forms are both listed and share a common entry containing only masculine forms.12 As we will see below (§4.3), typical examples are terms denoting professions such 2&jubilerka (f.) -jubiler (m.) 'jeweller': (5a) (5b)
jubilerka f.: ziotni k (m.) 'goldsmith' (grower (m.) 'engraver', rytoumik (m.) 'wood/copperplate engraver'). jubilerm.: ziotnik (m.) 'goldsmith' (grower (m.) 'engraver', rytoumik (m.) 'wood/copperplate engraver').
4 The Thesaurus lists a masculine or - less often - a feminine form, but no counterpart of the opposite gender, e.g. mecenas (m.) 'lawyer' or szwaczka (f.) 'seamstress'. Entries of type 1 have been chosen by the authors of the Thesaurus for personal nouns whose lexical gender is expressed by the stem. As to pairs related by derivation, we find this type in the domain of interpersonal relations and for nouns referring to traits of character (e.g. przyjaciel (m.) - pryjaciotka (f.) 'friend', kaprys'nik (m.) - kaprysnica (f.) 'capricious person', see §4.2), in the case of noble tides such as krol (m.) 'king' - krolowa (f.) 'queen', in the case of fairytale characters and the like (e.g. czarownik (m.) 'sorcerer' - czarownica (f.) 'witch') and for some professions (e.g. sprzedawca (m.) - sprzedawczyni (f.) 'shop assistant'). Entries of type 1 treat feminine and masculine terms of personal noun pairs as semantically distinct and as clearly marked for gender. They accurately reflect the fact that the masculine terms of the pairs in question cannot be used generically to refer to females and directly provide feminine synonyms for feminine headwords. Feminine synonyms are also provided in the second type of entry, which in the Microsoft Word Thesaurus is adopted occasionally in the field of kinship terms. In contrast to type 1 entries, however, type 2 entries treat masculine and feminine terms as semantically equivalent and only weakly marked for lexical gender, providing also synonyms of a headword's opposite lexical
POLISH MIECZNIKOWSKI
129
gender. The result is clearly odd, though admittedly the risk of confusion for the user is small, since the synonyms' lexical gender is usually marked morphologically. Type 3 entries, an economical lexicographical representation applied to personal nouns related by productive derivational procedures, treat feminine and masculine terms as otherwise semantically equivalent. This may seem advantageous in some cases, since in this kind of entry unwanted gender stereotypes are likely to appear. However, the fact that feminine synonyms have to be retrieved indirectly from the masculine forms provided is a serious shortcoming. The Thesaurus does not indicate whether a feminine equivalent of a given masculine synonym exists or not. In cases where one does exist, the form is not in fact provided, though it may not be fully predictable on the basis of general rules of word formation. For example, the Thesaurus proposes szef(m.) 'boss' as a synonym for dyrektorka (f.) '(female) director', but only a competent speaker of Polish will retrieve the existing feminine form szefawa (f.), which is derived in an atypical way. In cases where a given masculine synonym simply has no direct feminine equivalent in the Polish language, no information is given as to whether the masculine form can be used generically for feminine referents or not. So if stomatolog (m.) fairly obviously is an adequate synonym for dentystka (f.) '(female) dentist', the user is left to decide for herself/himself whether golibroda (m.) 'barber' (lit. 'shaves-beard') can be used as a synonym forfryzjerka (f.) '(female) hairdresser' (it can't). Type 4 entries, finally, are an adequate lexicographical solution for the considerable number of personal nouns that have no generally accepted counterpart of the opposite gender. As we will see below when discussing profession titles (see §4.3), the cases in which the Thesaurus fails to include feminine forms that actually exist in the Polish language are relatively rare. In this respect, the Thesaurus is more complete than conventional Polish dictionaries such as Sobol (1993), in which the majority of existing feminine personal nouns are not mentioned at all, either as headwords or within a lexical entry. 4.2 The crucial pair kobieta - m^zczyzna arid related problems According to Sobol (1993), kobieta (f.) means: (1) dorosty cztowiek plci zenskiej 'adult human being of female gender'; and (2) zona 'wife'; mezczyzna (m.) means dorosty cztowiek plci meskiej 'adult human being of male gender'. In contrast to this succinct description, the Microsoft Word Thesaurus provides 20 meanings for each term as well as 66 additional synonyms for kobieta and 62 additional synonyms for mezczyzna.1* In important respects the two entries are organized symmetricallv. So both contain, in the same order: • what one could term true synonyms, meaning roughly 'woman' or 'man' in different social classes and in different registers;
130
• • • •
GENDER, LANGUAGE AND NEW LITERACY
terms meaning 'old woman'/'old man'. terms used to refer to a woman/man with a specific marital status; terms meaning 'mother'/'father' and 'godmother'/'godfather'; terms used to refer to a young woman/man.
Aside from these sets of comparable synonyms, the entries of kobieta and mezczyzna each contain specific synonyms expressing different types of gender stereotypes. Kobieta is linked, on the one hand, to 11 terms related to childcare and education of small children (mamka, niania, nianka, piastunka; opiekunka, guwernantka, bona, przyzwoitka, wychowawczyni, przedszkolanka, sunetliczanka). On the other hand, kobieta has a bunch of emotively connotated synonyms. The positive ones, listed first, are related to the ideas of beauty, purity and innocence (ptec piekna, biatogtowa, gptabeczka; anielica, hurysa, pieknosc, slicznotka, miss, madonna) or have a diminutive character (kobiecina, babina).1* The negative ones are divided into two sets: kumoszka 'gossip', listed together with the term pnyjaciotka '(girl)friend', and a set of 13 synonyms headed by the term ztosnica 'cross-patch', 'irascible woman'. The latter set, particularly visible because of its size, deserves some special attention. In contrast to the other synonyms specific to kobieta (rather than organized symmetrically within the pair kobieta - mezczyzna), only a minority of the synonyms in this set, denoting mythological female figures, lack a masculine counterpart in the Polish language (hetera 'Hetaera', 'shrew', rnegiera 'Megaera', jedza 'shrew', wiedzma 'witch', ksantypa 'Xanthippe'). The remaining eight synonyms, which the Thesaurus lists first, are derived from masculine forms, with which they form pairs: ztosnik (m.) -ziosnica (f.) 'crosspatch', 'irascible person', kaprysnik (m.) - kaprysnica (f.) 'capricious person', grymasnik (m.) - grymasnica (f.) 'moaner', 'fussy person', sekutnik (m.) -sekutnica (f.) 'unbearable/quarrelsome person', ktotnik (m.) - ktotnica (f.) 'wrangler', 'quarreller', diabd (m.) - diablica (f.) 'devil', piekielnik (m.) - piekielnica (f.) lit. 'inhabitant of hell', herod (m.) 'Herod', 'cruel person' -herodbaba (f.) 'female Herod', 'despotic woman'. That these terms appear as synonyms of kobieta, but not of mezczyzna, is in line with a further peculiarity, namely that they themselves have entries of type 1. Indeed, the entries of the feminine forms consist of the synonym set headed by ziosnica exactly as it appears under the headword kobieta; they then indicate kobieta as a similar term. In contrast, the masculine forms of the above words are either absent in the Thesaurus (ztosnik, ktotnik, piekielnik, herod) or have entirely different entries of type 1, not containing mezczyzna as a similar term (kaprysnik, grymasnik, sekutnik, diabet). With regard to these pairs, it seems fair to criticize the Thesaurus lexicographers for having introduced a gender asymmetry that has no systemic justification. The gender-specific synonyms of mezczyzna form three sets. First, we find terms indicating an unknown individual (indywiduum (nt.), osobnik (m.) '(unknown) individual', ktos (m.) 'someone', iks (m.) 'X', iksinski (m.) 'Xinski', 'mister X'). The association of these expressions with mezczyzna
POLISH MIECZNIKOWSKI
131
reflects a male bias in their interpretation (see Jaworski 1986). This becomes all the more clear if one considers the fact that in the entries of each of these expressions the lexicographers indicate mezczyzna, but not kobieta, as a similar term - despite the fact that they can perfectly well be used to refer to females.15 It would have been preferable to treat these expressions like the similar terms osoba (f.) 'person' or postac (f.) 'person', 'figure', 'character', for which the Thesaurus indicates only gender-neutral similar terms. As to kobieta and mficzyzna, since both nouns can be used to refer to an unknown individual, it would have been adequate to indicate in both entries a set of gender-unspecific words - including osoba and postac. A second small set of synonyms specific to mficzyzna is composed of chiop 'peasant', 'man' and story kon lit. 'old horse'. These terms, especially chtop, are frequently used in colloquial speech. They are listed after the lexemes denoting young men, which indicates perhaps that the lexicographers consider them typical of young people's language. The entry of kobieta does not contain the feminine correspondent of chtop, viz. the very frequent baba (f.) 'woman' - which has no entry of its own, either.16 Might it be that this inadequate - lexicographical decision has been driven by politeness concerns with regard to women, since baba and chtop as synonyms of kobieta and mficzyzna belong to a low stylistic register and baba can have negative connotations? Let's conclude this section by considering the third set of synonyms given for mficzyzna. This entry gives as many as 23 synonyms, grouped around four meanings: narzeczony 'fiance' (kandydat na meza, kawaler, konkurent, starajaty, amant, wielbicieL, adorator, zalotnik), zakochany 'lover' (lit. 'who has fallen in love') (zadurzony, zabujany), sympatia (m.) 'beloved one'17 (bliski sercu, luby, mify, najdroiszy, mitosc, ukochany, umitowany), and wybranek 'bridegroom' (lit. 'the chosen one') (oblubieniec, pan miody). These synonyms cover a large set of roles related to (courtly) love, from the man actively courting a woman (narzeczony etc.) to the man at the mercy of love (zakochany etc.), the man who is the object of a woman's love (sympatia etc.) and finally the man chosen by a woman as a husband-to-be (wybranek etc.). While according to the individual entries of these terms some of these roles, in particular the last two types, are also attributed to women, significantly the corresponding feminine terms do not appear as synonyms of kobieta. There is no straightforward interpretation of this lexicographical decision. On the one hand, it seems as if the authors of the Thesaurus judged men more likely to be lovers and beloved ones than women, and this positive evaluation of men in partnership relations contrasts sharply with the negative evaluation of women in interpersonal relationships in general, as expressed in the previously discussed synonyms of kobieta clustering around the ideas of quarrelsomeness (ztosnica etc.) and passion for gossip (kumoszka). In comparison with the thesauri of other languages, on the other hand, it is noteworthy that the domain of love and partnership is given such importance in the entry of mezczyzna at all unlike certain other features commonly associated with masculinity such as strength, not to mention features belonging to the conceptual field of
132
GENDER, LANGUAGE AND NEW LITERACY
'human being', which are lexicalized separately in Polish by the word cztowiek and its derivatives. 4.3 Some observations on the treatment of professional titles in //# Thesaurus
Professional tides form an unstable area of the lexicon, subject to variation and rapid change, reflecting extralinguistic social, technological and political developments. This is also true with regard to Polish feminine and masculine forms of professional titles: as elsewhere, the growing female presence on the labour market in the past 150 years has made it necessary to develop new linguistic means to refer to women entering professional domains previously reserved for men. Corresponding to more general tendencies in the domain of personal nouns, speakers of Polish have the possibility either to form new feminine forms by derivation or to use existing masculine forms, while having recourse to morphosyntactical strategies (agreement, reduction of inflection) and the use of pani or proper names to make it clear that a female is intended (see §2). The morphological procedures used for the derivation of feminine forms do not show a great amount of variation: in the overwhelming majority of cases, only one suffixed feminine derivative is possible for any given professional tide. However, there is considerable inconsistency as to the set of professional tides that form feminine variants and as to eventual semantic differences between feminine and masculine forms. In a feminist perspective, it has been recommended that existing derived feminine forms be used consistently and innovative forms such as psycholozka (f.) (derived from psycholog (m.)) be promoted (Miemietz 1993, 1996; Koniuszaniec & Blaszkowska 2003), in the hope that the negative connotations and restrictions of meaning affecting many derived feminine forms will disappear thanks to increased frequency of use. As mentioned earlier (see §3), these recommendations are quite recent. The question of whether the consistent promotion of feminine derivations is desirable in all respects in Polish, and whether in themselves they are likely to bring about semantic changes favourable to the image of women has not been debated in depth. Now how does the Microsoft Word Thesaurus represent innovative feminine forms, and what image does it give of women's and men's roles in the professional world? In what follows, die first of these two questions will be discussed by examining a corpus of professional titles and agentives, retrieved by Miemietz (1993) in 1,254 job announcements published in the Polish newspaper Zycie Warszamy in 1992. The second question will be addressed on the basis of selected Thesaurus entries of type 1, i.e. proposing separate descriptions of feminine and masculine forms. The above-mentioned corpus consists of 177 professional tides,18 of which 80 are not listed in the Thesaurus. Of die remaining 97 nouns, 69 have entries of type 3 according to the typology proposed under §4.1, 9 have entries of type 1 and 19 have entries of type 4 (of which 6 feminine and 13 masculine ones). This overview shows that professional titles are most often presented in
POLISH MIECZMKOWKI
133
type 3 entries. A simple morphological analysis reveals that type 3 entries have been chosen mainly on formal grounds. This set of entries consists in fact of a set of pairs with feminine forms ending in -ka as well as pairs of the type -ca (m.) -czyni (f.), resulting from suffixation procedures that are treated as fully regular and productive. This lexicographical strategy copes with variation in language use by automatically applying word formation rules. It makes the Thesaurus innovative in some respects (but even in the Thesaurus we don't find pairs of the type psychobg (m.) - psycholozka (f.)), while producing a certain amount of over-generalization. Thus the Thesaurus contains feminine forms that are neither attested in Miemietz' (1993) corpus nor listed in the dictionary she cites, Szymczak (1978-81). An example is kierowczyni (f.)' (female) driver'. Others are pairs of nouns ending in -tor (m.)/-torka (f.), nouns ending in -yk (m.)/-yczka (f.) - or nouns ending in -rz/-rka, which the Thesaurus lists without distinguishing between common forms such as lekarka (f.) '(female) doctor' and (still?) uncommon forms such as tokarka (f.) '(female) turner' or stolarka (f.) '(female) carpenter', 'joiner'. In some cases, the female forms in question have primarily a non-personal meaning, which the Thesaurus does not indicate (e.g. tokarka 'lathe', stolarka 'carpentry', 'woodwork'). On the other hand, the rule-based approach adopted by the lexicographers fails to account for a certain number of atypical derivations. It is comprehensible that the Thesaurus has only masculine entries (type 4) for ciesla (m.) 'carpenter', poligraf (m.) 'typographer', 'printer', prezes (m.) 'president' and a set of nouns ending in -log (even if this case is debatable) and -owiec. But, quite incorrectly, it also excludes a set of atypically derived, but quite common nouns: mistrzyni (f.) '(female) master' (< mistrz m.), cudzoziemka '(female) foreigner' (< cudzoziemiec (m.)), and uczennica (f.) '(female) pupil' (< uczen (m.)). Other examples in -ica or -nic-a such as the nouns treated under §4.2 (ztosnica etc.) show, by the way, that the Thesaurus does not in fact contain any type 3 entries for pairs in -0/-ica or -nik/-nic-a. Having been separated from their masculine correspondents, some forms in -ica have apparently been forgotten by the lexicographers. To sum up, the vast majority of feminine professional titles are represented in the Thesaurus in the form of type 3 entries, as a result of a mechanical application of derivational rules. The advantage of this procedure, namely that it at last ensures the inclusion of feminine forms also in professional domains that are dominated by men (e.g. instalatorka), is outweighed not only by the general weaknesses of this intrinsically asymmetrical type of entry (see §4.1), but also by the omission of forms (e.g. mistrzyni) and meanings (e.g. tokarka 'lathe') that deviate from the rule applied. Now what about those feminine titles that are treated in specific entries, independently of an eventual masculine counterpart? A first group of terms consists of those terms in the Miemietz (1993) sample that appear only as feminine forms in the Thesaurus (entries of type 4): bufetowa (f.) 'woman serving at a buffet', gosposia (f.) 'housekeeper', higienistka (f.) '(female) hygienist', hostessa (f.) 'hostess', krawcowa (f.)
134
GENDER, LANGUAGE AND NEW LITERACY
'dressmaker', szwaczka (f.) 'seamstress'. Some of them have no masculine correspondents in the Polish language (gosposia, hostesses szwaczka). For others, there are corresponding masculine forms (bufetowy (m.), higienista (m.), krawiec (m.)); but also in Miemietz' corpus these lexical items appear only in their feminine form, and they denote professions typically associated with women. In these cases, the decision to explicitly list the feminine forms and to exclude the masculine ones is not motivated by the language system, but rather is related to frequency in language use and reproduces social stereotypes. With regard to those titles that have two separate entries, a feminine and a masculine one (type 1), the synonyms provided by the Thesaurus reflect gender asymmetries that are lexicalized in the Polish language, but also add unnecessary stereotypes, considerably compromising the usefulness of the Thesaurus. It may be acceptable, on the one hand, that the Thesaurus indicates items such as siostra 'sister', potoina 'midwife' and akuszerka 'midwife' as synonyms for pielegniarka (f.) 'nurse' but not for its masculine equivalent pielfgniarz (m.); these terms do indeed exist only in the feminine form in Polish and are necessarily applied to female referents. But pairs such as kucharz (m.) - kucharka (f.) 'cook' are more problematic. The Thesaurus indicates kuchmistn (m.) 'master cook' and szef kuchni (m.) 'chef as synonyms for kucharz, whereas for kucharka it suggests podkuchenna '(female) cook's help', pomoc kuchenna (f.) 'kitchen help', kuchcik (m.) 'kitchen help', 'cook's apprentice' and pomywacz (m.) 'dish washer'. It is true that women are more often employed as cooks in a subordinate position and that, as a consequence, the feminine term may thus be seen as reflecting this situation. To avoid this connotation of low prestige, when referring to women, the language user has the option of using kucharz (m.) as a generic masculine. But in contrast to titles such as doktoror inzynier, the feminine form will often be preferred, and any negative connotations will simply be neutralized by the context (e.g. Ona jest zawodowq/doskonatq kucharkq 'She is a professional/ excellent cook'). The synonyms for kucharka provided by the Thesaurus clearly do not account for this type of use. To retrieve adequate feminine synonyms, the user is constrained to look up the masculine entry and to form feminine variants of the synonyms provided there (kuchmistrzyni (f.) '(female) master cook' and szefowa kuchni (f.) '(female) chef are attested lexical items in Polish). That the Microsoft Word user will make this laborious detour is all the more unlikely in the case of pairs such as tancerz (m.) tancerka (f.) 'dancer', which exclude the masculine form with reference to women.19 He or she will therefore stick with danserka (f.) '(female) dancer' and girisa (f.) 'chorus girl' as possible alternatives of tancerka, without noticing the useful items baletmistn (m.) 'ballet-master' (f. baletmistrzyni) and cAoreogra/'choreographer', indicated as synonyms of tancerz (m). 5 Conclusio n The Polish version of the Microsoft Word Thesaurus may be a useful memory aid in many respects. But when it comes to reference to females by means of
POLISH MIECZNIKOWSKI
135
personal nouns, it is hardly of any help to its user. A large proportion of lexically female personal nouns turn out to owe their inclusion only to the mechanical (and therefore error-prone) application of derivational rules, and have only lexically male synonyms, irrespective of the latter's referential properties. The treatment of pairs of personal nouns in separate lexical entries helps avoid these shortcomings and leads to satisfactory results in some cases. However, the analyses proposed in this chapter have shown that this strategy is no guarantee for a careful choice of synonyms either. The lexicographers often exaggerate the semantic differences between corresponding items with female and male gender, relating them to completely separate sets of synonyms. These sets, accessible from any of their members, usually form specifically male or female semantic fields. The resulting images of gender roles might correspond to current linguistic routines in modern Polish society, as Miemietz' (1993) corpus of job advertisements suggests at least for the domain of professional titles. Just as at the level of word formation rules, however, reproduction of what is (stereo) typical does not necessarily support the user in the creative process of text production. Notes 1 A good collection of links to available on-line dictionaries for Polish can be found on the website of the Ruhr-Universitat Bochum (, consulted in January 2005). Among others, there exists an on-line version of the monolingual dictionary edited by Szymczak (197&-81) and revised in 1992 and 1995 (, consulted in January 2005), which also offers the possibility of searching the corpus on which the dictionary is based. 2 A noun's grammatical gender is defined by the pattern of inflectional forms it selects in the sentence by virtue of agreement relations (Wiese 1968; Wrobel 2001). 3 If personal pluraKa tantum (collectives such as panstwo pi. 'Mr and Ms', 'ladies and gentlemen') and non personal pluralia tantum (e.g. saniepl. 'sleigh') are counted as categories of their own, there are even seven gender classes (see Wrobel 2001). 4 This condition seems not to apply necessarily in the case of plural referential uses. With regard to this context - which can be considered less definite than a referential singular context, but still more definite than the predicative use - existing descriptions do not agree as to the acceptability of inflection (Miemietz 1993: 22). 5 This interpretation agrees with the description given by Barbel Miemietz, who discusses in detail this type of nouns and their treatment in the literature (1993: 16-25). See also the interesting proposal made by van Leeuwen-Turnovr.ova (2003), according to which the resistance to derivation (of feminine forms) and inflection corresponds to a tendency of the Polish language to retard linguistic integration of the Other, appearing also with regard to words of foreign origin. This tendency is found in Russian as well, but not in Czech. In contrast, Handke (1994b) relates the uninflectedness of feminine forms to a more general tendency of Polish to reduce nominal inflection. 6 Useful bibliographical overviews can also be found in Miemietz (1993,1996) and inBlaszkowska(2002).
136
GENDER, LANGUAGE AND NEW LITERACY
7 The latter aspect has been treated by Kwiryna Handke, who arrives at the following conclusions: 1) The intellectual sphere dominates in the behaviour of men, who are usually more concerned with the essence of things than with their secondary properties or 'decorative frame', which in turn women prefer. So when buying a car, for instance, men are primarily interested in technical and functional values, and women - above all in external values, in particular in colour. ... 2) The psychical sphere dominates in the behaviour of women, who are more subject to emotions than men. (1994a: 26, my translation) Further sociolinguistic studies on Polish are needed to assess the way the author's analysis reproduces widespread stereotypes on female and male behaviour. 8 According to the historical overview proposed by Titkow (1996), gender roles in Polish society have been influenced by, among others, the Roman Catholic Church, by the ethos of aristocracy (granting women high status and the competence to be administrator of a family's estate), by women's participation in the struggle for national independence, above all as mothers maintaining and passing on the national cultural and linguistic heritage (hence the stereotype of matka polka, the 'Polish mother'), and by the influx of women into the labour market since the second half of the nineteenth century. 9 Among other things, a Plenipotentiary for Gender Equality has been appointed, see , consulted in January 2005. 10 Rzqd o sytuacji kobiet w Polsce, see . consulted in January 2005. 11 In this and in the following examples the meanings are in bold, whereas the items given for each meaning stand in brackets, in italics. 12 Double gender nouns such as sierota (f./m.) 'orphan' can in general be ranged in this category; an exception is sympatia (f.) 'girlfriend', 'fiancee' - sympatia (m.) 'boyfriend', 'fiance', which has two different entries (type 1). 13 Let us note that the Thesaurus does not contain the adjectives ifski 'feminine', 'female', hobiecy 'womanly', 'feminine' and meski 'male', 'masculine'. We find only kobifcoasan adverb, with a set of adverbial synonyms centred around the restricted - and stereotypical - meaning of female attractiveness (podqgajqco, ponftnie, apetycznie, mile). 14 There is a similar asymmetry in the pair matka 'mother' - ojriec 'father' described mficzyna earlier. Whereas the affectively connotated forms mama and mamusia appear as synonyms of matka, the words ojaulek, tata and tatus (which have an entry in the Thesaurus) are not indicated as synonyms of ojciec. 15 Most misleading is the entry of iksinski, which is of type 3 (see §4.1): mezayzna ends up being a similar term not only of iksinski (m.), but also of its feminine equivalent iksinska. 16 Of the various derivatives of baba (see Kryk-Kastovsky 2000), we find in the Thesaurus the diminutive-like babinaand herod-baba 'female Herod'. 17 'Person of the opposite sex that one likes or with whom one is in love' (Sobol 1993, my translation). 18 Part of them appeared only as masculine forms, others only as feminine forms or in both variants. 19 See e.g. Dalewska-Gren (1994: 86).
POLISH MIECZNIKOWSK1
137
References Blaszkowska, Hanka (2002), Weibtiche Personenbezeichnungen im Deutschen und Polnischen: Eine hontrastive Studie aus der Sicht derfeministischen Sprachkritik (Nazwy osobowe w odniesimiu do kobiet w jfzyku niemieckim i polskim z perspektywy feministycznej krytyki jezykowej). Uniwersytet im. Adama Mickiewicza, Wydzial Neofilologii. Dalewska-Greh, Hanna (1994), 'Jezyk a plec - problemy z uzyciem rzeczownikow osobowych', in Kwiryna Handke & Hanna Dalewska-Greh (eds), Polszczyzna a/i Polacy u schylku XX wieku. Warsaw: SOW, pp. 85-96. Handke, Kwiryna (1994a), 'Jezyk a determinanty pki', in Janusz Anusiewicz & Kwiryna Handke (eds), Plecwjezyku i kulturze. Jezyk a kultura 9. Wroclaw: Wydawnictwo 'Wiedza o kulturze' Fundacji Uniwersytetu Wroclawskiego, pp. 15-31. Handke, Kwiryna (1994b), 'Przyczyny ograniczenia fleksji nominalnej we wspokzesnej polszczyznie', in Kwiryna Handke & Hanna Dalewska-Gren (eds), Polszczyzna a/i Polacy u schyiku XX wieku. Warsaw: SOW, pp. 74-S3. Handke, Kwiryna (1994c), 'Stosunek Polek do nurtow feministycznych i jezyka', in Kwiryna Handke &: Hanna Dalewska-Gren (eds), Polszczyzna a/i Polacy u schylku XX wieku. Warsaw: SOW, pp. 342-59. Jaworski, Adam (1986), A Linguistic Picture of Women's Position in Society: A Polish-English Contrastive Study. Frankfurt/M.: Peter Lang. Koniuszaniec, Gabriela & Blaszkowska, Hanka (2003), 'Language and gender in Polish', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins, pp. 259-85. Kryk-Kastovsky, Barbara (2000), 'Norm vs use: on gender in Polish', in Barbara Unterbeck & Matti Rissanen (eds), Gender in Grammar and Cognition. Berlin: de Gruyter, pp. 729-47. Levin-Steinmann, Anke (2003), 'Weltbild - Geschlecht - Sprache', in Jifina van Leeuwen-Turnovcova & Uta Rohrborn (eds), Beitrdge des Gender-Blocks zum XIII: Internationalen Slavistenkongress in Ljubljana 15.-21. August 2003. Munich: Otto Sagner, pp. 147-67. Linke, Angelika (2002), 'Das Wort in der feministichen Sprachreflexion. Eine Ubersicht', in D. Alan Cruse; Franz Hundsnurscher; Michael Job & Peter Rolf Lutzeier (eds), Lexikologie/Lexicology: Ein internationals Handbuch zur Natur und Struktur von Wortern und Wortschatuen. HSK 21.1. Berlin: de Gruyter, pp. 121-8. Miemietz, Barbel (1993), Motivation zur Motion: Zur Bezeichnung von Frauen durchFeminina und Maskulina im Polnischen. Frankfurt: Peter Lang. Miemietz, Barbel (1996), 'Ein marginales Phanomen? Zum Ausdruck des Sexus im Polnischen', in Walter Koschmal (ed.), Die Frau in der polnischen Gegenwartskultur. Cologne: Bohlau, pp. 125-83. Sobol, Elzbieta (ed.) (1993), Maty stownik jezyka polskiego PWN. Warsaw: Wydawnictwo Naukowe PWN. Szymczak, Mieczyslaw (ed.) (1978-81), Slownik jezyka polskiego (3 vols). Warsaw: Wydawnictwo Naukowe PWN. Titkow, Anna (1996), 'Frauen unter Druck?', in Walter Koschmal (ed.), Die Frau in der polnischen Gegenwartskultur. Cologne: Bohlau, pp. 31-67. van Leeuwen-Turnovcova, Jifina (2003), 'Genderspezifische Aspekte der Uberwindung von Diglossie und normative Traditionen im slawischen Areal', in Jifina van Leeuwen-Turnovcova & Uta Rohrborn (eds), Beitrdge des Gender-Blocks zum XIII: Internationalen Slavistenkongress in Ljubljana 15.-21. August 2003. Munich: Otto Sagner, pp. 45-74.
138
GENDER, LANGUAGE AND NEW LITERACY
Wiese, Eva (1968), 'Zu einigen Problemen der Kategorie des Genus von Personenbezeichnungen in der polnischen Sprache der Gegenwart', in Rainer Eckert, Ernst Eichler, Miroslav Komarek & Rudolf Zimek (eds), Zur grammatikalischen und lexikalischen Struktur der slawischen Gegenwartssprachen. Halle: Niemeyer, pp. 95-104. Wrobel, Henryk (2001), Gramatyka jfzyka polskiego. Krakow: Spolka wydawnicza 'Od
nowa Links http://sjp.pwn.pl/ http://www.oska.org.pl/articles.php?id=153 http://www.rownystatus.gov.pl http://www.ruhr-uni-bochum.de/wwwoerterbuch/polnisch/polnisch.htmtm
9 The perspective on gender in the Portuguese Word Thesaurus Maria Aldina Marques
1 Outlin e of gender-marking and gender-construction in Portuguese 1.1 Categorization of noun gender in Portuguese - natural gender and grammatical gender: from semantics to grammar
Existing since time immemorial in different societies, languages are not only extremely influential, but, more precisely, crystallize the way the speaker perceives the world. The 'sexed' feature of life is arrived at on the basis of the categorization of grammatical gender, despite the formal nature of this category. In Portuguese, gender as a procedure of grammatical categorization is characteristic of nouns, adjectives, participle forms, determiners and pronouns, in spite of its different morphological and syntactic functioning. Traditional Grammar has circumscribed, and moreover is still circumscribing the analysis of grammatical gender in Portuguese. A logical perspective of grammar and a representation of grammatical gender based on semantic features support the following definition by a nineteenth-century Portuguese grammarian: e como todos os animaes naturalmente se distinguem em duas Classes ou Generos segundo os dous sexos de macho e de femea: os Grammaticos puzerao os nomes dos primeiros na Classe ou Genero Masculine, e os dos segundos no Feminino. Estas sao as classes naturaes em que entrao so os animaes. Todos os mais seres que nao tern sexo algum, deveriao ser arranjados na Classe, ou Genero Neutro, isto he, formarem todos uma terceira Classe ... Das Classes naturaes, a significacao mesma determinava o seu genero; das arbitrarias, so a terminac.ao dos nomes analoga a dos primeiros, he que podia determinar. Daqui a divisao do genero dos Nomes, ou pcla sua significac.ao, ou pela sua terminac.ao. (Jeronimo Scares Barhosa 1822: 123-4)'
There are thus two criteria for analysing grammatical gender in Portuguese. The first is a semantic one, based on male/female distinction, and the second is morphological, based on noun termination (-o/-a, masculine and feminine gender-marking suffixes, respectively)."
140
GENDER, LANGUAGE AND NEW LITERACY
In fact, Traditional Grammar has misunderstood the question, in as much as it makes an analysis that amalgamates semantic and syntactic features, explains syntactic rules via semantic properties, and defines Portuguese noun variation as an inflectional category of gender. 1.2 Portuguese noun gender: a syntagmatic category
Some Portuguese grammars present an inflectional category of gender based on the noun variation as presented in Table 9.1. However, this morphemic variation is a redundant morpho-lexical process, overestimated by Traditional Grammar. In fact, this is a function that characterizes animate but also inanimate beings, and must be interpreted as a derivational process, referring to several aspects of 'reality'. Even though some of those aspects are related to sex (1), and some others are not (2), the morphological mechanism is the same: (1) (2)
o menino 'the boy' - a menina 'the girl' ogato 'the male cat' - a gata 'the female cat' o barco 'the boat' - a barca 'the bark' ofruto '[a specific] fruit' - afruta 'the fruit [in general]' o mate 'the brushwood' - a mala 'the wood'
The persistence of the traditional interpretation enhances theoretical confusion, and, as such, also a pragmatic one (e.g. the real effects of a speaker's usage), between feminine and masculine grammatical gender and female and male natural gender, the grammatical category being perceived as a semantic class. More recent analyses have reconsidered the question of the relationship between grammatical gender and natural gender, stressing the divergent nature of the categories under consideration. In Portuguese, all nouns have grammatical gender; they necessarily belong to either the feminine or to the masculine class. This sub-classification is a lexical feature, but its linguistic manifestation is syntagmatic. Gender (of nouns) is, first of all and in a systematic way, a syntagmatic process, marked by determiners (Table 9.2). Considering that the Portuguese language condenses the three categories of Latin into two categories, neuter Latin nouns have been included in the formal masculine and feminine classes, in an arbitrary way. Beyond this,
Table 9.1 Morphemic variation in masculine and feminine nouns in Portuguese Feminine
Masculine
a gata 'the female cat' a menina 'the girl'
o goto 'the male cat' o menino 'the boy'
PORTUGUESE MARQUES
141
Table 9.2 Gender as a syntagmatic process marked by determiners Feminine
Masculine
a tribo 'the tribe' apedro'thestone' a testemunha 'the witness' a ponte 'the bridge' a^or'the flower' a fet 'the law'
o liwo 'the book' o planeta 'the planet' o cinema 'the cinema' odente 'the tooth' o girassol 'the sunflower o amor 'the love'
throughout the ages, some words have changed from one class to anodier, without any known explanation (for example, mar- 'sea' and fim - 'end', initially feminine, are now masculine). Neologisms are added (sometimes with 'fluctuations') into one of the two categories. 1.3 Grammatical gender and natural gender
Using verbal language is a basic act of social life; thus making language a cultural phenomenon. So, assuming that life is 'sexed', verbal language has certain linguistic devices with which to mark natural gender. That is to sav that language represents, in different linguistic forms, the conceptualization of this 'fragment' of the world. In Portuguese, there are several mechanisms to signal the reference to female and male: 1 Lexical derivation: o galo die cock - a galinha die hen o actor the actor - a actriz the actress 2 Lexical variation: o homem die man - a mulher die woman o boi die bull - a vaca die cow 3 Compound nouns3 a dguia macho die male eagle - a dguiajemea die female eagle The forms listed above as inflectional (goto - gata) are simply a manifestation of the derivational mechanism. 1.4 Formal categories and masculine form functions
The structural approach analyses the feminine/masculine category relationship as a structural one between marked and unmarked terms, respectively:
142
GENDER, LANGUAGE AND NEW LITERACY
o termo marcado 'feminine' sempre [corrigido em nota de rodape para 'normalmente'] 'ser da especie feminina', o termo nao marcado, masculine, significa ora 'ser da especie masculina', ora o conceito generico ou o conjunto universal dos seres nele abrangidos, sem distincao de especies, ora uma coleccao particular de membros de uma e outra especie. (Herculano de Carvalho 1984: 637-8)"
As with other languages, in Portuguese the 'masculine' form is used in specific contexts, with a denominated generic value (1), in which we can include the so-called 'masculine plural' (2): (3a) (3b) (4a) (4b)
O gato e urn animal domestico. The cat is a domestic animal. Os gat os sdo animais domesticos. Cats are domestic animals. Eu encontreiafamilia Silva (f., sg.). Eles (m., pi.) sdo muitosimpdticos. I met the Silva family. They are very nice. Vi a Ana (f.), a Maria (f.) e o Pedro (m.). Eles (m., pi.) foram ao cinema. I saw Anna, Mary and Peter. They went to the cinema.
In spite of the formal coincidence (syntactic agreement, in both cases, is made with adjectives and determiners in the masculine), this must be considered as a different use. The proposal is that, in this case, there would be a three-term semantic relationship here: two co-hyponyms (gato,- goto) and a hypernym (gato2), with neutralization of the semantic specification of 'sexed' that exists in the pair. It is a lexically non-specific term. Consequently, between the terms of the pair gato- gata there is a formal asymmetry, but not a semantic asymmetry. The same process of neutralization of gender can be seen in other kinds of lexical constructions. That is the case of serhumano 'human being' that refers to mulher'woman' and homem 'man'. This kind of semantic neutralization is marked in the lexicon itself, with the 'sexless'5 representation of 'sexed' beings: a crianfa 'the (f.) child' o bebe'the (m.) baby' apessoa 'the (f.) person'
The co-existence of these two mechanisms gives the 'masculine' form the statute of an unmarked term, present, for instance, in the stereotyped order of reference in grammars that refer to this subject (first, the 'masculine' and then the 'feminine'), and in the lemmatization.
PORTUGUESE MARQUES
143
2 Socia l gender: from mulheres-a-dias ('dailies [women]') to Homens de Estado ('Statesmen') The creation of lexical neologisms accompanies social evolution in general and that of institutions in particular. Since the Revolution of the 25 April 1975, which established a democratic regime in Portugal, Portuguese society has been characterized by transformations in which the opening of work opportunities in the labour sector, in politics and in every social domain6 to women is striking. The new social interaction generates linguistic changes, and the Portuguese language reflects this evolution. It especially includes new terms for new specifically female professions. Fields in which women had no representation, such as in government and law, now encompass designations (and roles, actually performed by women) such as a primeiraministra 'the (female) Prime minister', ajuiza 'the (female) judge', a mulher de Estado 'the stateswoman'. The traditional separation of jobs has narrowed; women have entered a labour universe previously almost completely masculine. The opposite is not so noticeable. There are no homens-a-dias ('dailies [men]'), but there are mulheres de Estado7 'stateswomen'. At the same time, there is an euphemization in the naming of less prestigious (feminine) jobs, from mulher-a-dias 'daily' and criada 'maid' to empregada domestica 'domestic [female] worker'. The inequalities between women's and men's roles can be illustrated by full names and address forms. Indeed, full names and treatment forms constitute a special area in the representation of institutional and social relationships between men and women. In spite of possible variations, in Portuguese, a full name prototypically follows this sequence: First name + mother's family name (usually her father's family name) + father's family name (the father's family name, with very few exceptions): Ex. Manuel Ribeiro da Silva Rita Ribeiro da Silva
Prototypical uses:* men are known by their last name (their father's family name): Sr. (= senhor) Silva 'Mr Silva', whereas women are known by their first name: sr" (= senhora) / D.a (= dona) '±Mrs Rita'. Traditionally, marriage did not bring name changes with it. That is to say, women did not adopt their husbands' family name. Nowadays, new habits prevail and a great number of women adopt their husband's family name. Portuguese law allows three possibilities: to adopt the husband's family name, to adopt the wife's family name, or each to maintain their own family name. The male line prevails over the female although it is not as systematic as it is in other European cultures.
144
GENDER, LANGUAGE AND NEW LITERACY
3 Dat a analysis 3.1 Relevant characteristics 3.1.1 The Thesaurus - organization We have worked on the Portuguese version of Word9, Microsoft Windows XP. In general terms, there are two significant points to be made about the organization of this version of the Thesaurus. First, the words selected as 'synonyms' are listed in alphabetical order; the word chosen as 'meaning' frequendy corresponds to the first 'synonym'. Second, the 'meanings' are categorized (noun, adjective, verb), but there is a lack of clarity regarding those syntactic categories. There are words listed as adjectives, for example, when they are nouns (see the case of the noun aluno 'student', marked as an adjective). There are other gross mistakes that allow, for instance, die amalgamation of the noun governante (o/a governante, 'the member of a government' (m./f.)), with the noun governanta (from the pair a governanta - o mordomo 'the housekeeper - the butler'). Finally, a further aspect of these mistakes has to do with a clearly erroneous notion of the lexicon as a nomenclature that generates undue associations. In fact, the synonymic relationship is considered as one of a complete identity. This authorizes the reasoning a = b = c, then, a = c.
3.1.2 Ward Thesaurus inconsistencies We can list additional misuses, namely regarding the usage without explicit mention of: 1 Slang: penico 'chamber pot' is a synonym for funciondrio 'servant', 'employee', etc. 2 Popular register: so, corruption of the word senhor 'Mr'. 3 Familiar register: enforcado 'hanged' for casado 'married man'. 4 Old words from Brazilian Portuguese in a European Portuguese dictionarv: sid, iaid, sinhd (familiar terms of address for women from slavery times). 5 Metaphors: revolver 'pistol' for amigo 'friend' (a metaphor from Western movies). 6 Old literary terms: dulcineia for namarada 'girl-friend'. There are also some other kinds of mistakes: 1 Inventories based on graphic similarities: bardo 'baron' for the term vardo 'male','man [of respectability]'. 2 Svntactic inconsistencies: amiga 'friend' (f.) is a noun, but only verbs are suggested as synonyms.
PORTUGUESE MARQUES
145
3 Absence of the most frequent meanings; for example, the pair filha-filho
'daughter' - 'son' does not include the term descendente 'descendant'. 3.1.3 Quantitative analysis Some 31 pairs were listed within three areas: family; social relationships; economic and professional activities. The amount of information is clearly asymmetric and there is a prevalence of synonyms for the masculine in relation to the feminine, respectively 353 as against 230. Only eight pairs were found in which there is an equal number of synonyms: noiva - noivo 'bride' - 'bridegroom'; soUeira - soUeiro 'spinster' 'bachelor'; casada - casado 'married woman' - 'married man'; directora - director 'female director' - 'male director'; funciondria - funciondrio 'female employee' - 'male employee'; criada - criado 'woman-servant' - 'manservant'; professora -professor'female teacher' - 'male teacher'; medico, - medico 'female doctor' - 'male doctor'. Except in one case (criada - criado}, the list of feminine terms is literally copied from the masculine in a rough extension process in which simple syntactic agreement is not respected. Using the Word Thesaurus instructions, an automatic substitution in a given sentence, such as: (5a)
a professora chegou Thef teacher f arrived.
would have the following syntactically incorrect result: (5b)
a doutor chegou
Thef doctor,, , arrived.
The number of synonyms for the feminine term prevails over the masculine only in three pairs of words: esposa - marido 'wife' - 'husband'; filha - filho 'daughter' - 'son'; and governanta - governante 'housekeeper/female member of a government' - 'male member of a government'. In two cases, meanings and synonyms for the other term of the pair were not found: juiza 'female judge' (only the masculine form occurs) and prostituto 'male prostitute' (only the feminine form occurs). 3.2 Ideological representation of 'women' and 'men'
3.2.1 The pair mulher- homem 'woman' - 'man' The analysis of the pair mulher - homem 'woman' - 'man' and their synonyms allows us to consider a series of semantic relationships which reveals completely different fields for each member of the pair:J
146
(6a)
(6b)
GENDER, LANGUAGE AND NEW LITERACY
mulher: damn (adj.) (dama 'lady', 'concubine', dona 'Mrs' (title), 'mistress', 'owner', esposa 'wife', Jemea 'female', senhora 'lady'). dama: actri z (subs.) (actriz, amiga, concubina, Jemea, dona, madona, manceba, marafona, meretriz, mulher, prostituta, rainha, senhora). dona: Not found (Not found). esposa: companheir a (subs.) (companheira, conjuge, consorte, costela, dona, metade, obrigofdo, mulher, patrva, senhora, velha); abraca r (verbo); marido (subs). Jemea: barrega (subs.) (barregd, colchete, concubina, dama, donzela, fechadura, mde, manceba, meretriz, mofa, mulher, prostituta). senhora: senhor a (subs.) (senhora, ama, dama, esposa, iaid, madama, madona, matrona, sia, sinhd); amo (adj). homem: alguem (subs.) (alguem 'someone', esposo 'husband', humanidade 'humanity', humano 'human', individuo 'individual', mortal'mortal', operdrio 'workman', pessoa 'person', sujeito 'fellow', vardo 'male, man [of respectability]'). alguem: algum (algum, certo, nenhum, qualquer, tat), esposa marid o (subs.) (marido, conjuge, consorte, dono, homem, patrdo, vardo); abracar (verbo). humanidade. afabilidad e (subs.) (afabilidade, agrado, benevolenda, benignidade, bondade, brandura, caridade, carinho, clemencia, compaixdo, filantropia, gente, mundo, sensibilidade, universo, urbanidade, belas-artes). humano: afavel (adj). (afdvel, beneficente, benfazejo, benigno, bom, bondoso, caridoso, clemente, compassivo, humanitdrio, indulgente, manso, natural, pacato, pacifico, piedoso, sensivel). individuo: indivisive l (adj.) (indivisivel, indiviso, insepardvel); alma , (adj.).
mortak caduc o (caduco, capital, cruel, deleterio, efemero, enfadonho, excessivo, fastidioso, fatal, figadal, Jrdgil, funereo, funesto, insuportdvel, Idnguido, letal, moribundo, mortifo, morttfero, passageiro, perecivel, perigoso, profundo, transitorio)', homem (adj.) Basically die first group of synonyms determines two spheres for 'woman': • a family sphere: esposa 'wife' (whose synonyms are: costela 'rib', metade 'half), obrigafdo 'obligation', velha 'old woman' and mde 'modier'. The word senhora 'lady', 'mistress', with synonyms such as sinhd, iaid (discussed previously), etc., is also restricted to the family area. • a sexual sphere: Jemea 'female', prostituta 'prostitute'. Besides, the list of synonyms which is available in the Thesaurus for each of the synonyms of mulher (dama, dona, esposa, Jemea, senhora) does not introduce any odier semantic field. It only reinforces these initial domains. Derogatory terms, slang and popular words, are only found in the construction of the domain of'woman' (mulher. costela, metade, obtigafdo, patroa, velha, barregd).
PORTUGUESE MARQUES
147
The human dimension is not presented in the same way for 'woman' mulher and 'man' homem. For example, the relationships between homem 'man', pessoa 'person' and mulher 'woman' are presented differently. Homem has pessoa as a synonym and vice-versa; but mulher does not have pessoa as a synonym - only pessoa selects mulher. The relationship established is direct for the first pair (homem - pessoa) and indirect for the second one (pessoa mulher): homem 'man' <-> pessoa 'person' —» mulher 'woman' The pairs masculino - feminino ('masculine' - 'feminine') andjemea- macho reinforce the previous analysis. (7a) (7b)
(8a) (8b)
feminino: feme o (femeo, femeal, femeeiro, feminil, mulher, mulherengp, mulheril). masculino: varonil (adj.) (varonil, animoso, corajoso, denodado, destemido, energico, forte, heroico, ilustre, incisivo, inclito, intrepido, mdsculo, robusto, valoroso, viril). Jemea: barrega (subs.) (barregd, colchete, concubina, dama, donzela, fechadura, mde, manceba, meretriz, mofa, mulher, prostituta). macho: amant e (amante, amdsio, colchete, corajoso, destemido, esforgado, forte, grilhdo, masculino, mdsculo, mulo, robusto, valentdo, valente, vardo, varonil, vigoroso).
On the other hand, the setting for homem 'man' (including also what we have defined as a generic use) is extremely positive and wide (familiar, social, sexual, professional, moral, spiritual spheres). In this representation, homem 'man' is marked by his strength, energy, respectability, spiritual and sexual dominance ('brave', 'strong', 'virile', 'individual', 'worker', 'heroic', 'boss'); mulher 'woman' is characterized by her family and sexual submission ('rib', 'other half, 'obligation', 'old woman', 'prostitute'). In other words, the semantic features of 'man' (homem) are globally characterized by control, activity and leadership in contrast with the passivity and dependence (or simple 'feminine copy' of masculine) of 'woman'. The exception is the term rainha 'queen', but it is weaker than the other pair term ra'king'. 3.2.2 Other crucial pairs The following pairs, esposa - marido 'wife' - 'husband'; companheira - companheiro 'female companion' - 'male companion'; avo - avo 'grandmother' 'grandfather'; dama - cavalheiro 'lady' - 'gentleman'; senhora - senhor 'lady' 'gentleman'/ 'mistress' - 'mister'/ 'Mrs' - 'Mr'; patroa - patrdo 'wife' - 'husband' [popular]/ 'female boss' - 'male boss' and ministra- ministro 'female minister' - 'male minister' are a representative sample of the basic semantic
148
GENDER, LANGUAGE AND NEW LITERACY
areas we have determined. They are grouped into three domains: (1) the family sphere; (2) the social sphere: woman/man relationship; and (3) the economic and professional sphere. 1 Family sphere, the pairs esposa - marido, companheira - companheiro and avo avo illustrate the family sphere. The frame constructed for the feminine terms has the same features we have listed above; it is a negative representation of the family relationship (depending on man's leadership and sexual submission). For instance, the word 'grandmother' has only one endearing term as a synonvm: vovo 'granny'. The masculine image is completely different. Man is presented as a leader and a man of the world, of 'all the worlds', present and past, in the sense that the ancestry line is another important feature of the representation. In addition to the particularities already mentioned, the third pair, companheiro - companheira, is a transition pair between the family sphere (feminine term) and social sphere (masculine term). (9a)
(9b) (lOa) (lOb)
(11 a) (lib)
esposa. companheir a (subs.) (companheira, conjuge, consorte, costela, dona, metade, obrigafdo, mulker, patroa, senhora, velha); abraca r (verbo); marido (subs) marido: marid o (marido, conjuge, consorte, dono, esposo, homem, patrdo, vardo). companheira: coleg a (subs.) (colega, concubina, consorte, esposa, Jemea, pdreas); companheiro (adj.). companheiro: companheir o (adj) (companheiro, acompanhador, adjunto, amdsio, amigo, associado, camarada, co-associado, colega, condisdpulo, confidente, confrade, consocio, consorte, esposo, insepardvel, irmdo, parceiro, revolver, socio, unido). avo: vovo (subs.) (vovo); anciao (adj.) . avo: anciao (adj.) (anciao, antepassado, avoengo, primogenitor, progenitor, vovo, ascendencia, ascendentes, genealogia, maiores).
2 Social sphere: the woman / man relationship: As examples of nouns representing social relationships, the pairs dama - cavalheiro 'lady* - 'gentleman' and senhora - senhor ('lady' - 'gentleman'/ 'Mrs' - 'Mr') have the same meanings we have already shown. The representation of woman as a lady (dawtf and senhora) is predominantly attached to the two semantic spheres 'sexual object' and 'housewife'. The relationship between these aspects is very strong, as we can see from the presence of the same 'synonyms' for both cases: dona, madona, senhora. Those two aforementioned areas are organized around this nuclear domain. The terms cavalheiro 'gentleman' and senhor 'lord'/'Mr', on the contrary, recall the usual positive features: physical and moral strength, power, and civility.
PORTUGUESE MARQUES
149
(12a) damcc. actri z (subs.) (actriz, amiga, concubina, Jemea, dona, madona, manceba, marafona, meretriz, mulher, prostituta, rainha, senhora). (12b) cavalheim cavalheir o (adj.) (cavalheiro, brioso, bizarro, cavaleiro, cavaIheiresco, cavalheiwso, cartes, delicado, digno, distinto, fidalgp, gentil, gentilhomem, nobre, senhor). (13a) senhora: senhora (subs.) (senhora, ama, dama, esposa, iaid, madama, madona, matrona, sid, sinhd);amo (adj.). (13b) senhor. amo (adj.) (amo, chefe, Deus, dom, dominador, dono,fidalgp, imponente, magnate, pai, patrdo, possuidor, proprietdrio, senhoril, senhoria, so, soberano). 3 Economic and professional sphere: The prevalence of the masculine forms (and masculine world) is evident. Starting with the pair patroa - patrdo ('female boss' - 'male boss'), it is obvious that it is the masculine term that enables us to include them in that sphere. There are no 'synonyms' for patroa related to the professional area. They all belong to a private sphere, once more derogatorily represented (patroa is a popular term for 'wife'; 'housewife'): (14a) patroa: ama (subs.) (ama, amdsia, dona, esposa, madama)', patra o (subs.). (14b) patrdo: patra o (subs.) (patrdo, amo, arrais, bianco, chefe, dono, empregador, esposo, orago, padrdo, padroeiro, pai, patrono, protector, senhor). In the same way, the pairs ministra - ministro 'female minister' - 'male minister', juiza - juiz 'female judge' - 'male judge' and medica - medico 'female doctor' - 'male doctor' are ideologically represented. In contrast with current practice in modern Portuguese society, the Word Thesaurus docs not include the word ministra ('woman minister') as 'member of the government', but presents only archaic synonyms (e.g. 'servant'). With respect to the pair^uzza -juiz ('femalejudge' - 'malejudge'), only the latter (juiz) is registered. Something similar happens where the last pair, medica - medico 'female doctor' - 'male doctor' is concerned. The inventory of synonyms is the same for both terms, that is to say, an inventory in the masculine form, showing a lack of autonomy for the feminine term. ministra: criad a (subs.) (criada, medianeira, servo, servente); ministra r (verbo); administrador (subs.). (15b) ministro: administrado r (subs.) (administrador, auxiliar, embaixador, escravo, executor, governante, instrumento, juiz, magistrado, medianeiro, padre, peru, pregador, sacerdote, servidor, servo, urubu-cafador); ministra r (verbo). (16a) juiza: not found (not found) (16b) juiz: juiz (subs.) (juiz, apreciador, drbitro, conhecedor, julgador, justifa, magistrado, sentenciador). (15a)
150
GENDER, LANGUAGE AND NEW LITERACY
(17a) medico: medico (adj.) (medico, clinico, curador, doutor, esculdpio, facultativo, fisico, galeno, medicinal, terapeutd). (17b) medico: medico (adj.) (medico, clinico, curador, doutor, esculdpio, facultative, fisico, galeno, medicinal, terapeutd).
To sum up, the quantitative lists presented above can be recalled so that they can be reinterpreted in a qualitative way. In fact, the female world either does not exist in the Word Thesaurus outside the family/sexual sphere, or it is an extension of the male world. The occurrence of derogatory terms in order to characterize female semantic fields stresses the unbalanced representation of women's and men's roles in modern societies. These are obsolete social and cultural stereotypes, similar to those of nineteenth-century grammars. The results of our analysis are consistent with the definition of noun gender given, in 1823, by the Portuguese grammarian Jeronimo Scares Barbosa: Sao do Genero Masculino todos os nomes Substantives, que significao macho, assim proprios, como appellatives, ou sejao de homens, como Andre, Rei, ou de brutos, como Bucephalo, Cavallo, o u de profissoe s e ministerio s proprio s d o homem , como Propheta, Patriarcha, Magistrado, Sacerdote, e ainda aqueles que sendo femininos quando significao couzas ou ac^oes, passao a designar varies officios proprios do homem, como o Atalaia,... Sao do Genero Feminino todos os nomes Substantivos, que significao femea, ou sejao proprio s d e mulher , como Matildes, Ignez, o u Appellatives d e officios , e couzas, qu e Ihes pertencem, como Rainha, Mai, Avo, Madrasta, Costureira, Tecedeira; ou de brutos, como Egoa, Vacca. (Jeronimo Scares Barbosa 1822: 126-7).I0
4 Conclusio n The Word Thesaurus is undoubtedly a pre-structuralist dictionary that regards language as a nomenclature. The contextual relationships of meaning are ignored. The Word Thesaurus is a list of words without any criteria of frequency, similar functions and discursive registers. It is the result of an atomistic conception of the lexicon.11 The word-to-word relationship causes a radical decontextualization of word meaning, and introduces, consequently, semantic and pragmatic bias. The alphabetical word order used in the Word Thesaurus that we are studying does not prevent the construction of ideological domains concerning the social relationships of women and men, because it depends, especially, on lexical choices. There is also a rough and inadequate interpretation of grammatical gender. The Word Thesaurus is a dictionary providing a choice of stylistic variations, useful only to those users who have very good skills in the Portuguese language, but useless to those who want to know the meaning or use of a particular word. Furthermore, the Word Thesaurus gives an outdated perspective of Portuguese society, especially of female roles therein.
PORTUGUESE MARQUES
151
The relationship between the way the world is perceived and its linguistic representation is not simple. That is why die unbalanced image of die female and male social roles cannot be corrected simply by some (is it possible?) changes in grammatical rules, because experience shows that we cannot just degender languages. Prejudices will always find other ways of appearing. With regard to verbal language (as well as to everything else), ignorance of language function is die first enemy to combat. Besides, verbal language is not a code and the act of speaking is not innocent. So, an 'aseptic linguistic reference' is not desirable and, moreover, it is impossible. In fact, the 'construction' of a hypodietical aseptic 'degendered' language would be an erroneous approach to solving the problem of social inequalities: changing the words is not enough. To hide diem is, however, uie worst possible solution to eliminating this biased situation. Acknowledgement I wish to thank Alvaro Iriarte for die fruitful discussion on dictionary organization, Ana Ribeiro, Clara Oliveira and Joanne Paisana for the review of the English version. I owe a particular debt of gratitude to an anonymous reviewer for die meticulous comments he provided on an earlier English version of this chapter. Notes 1 Our translation is: and as all animals are naturally subdivided into two classes or Genders according to the male and female sex, Grammarians put the nouns relating to the former into the masculine Class or Gender, and the latter into the feminine one. These are natural classes formed only by animal nouns. All the sexless beings should be included in the Neuter Class or Gender, i.e. they should form a third Class. [...]. With regard to the natural classes, the meaning itself determined its gender; only the noun ending could indicate arbitrary classes. Consequently, noun gender classification is effected either through the meaning of the noun, or by its termination. (Jeronimo Scares Barbosa 1822: 123-4) 2 And these are not the only ones. There are other possibilities: ponU, juri, jlar, etc. 3 It's a sporadic linguistic mechanism used when it is important to signal the male or the female. However, it's important to our study because certain feminine derivations from a masculine form may cause semantic ambiguity (that is the case ofopoUda 'the (m.) policeman' -apolida 'the (f.) police force' / 'the (f.) policewoman'). In order to avoid this situation, we can use a similar process for the representation of 'feminine': for clarity, we will then have a mulher-policw 'the (f.) policewoman'. 4 Our translation: the marked term 'feminine' always [corrected in a footnote to 'usually'] 'feminine being'; the unmarked term, masculine, sometimes means 'masculine being', sometimes the generic concept or the universal group of all beings
152
GENDER, LANGUAGE AND NEW LITERACY included in it, without any distinction of species, sometimes a particular collection of members of one or another species.
5 6 7
8 9 10
The author could not help considering that it is not after all clarified why the 'masculine' is the unmarked term; and he advances the following hypothesis: 'maybe,... in all kinds of society, even in the matriarchal ones, it is the man who is concerned with the activities essential to the community life' (1984: 638) (our translation). The relevance of that assumption signalizes that the masculine form of organizing societies has conditioned this grammatical categorization. By 'sexless', we mean 'without specification of the female or male sex'. This feature is not semantically relevant. Even the most conservative institutions, like the army, have allowed women the possibility of developing a career there. Dictionaries still do not register these forms, yet through a small, informal inquiry, we were able to ascertain the acceptance, by Portuguese speakers, of these new forms. The question is simplified here. The forms of treatment are very complex, we are pointing out the most frequent and unmarked ones. In the following examples, the meanings are in bold, whereas the items given for each meaning stand in brackets, in italics. Our translation: Belonging to the masculine gender are all nouns that mean male , including proper nouns or common nouns, including male nouns, such as Andrew or king, or animal nouns, such as Bucephalus, horse, or professions and tasks natural to men, such as prophet, patriarch, judge, priest, and even those that, being feminine when meaning things or actions, designate several jobs natural to men, such as oAtalaia ('the sentinel') . . . . Belonging to the feminine gender are all nouns that mean female, including those natural to women, such as Mathildas, Ignez (women's names), or common nouns of jobs and things that belong to them, such as queen, mother, grandmother, stepmother, dressmaker, woman 'weaver', or animal nouns such as mare, cow, fox.
11 For an overview of the organization of dictionaries, see Iriarte Sanroman (2001).
References Barbosa, Jeronimo Scares (1822), Grammatica Philosophica da Lingua Portugueza. Lisbon: Academia das Sciencias. Carvalho, J. Herculano (1984), Teoria da Linguagem, Vol. II. Coimbra: Coimbra Editora. Iriarte Sanroman, Alvaro (2001), A Unidade lexicogrdfica -palavras, colocafoes, frasemas, pragmatemas. Braga: Universidade do Minho, CEHUM (Coleccao Poliedro).
10 Word's Spanish Thesaurus, some limits of automaticity Pura Guil
1 Introductio n Spanish is widely estimated to be third among world languages as regards the number of speakers (around 380 million) and the number of countries (around 20) of which it is the official language. Despite this enormous geographical spread (native speakers are found throughout the five continents), it enjoys a high degree of linguistic homogeneity, especially at its more educated levels, so that the two broad varieties which are generally recognized Peninsular European (or Castiliari) and Latin American - are for the most part mutually comprehensible. Contributing to this has undoubtedly been the unifying work of the Real Academia Espanola de la Lengua, integrated into the Asociacion de Academias de la Lengua Espanola, a body comprised of the world's 22 existing Academies and whose goal is 'trabajar en pro de la unidad, integridad y crecimiento del idioma comun'.1 The analvsis which follows will consider data relative to the Peninsular European variety as presented in the Thesaurus of Microsoft Office Word 2003, version 11.0 CSinonimos Espanol: (Espafia - alfab. Internacional)') ,2 although, as we will see, it will also address, in random fashion and without further specification, examples which, despite the profound homogeneity of the Spanish language alluded to previously, clearly belong to the Latin American variety. This choice of the Castilian variety was determined as much by two factors. On the one hand, it is this variety which was examined in Bazzanella et al. (2000), which is the starting point of the present work. On the other hand, the Castilian variety, given the accelerated pace of social changes occurring in Spain in the past 30 years, with their consequent impact on the language, has been the object of a much greater number of studies on gender and variety than the Latin American varieties (as observed by, among others, Nissen 2003: 252-3). In this chapter, §2 is dedicated to a brief description of Spanish gender markers, while §3 will trace a general panorama of the roles attributed to both women and men in Spanish society and the changes that have taken place in the last thirty years. §4 will deal with our study of the categorization of masculine and feminine terms in MS Word Thesaurus for Spanish,
154
GENDER, LANGUAGE AND NEW LITERACY
detecting changes that might have been introduced between versions 6.0 and 11.0. 2 Gende r in Spanish Spanish has two genders, feminine and masculine, which in animate nouns, and specifically in those denoting human beings - which is what concerns us here - are determined by the sexual differentiation of their referents. Nouns, whether singular or plural, agree in gender with their determiners and modifiers, as well as with the anaphoric personal pronouns in the third person, singular and plural, and die deictic first and second person plural. Formal gender expression in nouns for persons adopts a variety of formats (Ambadiang 1999: 4854): 1 common root with gender opposition marked in the suffix: hermano (m.) 'brother' - hermana (f.) 'sister'; 2 diverse root with gender opposition marked in the suffix: yemo (m.) 'sonin-law' - nuera (f.) 'daughter-in-law'; 3 invariable noun with gender opposition expressed by its determiners and modifiers rather than its own morphology: el mdrtir (m.) - la mdrtir (f.) 'the martyr' (m./f.); 4 pair of invariable nouns having differing genders ('heteronyms'): hombre 'man' - mujer1woman'; 5 (in)variable noun with the modifier mujer placed before it: el obispo (m.) 'the bishop' - la mujer obispo (f.) '(the lady) bishop* (although gender indication through nominal composition does not exist in Spanish, this procedure is the closest we find to it). In each of these two classes of grammatical gender, masculine and feminine, we find lexemes which have only one form and consequently do not determine sex: they may refer equally well both to women and to men (compare: (la) victima (f.) 'the victim', (el) cadaver (m.) 'the corpse'). These are the socalled 'epicene nouns'. In other cases, however, the masculine noun form must be used to make a generic reference, in the plural referring to a group comprised of both women and men (los padres, 'fathers', but also the generic 'parents'). In the singular they may also refer to the category (el hombre es mortal 'man is mortal'). What needs to be understood is that, as Nissen has astutely noted (2003: 257), only epicene nouns (las personas 'people') allow reference to a group comprised of both men and women, or only men, or only women, while masculine nouns used generically may not be applied to a group comprised only by women, precisely because the noun in question has a feminine form (los padres 'parents, fathers' - las madres 'mothers'). Consequently, the denomination 'epicene' given by the Academia (1989: 2.2.6A) to these last is inappropriate. Although the unstressed endings -a and -o are not die only ones and are by no means exclusive to feminine nouns for women or masculine nouns for
SPANISH GUIL
155
men, they are certainly those with which gender distinction is most frequently associated. In most cases, the feminine form is obtained by derivation through the regular inflection of masculine noun forms, as the Real Academia Espanola has noted (1989: 2.2.66). On rare occasions, for social or historical reasons, morphological derivation has taken an inverse course, that is to say, with a masculine form being obtained from a feminine noun: viuda (f.) 'widow' - viudo (m.) 'widower' (Real Academia Espanola 1989: 2.2.66), although in reality this inverse procedure is synchronically imperceptible. In a certain number of cases, Spanish also possesses independent lexical units, which may or may not be formally marked by specific gender morphemes, and which include the semantic features [+ female] or [+ male] (compare cases 2 and 4 above). Consequently, most nouns of kinship in Spanish, as in other Romance languages, are distinguished through morphological alternation (hijo (m.) 'son' - hija (f.) 'daughter'), with lexical gender also represented: madre (f.) 'mother' - padre (m.) 'father', mama (f.) 'mum' papa (m.) 'dad', nuera (f.) 'daughter-in-law' -yerno (m.) 'son-in-law'. More of this occurs with nouns of profession, position, rank, trade, activity, etc. (i.e., 'title' nouns), as generally those for women have been derived morphologically from those for men: zapatero (m.) 'shoemaker' - zapatera (f.) '(female) shoemaker', abad (m.) 'abbot' - abadesa (f.) 'abbess', rey (m.) 'king' - reina (f.) 'queen', actor (m.) 'actor' - actriz (f.) 'actress'. We also find cases of lexical gender differentiation, however. For example, while the masculine preceptor, designating a private tutor for children, has the morphological derivation preceptora (f.), this latter word is by no means common, the exclusively feminine institutriz (Moliner 1996: preceptor) being more widely employed. Along with terms belonging to more formal registers for designating the two members of a married couple,3 in common use mujer (f.) 'woman' also signifies 'wife', while hombre does not signify 'husband', a specific term, marido, being used for this. Clearly degrading is the use of the pair hembra varon (or hembra - hombre) for the human species (as it appeared in the Spanish Civil Register until its modification in 1993 to mujer - varon or hombre), given that the term varon (m.) '(human) male' is applicable exclusively to men, while hembra (f.) 'female' may refer not only to women but also to animals. At the same time, hembra forms a symmetrical pair with macho (m.) 'male', these terms referring to the sexes of animals, and so only this pair should be admissible if extended to human beings4 (Garcia Meseguer 1994: 28-30). Apart from the problems arising from the derivation of the feminine from the masculine in nouns of title, which we will examine in §3, the distinct value assigned in some cases to the feminine form of a lexeme must be noted. The indeterminates individua, projima, sujeta orfulana, as well as other derivatives, such as parienta or socia, possess a clearly pejorative value lacking in their masculine counterparts (individuo, projimo, sujeto, fulano, pariente, socio). Likewise, militara, generala, or coronela are used, or have been used, to deprecate or poke fun at the wives of the militar (m.) 'soldier', 'military officer', general (m.) 'general' or coronet (m.) 'colonel'. Used figuratively, zorro
156
GENDER, LANGUAGE AND NEW LITERACY
(m.) 'fox' refers to the man who is clever and astute, and although this value should be retained when the term is applied to a woman (zorra (f.) 'female fox (vixen)'), this association is in fact absent, and that of 'slut' found in its place. Similarly, while un hombre publko refers to a man concerned with public affairs, una mujer publica, at least traditionally, signifies prostituta (f.) 'prostitute'. In the same way, in Spanish, we find asymmetry in terms of courtesy such as senora (f.) - senor (m.), the latter being independent of the marital status of the man to which it is applied, while senora is used to address a married woman, just as senorita (f.) is used for an unmarried woman; that is, in both cases the woman is not considered in terms of her own individuality but rather in her relation to men. It is true that senorito (in.) also exists, although it does not designate a 'young, single man' or 'bachelor', but rather a man who is 'young, well-to-do and idle'. Lastly, and although separated from the question of gender properly speaking, a few other cases deserve to be enumerated, which may be considered to show the way in which women and men are represented in Spanish culture and society. The diminutive suffix in Spanish is applied to proper names, particularly those of children and women. Men are rarely called by the diminutive form and only within a familv (or otherwise verv intimate) context. As Maria Moliner remarked as early as 1970, outside these situations its use indicates affection but at the same time does not concede much importance to the man being referred to (Moliner 1996: diminutivo). By contrast, it is quite normal to use hypocorisms ('pet names') when addressing women, even outside family or intimate contexts. Although they are terms of affection, such terms may also reveal an attitude of condescending superiority, especially if similar terms are not used with the men present in the same situation and if the woman is not in a position to react by using this same form of address for her interlocutor. There is a tendency to use, both allocutively and designatively, for adult women such nearly synonymous terms as chica (f.) 'girl', chiquilla (f.) 'very young girl', pequena (f.) 'little one', nina (f.) 'little girl', nena (f.) 'baby' and the like, all of these expressing a protective attitude which relegates the woman to a permanent immaturity or 'underage' status (and which women tend to accept as an illusion against the implacable process of aging). When used in asymmetrical situations, however, for example in the workplace, they may conceal a desire to infantilize their subject and thereby undermine her authority and worth (see Garcia Mouton 2003: 197-8). 3 Socia l roles and social change At the outset of this chapter, reference was made to Nissen's observation (2003: 252-3) attributing the conspicuous production of studies on 'gender and language' in Spain to the accelerated rate of social change experienced after the death of Franco in 1975. It is one with which, naturally,
SPANISH GUIL
157
we are in agreement, provided that it is not allowed to obscure the fact that important precedents exist: from the vindication in the nineteenth century of the right to education, to the demands of the first women's organizations in the early twentieth century for universal suffrage, obtained under the Second Spanish Republic (1931-36). After the Civil War (193639), the newly installed Franco regime interrupted the gradual integration of women into political and public life. This lull, however, was to be only temporary, lasting as it did only as long as the bleak, dormant period of the dictatorship.5 Proof of this was the explosion of new vitality that accompanied the political transition towards democracy, along with the country's corresponding modernization, which is a source of considerable pride and satisfaction for the immense majority of Spanish men and women today. In 1983, the Institute de la Mujer was created with the express purpose of promoting policies of social equality for both sexes, following the principle of non-discrimination recognized in the Spanish Constitution of 1978. The access of women to all levels of academic and professional life grew from a minority of cases to a generalized phenomenon. The current participation of women in the most varied social, economic and cultural spheres has since risen to levels which, though still not equal to those of men, nevertheless represent a degree of progress unimaginable only 30 vears ago. The changes which have appeared in the language are clear indicators of the transformations which Spanish society has undergone, from the awareness of the sexism inherent in the Spanish inflection system to the search for appropriate tides in the workplace. While the purpose of the feminine gender is to indicate that the referent of the noun is a woman, the masculine not only indicates the male sex of the referent, but also may have a generic use, as well as serving as the deciding factor in processes of syntactic coordination involving nouns of opposing genders. This functional asymmetry, enhanced by the phenomena of modifier and determiner agreement, has the effect of hiding women behind men, and thus of relegating them to a secondary position in which they can more easily be ignored (Garcia Meseguer 1994; Ambadiang 1999; Garcia Mouton 2003). With the massive access of women to the professional world, the problem is to find acceptable tides for professions, positions and trades. Although the language system in theory allows for the gender-based splitting of most of the nouns concerned, the question was not. so simple 20 years ago. Many of them had traditionally only been used in the masculine, and the new feminine forms, although properly derived, seemed shocking or lacking in euphony. Even so, precisely because of their novelty, they were not encumbered with the connotative weight borne by certain other titles which had undergone gender splitting long before. Thus, for example, la medico (f.) 'die (female) doctor', referred more often to 'the doctor's wife' or la alcaldesa (f.) to 'die mayor's wife', than to the woman who held the corresponding profession or position (something that was exceptional until quite recent times), and this
158
GENDER, LANGUAGE AND NEW LITERACY
seemed to invalidate these terms when they were applied to women in newly emerging professional positions. In other cases, as we saw earlier, the feminine and masculine forms of the same lexeme have acquired different values, the feminine corresponding to a less prestigious profession than the masculine. Such is the case of secretaria (f.) 'secretary', a job traditionally performed by a woman and denoting a subordinate administrative employee, while secretario (m.) 'secretary/secretary-general', refers to a post of a certain importance within an organization or company, one which is traditionally occupied by a man. A feminine form may even be assigned a negative connotation from which the masculine is exempt: verdulero (m.) merely designates a man who sells vegetables, but verdulera (f.) has the additional meaning of a 'brazen and sharp-tongued woman'. All this type of casuistry, which in one way or another confers less prestige on nouns denoting female tides, has led many professional women to prefer the masculine form when referring to their educational tides, jobs or professional posts, if for no other reason than to lay claim to a certain professional status or position in the hierarchy: arquitecto (m.) 'architect', ingeniero (m.) 'engineer', notario (m.) 'notary'. Only in those jobs or professions traditionally done by women has adoption of the masculine form not been the norm: maestro (f.) ' (primary school) teacher', mecanografa (f.) 'typist', comadrvna (f.) 'midwife', farmaceutica (f.) 'pharmacist'. This lesser prestige of the feminine form is confirmed if we observe the resistance of our society to creating masculine derivatives from feminine forms for professions typically carried out by women but nowadays open to men as well. This is the case of azafata (f.) 'stewardess, air hostess', for which the epicene auxiliar de vuelo 'flight attendant' has been coined, or the similarly epicene auxiliar tecnico sanitaria (or ATS), which can be used for both the traditional enfermera (f.) 'nurse' and the masculine enfermero (m.) '(male) nurse', as if replacing the feminine noun somehow conferred a higher social status on the profession. Naturally, a generalized trend has developed of recommending, within many diverse spheres, that women be brought out from under the shadow of men and that everything which prevents their complete visibility be eliminated from the language. To this effect, it has been recommended, among other things, that generic masculine nouns, e.g. los espanoles 'Spaniards', be replaced by epicene forms, lapoblacion espanola (f.) 'the Spanish population', or by split forms, las espanolas (f.) y los espanoles (m.) 'Spanish men and women'. As happens in other countries, however, these types of reform, while succeeding in drawing the public's attention to our culture's ingrained androcentrism, take a long time to have any effect, and frequently they are only taken up by committed minorities or fail to appear outside the official written language. In the case of split forms, while this often tedious practice is certainly at times necessary in order to clarify reference, it tends to be received with scepticism through being associated with a certain brand of political rhetoric, heard especially at election times and aimed at the ballot box.
SPANISH GUIL
159
More successful have been recommendations for the use of title nouns with explicit gender markers, such as lajueza (f.) 'the (female) judge', lajefa (f.) 'the (female) boss'; and this despite the fact that in cases like these the determiner alone (la/el juez, la/el jefe) would be enough to mark gender. Contributing to this, however, has been an ongoing process of change in society, which has been increasingly evident in recent years and which has led to the use of feminine forms to designate women in all manner of posts: it seems that in the Spanish judiciary there are now more women than men, while half of the current government is comprised ofministras (f.) '(female) ministers'; when the government's first Vicepresidenta (f.) '(female) Vicepresident' was appointed, no one assumed any longer that this referred to the wife of the Vice-president. Even so, Ambadiang (1999: 4867-70) claims a greater degree of gender splitting in Latin American Spanish of nouns which remain invariable on the Peninsula, attributing this to the propensity toward inflectional regularization characteristic of the former variety (consider la testiga 'the female witness' vs el testigp 'the male witness'), in contrast to the latter variety's tendency to adopt an array of different solutions, seemingly in obedience to some precepts of economy in the use of the language system: thus the feminine may be marked both by inflection and by the determiner (la huespeda 'the female guest' vs el huesped 'the male guest'); or uniquely by the determiner (la testigo 'the female witness' vs el testigo 'the male witness'); or by contradictory combination of both solutions (la medico 'the female physician' vs el medico 'the male physician'). 4 MS Word Thesaurus for Spanish We have taken as our point of departure the analysis proposed in Bazzanella et al (2000: 203-8), which was conducted on 24 pairs of terms registered in the Spanish Thesaurus of Word 6.0. This gives us a basis for detecting any changes which have been made in Word Version 11.0. Initially, we intended the entries analysed in the later version of the Thesaurus to be the same as those in the earlier study, but as our work proceeded we felt the need to minimally but significantly widen our original set of terms. 4.1 Spanish in Word 6.0 On using this version of the Thesaurus, the lack of clearly defined lexicographical criteria in its compilation becomes immediately evident. Entries, 'meanings' and 'synonyms'6 are selected and included/excluded in an apparently chance manner and arranged haphazardly. It is true that the terms searched for were found in the feminine as well as in the masculine, except in one case: directora (f.) '(female) director' was not included, although director (m.) '(male) director' was, with one meaning (regente) and sixteen synonyms (negente, rector, jefe, presidente, encargado, dirigente, directive, dignatario, guia, autoridad, principal, encauzador, tutor, adiestrador, orientador, conductor). To explore terms denoting spheres of authority or
160
GENDER, LANGUAGE AND NEW LITERACY
power we selected only this pair and that formed by jefa (f.) and jefe (m.) '(female) boss' - '(male) boss'.7 da) (lb)
jefa: duen a (duena, senora, patrona, propietaria, ama, casera) jefe. directo r (director, dirigente, superior, regente, rector, superintendente, decano, senor, patron, patrono, amo, dueno, soberano, tirano, dictador, autocratd).
Compared to the rich lexicon offered for the masculine (although not all of the terms proposed can be considered acceptable), the image conveyed of the feminine seemed to exclude the possibility of a woman being a director; at most, she might be considered jefa only on the basis of wealth. Incomprehensibly, no meaning was given for the entry soltera (f.) 'unmarried woman', with reference made instead to soltero (m.) 'bachelor', defined as celibe and among whose synonyms appeared virgen 'virgin' (celibe, libre, suelto, mow, mancebo, casadero, nubil, virgen, doncel, solteron). A search for virgen, however, yields the surprising result that the meaning donceUa 'virgin'; 'handmaiden' as well as its synonyms are offered in the feminine, as if this condition were exclusively the province of young women (donceUa, adolescente, pubescente, impuber, chiquiUa, chica, moza, zagala, muchacha, senorita, damisela). Virginity in an adult or elderly woman seemed either not to exist, or to be simply not worth considering. The basic pair mujer - hombre received asymmetric treatment as regards the number of meanings given. Varon '(human) male' was the only meaning offered for hombre (m.) 'man', along with a series of synonyms which, apart from the first two, possess a generic value, however strange they may seem in certain cases: varon, macho, persona, individuo, mortal, criatura, semejante, projimo, ser humano, sujeto, cristiano, quidam, personaje, ente, ser, fulano. In contrast, three meanings for mujer (f.) 'woman' were listed: one related to gender, hembra 'female', another denoting low socio-professional status, criada 'maid', and the last referring to matrimony, esposa 'wife'. As will be obvious, this asymmetry in number of meanings extends also to their quality. Synonvms offered for the meaning of hembra 'female'8 appeared arranged in a series ordered sequentially according to the phases of biological development (unlike the treatment accorded to varon). These began with adulthood, designated in terms relative to social status (senora, dama), less marked than the terms which followed (duena, matrona, ama - antiquated and out of current use), and then continued into the youngest phase, represented by the lexeme senorita (see §2 above), and followed by donceUa (i.e. virgen, as we have just seen). Here the chronological sequence was interrupted to offer two surprising tenns as synonyms: venus 'Venus' and eva 'Eve'; with their connotations of 'very beautiful woman' and 'temptress' these reproduce the widelv held stereotype which attributes to women specific physical properties that give them power of attraction over men. The next term, nina 'little girl', represents riot so much a leap into childhood (see §2 above) as a common reference term for young women, fitting into the
SPANISH GUIL
161
chronological phase denoted by the following synonyms: joven, chica, muchacha, moza and zagala (the two last being antiquated or rural). Note, once again, the absence of lexemes referring to the final stages of a woman's life, such as vieja or anciana, as if in old age a woman became not only excluded from aesthetic or sexual considerations, but ended up lacking gender as well. Gender-biased choices are evident also in the complementary pairs madre (f.) 'mother' - padre (m.) 'father' and hija (f.) 'daughter' - hijo (m.) 'son'. In the first case, it was evident that in order to be a madre the basic ability to produce a new being is enough, while padre also implies the creation of a 'line': (2a) (2b)
madre: matron a (matrona, senora, ama, mujer, hembra, dama, madraza, mama). padre: progenitor (progenitor, procreador, padrazo, papa, cabeza, padrastro, familiar, ascendiente, autor de los dias).
Even more striking was the presentation of hija compared with that of hijo: (3a) (3b)
hija: nen a (nena, chiquilla, infanta, cria, chicuela, mocosuela, muchacha, chica, pequena, nina). hija reton o (retono, vdstago, descendiente, sucesor, heredero, familiar).
Just as the hija is reduced to a level of puerile immaturity, the treatment of hijo underscores the transmission of the family name and fortune through the masculine line, as if such a thing were onlv possible in the male case. On comparing the meaning cnada (f.) 'maid' of the entry mujer and its synonyms (sinnenta, doncetta, moza, camarera, fdmula, maritomes, chica, muchacha, servidora, asistenta, domestica, azafata, ninera) with those listed for criado (m.) 'manservant' (sirvient e siruiente, camarero, mozo, fdmulo. domestico, asistente, servidor, recadero, lacayo, mayordomo, ayuda de cdmard), it was at once clear that in neither case were the entries very accurate, as most were lexemes whose use could only be controversial, given the possible offence they could be likely to cause to their referents. There were nevertheless some observable differences between the two genders here: in the feminine entries we find with no indication whatsoever of appropriate register - lexemes such as fdmula or maritornes? only possible with a humorous or strongly pejorative value, values absent from the masculine entries, which, in contrast, included terms pertaining to high ranking personal servants (mayordomo, ayuda de cdmara). More balanced from the standpoint of gender was the data regarding the terms esposa (f.) 'wife' - esposo (m.) 'husband', although it was surprising to find companem as the meaning of esposo, rather than marido, which forms a complementary pair with mujer, the meaning given for esposa: (4a)
esposa: mujer (mujer, senora, conyuge, pareja, companera, costitta, consorte, contrayente, desposada, casada, media naranja, cara mitad).
162
(4b)
GENDER, LANGUAGE AND NEW LITERACY
esposcr. companer o (companero, marido, conyuge, hombre, consarte, pareja, contrayente, casado, cara mitad, media naranja).
The synonyms offered for these lexemes are, however, a clear example of the lack of selection and ordering criteria alluded to previously. In the case of esposa, the order of the first two synonyms may correspond to the common use10 of mujer among the middle and upper classes, and to senora in lower classes. It is not clear, however, why conyuge has been separated from consorte (both are used in formal, administrative or legal contexts) by the insertion, in the masculine, of hombre (unquestionably vulgar)11 and, in the feminine, pareja and companera (these usually being applied to a member of a couple who are not legally married), to say nothing of costilla 'rib', a humorous, colloquial denomination,12 carrying with it an undeniable sense of dependence and possession. After the references to aspects of the matrimonial ritual, the listings for both genders conclude with two expressions which are generally considered to be either affected or facetious: cara mitad 'other half and media naranja 'soulmate (literally, 'half orange')'. Something similar can be said of the entries offered for senora 'married woman', 'lady' - senor'man', 'gentleman': (5a)
(5b)
senora: dam a (dama, matrona, ama, duena, madre, cortesana, camarera, senor [sic!]); mujer (mujer, esposa, conyuge, companera, pareja, consorte, costilla, media naranja). senor, caballero (cabattero, noble, patricio, aristocrata, titulo, hidalgo, cortesano); amo (amo, dueno, patron, jefe, propietario, superior, cabeza, poseedor).
Arranged chaotically, and with no indication of usage, we find terms that are either antiquated (ama, duena, camarera, patricio, hidalgo, cortesano), colloquial (costilla) or clearly erroneous (senor (m.) appeared among the feminine entries). The usual meaning for cortesana as being 'high-class prostitute' was ignored, and the entries taken as a whole abounded in both gender discrimination (a woman is a senora through being the wife of a senor) and the stereotypically Hispanic obsession with grandeur. Stereotypical in the extreme, however, was the data given for the last pair examined: the adjectives femenino and masculina. (6a) (6b)
femenino: mujeri l (mujeril, femenil, suave, debil, afeminado, amujerado, gracioso, sutil, grdcil, delicado). masculincr. varonil (varonil, viril, fuerte, hombre, macho, redo, energico, vigoroso, valiente, hombruno).
The catalogue of properties attributed to each gender could hardly be more cliched and degrading.
SPANISH GUIL
163
4.2 Spanish mWord 11.0 In this latest version of Word the format of the Thesaurus has been modified, not only to include a new toolbar, called 'Reference', which allows the user to access a variety of resources, within her/his computer or on-line, without exiting the document being worked on. These include a variety of monolingual and bilingual dictionaries, Microsoft's Encarta Encyclopedia and several dictionaries of synonyms in other languages. Besides offering a list of possible options, the Thesaurus also permits the widening and linking of searches with only a click on the given term. The product reviews could not have been more flattering and so the level of expectation aroused in the user is very high. It is true, nevertheless, that in the corresponding 'Help' topic, one is warned that the 'appearance of apparently inappropriate terms is unintentional'.13 This increase in available information might well have brought with it a reduction in other information already included in the Thesaurus. However, in Word 11.0, we examined the same cases that appeared in Word 6.0 and even two more (directora, sotiera) - the first of these was absent and the second, although included, was only found in reference to the related expression (see §4.1).14 The number of meanings has been increased (to an average of 2.3 meanings per entry, as opposed to 1.2 in Word 6.0), as has the total number of synonyms in comparison to those in Word 6.0 (although their relative value in respect to the number of meanings has decreased proportionately: 7.75 in Word 11.0, compared to 10.1 in Word 6.0). In any case, the general impression that more information is given is also favoured by the clear differentiation among the given meanings on the basis of grammatical categories (often not the appropriate one), and by the inclusion of antonyms and related terms. The terms analysed are included in the on-line Appendix, available at www.lingue.unibo.it/wordthesaurus, and it is to this we now refer. The analysis began with the fundamental pair mujer 'woman' - hombre 'man' and here the continued absence of clearly defined lexicographical criteria was already in evidence. It is true that, with respect to the preceding version, there seemed to have been an attempt to eliminate the most extreme cases such as the inclusion of criada (f.) 'maid' as a meaning for mujer, and hembra 'female' has been replaced by senora 'lady', offered along with the new meaning joven (f.) 'young woman' and esposa 'wife', which was included in Word 6.0. Nevertheless, the same synonyms previously given for hembra are now divided between senora and joven, and remain nearly the same for esposa: again we find media naranjasaid cara mitad, although costilla has disappeared; Venus is now capitalized and eva omitted, but these are more than compensated for by the inclusion of hurt 'houri',15 beldad 'belle' and hermosura 'beauty', so confirming, this time explicitly and brazenly, that the only side of joven that is taken into consideration is the erotic-aesthetic one. The existence of older women is still ignored (neither vieja or anciana is listed), even though this time aduUa is found among the synonyms (inherited from hembra) for senora.
164
GENDER, LANGUAGE AND NEW LITERACY
So, however, is madre, which is given a social treatment in accordance with its traditional association as a cult object (a link not made in the cases of hombre, senor or padre). Asymmetrically, mujer is offered as an antonym for the entry hombre and, in addition to possessing among its meanings the generic persona 'person' and its corresponding synonyms, neither this nor the other three meanings listed for hombre coincide with those of its counterpart, mujer. Here, in place of social status (see senora above), gender distinction has been taken into account with the inclusion of varan '(human) male', although, incomprehensibly, among the synonyms we find cabattero 'gentleman'. As regards distinctions based on age, rather than maintaining, as with mujer, the invariable joven (the determiner indicating gender opposition), here the accent has been put on childhood, with nino 'little boy' given as a meaning and joven and adolescent* listed as its synonyms. There is a degree of correspondence, however, with the feminine term, in that esposo is given as a meaning, despite the shock (and hilarity) of finding as its synonyms tipazo 'person with a good figure' and castizo 'person having national or regional characteristics'.16 Similar inconsistencies are found with the pair hembra 'female' - varon '(human) male' (which we classified in §2 as degrading). At first glance, the amount of information offered seems to have increased over that of the previous version. Retained are the meanings mujer and hombre, but now with only five and six synonyms given for them respectively, in place of the eight and nine previously. While it is true that the number of semantic distinctions has been increased, their relevance is far from clear and, in some cases, there are obvious errors. Hombre 'man' is offered as an antonym of hembra 'female' in place of macho 'male' or even varon' (human) male';17 no less infelicitous is the case of senora 'married woman, lady', which is listed as an antonym for varon. There is little point in specifying that hembra is an adjective, or in presenting tomillo 'screw' as an antonym of hembra,™ without further clarification. Furthermore, it would have been better not to include as a 'related expression' (only) to varon the epicene ser 'being', or madre (among whose synonyms is doncetta 'virgin; handmaiden'(!)) as a 'related expression' to hembra. The term criada 'maid' does not appear as an entry in Word 11.0 although it does as a synonym of siruienta' (female) servant' (see note 14). As in the case of criado 'manservant', these entries seem to be rather more refined than in Word 6.0 (at least maritornes, fdmula and fdmulo have been eliminated). The term criado, however, apart from its meaning of sirviente, constitutes a clear illustration of the structure underlying Word 11.0's Thesaurus, which is based on a highly peculiar concept of polisemy, to say the least. Rather than restricting itself to showing the formal link between the noun criado 'manservant' and the past participle of the verb cnar'to breed, to raise (children or animals)', from which it is etymologically derived, the programme fills the listing for criado with all of the meanings and synonyms given for the term criar (what appears to have happened is that a derivation engine provided the listing of those past participles which coincide formally with the noun criado).
SPANISH GUIL
165
It is not indicated, however, that these forms are past participles, but rather (under the category of 'meaning') that they are verbs (V). In their lists of synonyms the verb from which these are derived is generally specified after each, in the infinitive, e.g. 'ocasionado (de ocasionar)'. It is striking indeed that for the entry esposa 'wife', the verb esposar 'to handcuff figures as the first meaning, with the meaning of sujeta 'restrain' and its corresponding synonyms, all of these in the second person imperative and with the corresponding infinitive given, followed immediately by the antonym libera 'liberate', in the same person and mood. The same occurs with esposo 'husband', where this time, asymmetrically, the first meaning given is that of marido 'husband' and as the formal relation here happens to be with the first person present indicative it is in this person that the synonyms are given. It is clear that the conjugation paradigm of the listed verbs has been introduced and items coinciding in their external formats have been matched within any grammatical category, reciprocally transferring their semantic structures, and independently of any other lexicographical consideration. In this way, enormous catalogues of forms are created which are superfluous, when not actually counterproductive, and result in the user feeling crushed beneath an avalanche of data which, in order to be used, must moreover be kepi simultaneously open on the screen and gradually scrolled down. This same automatic treatment is applied to convert masculine lexemes into feminine ones. For example, among the synonyms displayed for the entry esposa (defined as esposa (s.)), we find 'novia (de novio), companera (de compane ..., [sic]', with an indication in parentheses of the term from which it is derived and even with the predictable termination replaced by ellipses. This is surprisingly meticulous, if we compare this with the failure to show information related to possible linguistic varieties. This insistence on making explicit the derivation of feminine terms from masculine is, however, interpretable as a reflex of our androcentric culture in that, the routine having been established, the same method of dependency presents us with, for example, 'hembra (de macho)' (entry: esposa), "matrona (de patron)' and 'dama (de caballero)' (both under the entry senora), which, at least from a morphological standpoint, is totally absurd. Despite this automatic inversion, multiple asymmetrical aspects may be pointed out in the structuring of entries (only with feminine terms is a reference made to the other half of a gender pair - see, for example, hija — hijo, soltera - soUero, jefa - jefe, directora - director), as well as in purely lexical problems (e.g. listed as a meaning for hija we find a very strange noun, producta, which is simply a version with -a ofproducto, the corresponding listing for hijo, the same operation is performed with prelado 'ecclesiastical official' (term: jefe), resulting in prelada, a post which would certainly not be found in the church, at least not in the Catholic Church; in another case, the mechanism is not applied but, nevertheless, the masculine doncel appears among the synonyms for soltera). We will not, however, spend more time on this, as it seems merely anecdotal in comparison with what was still to come in our analysis. We would like, in any case, to at least state that we have found no
166
GENDER, LANGUAGE AND NEW LITERACY
terms which may be considered salacious or offensive or even simply related to sex (at least the most common ones that occurred to us), a fact that does reveal some degree of (self-) censorship. With the intention of widening the corpus, terms for professional tides which had not been considered in the Spanish section of Bazzanella et al (2000) - were searched on, and even nouns for high-ranking occupations were found, e.g. ingeniera, jueza. Nevertheless, the fact that the listings for these two entries began with the masculine form for each, as well as the symmetry of their structures, led us to suspect that this was the result of a simple automatic conversion. To judge the extent to which automated processes have been employed in the Thesaurus (with little or no revision of results), we decided to search for a series of lexemes, this time derived by alteration: mujerdta, mujerciUa, mujerona, mujerzuela, hombrecito, hombrecillo, hombron, hombreton. We found to our horror that the search process consisted simply of taking the original term19 (in this case mujeror hombre) and, without any type of precaution, applying a derivation engine which added the corresponding suffix to the meanings and synonyms. It is no wonder, then, that we find derivatives which, although possible according to the system, have not been approved by the norms of usage (madrita was listed for madre, rather than 'madrecita\ along with the similarly unusual madrittaand madruela); that altered terms appear which, for their belonging to a formal register would not easily admit this derivation (from consarte. consortita, consortilla, consartuela; from conyuge: conyugita, conyugiUaand even the impossible *conyuguela)\ that for the entry mujerzuela we find enumerated all of the meanings and synonyms for the entry mujer altered by the pejorative suffix -uela, however absurd these may sound ('damuela (de dama), duenuela (de duena), amuela (de ama), matronuela (de matrona)'), and without anywhere indicating dial this alteration, mujerzuela, has become lexicalized, taking on the specific meaning of 'prostitute'. Nor are the alterations derived from hombre acceptable, e.g. the absurd mortalito or mortalitto from mortal 'mortal', projimito or projimillo from projimo 'neighbour', semejantito or semejantiUo from semejante 'peer', or the impossible tipacitoor tipacillofrom tipazo, that strange synonym for esposo which we saw earlier. The last term searched for was the form hermosuruela, which we had found earlier as a synonym of mujerzuela, and which reproduces the entry hermosura '(a) beauty', with the addition of the suffix -uela. It confirmed for us that it would be a waste of time to take our analysis any further: we now faced a veritable catalogue of nonsense, of nearly indecipherable derivatives (esteticuela from estetica, idealuelofrom ideal), without the slightest respect for the phonotactic restrictions of the language (*beldadzuela from beldad, *divinidadzuela from dhrinidad), in short, an authentic assault on the Spanish language.
SPANISH GUIL
167
5 Conclusio n Although some small effort seems to have been made on this score, the principal source of symmetry in the treatment of women and men seems to come from the automatic conversion of masculine nouns into feminine ones. However, the lack of revision of the results obtained, along with the absence of defined lexicographical criteria (apart from the systematic censorship), are responsible for the extreme unreliability of this tool, which - as was the case with version 6.0 - selects and orders meanings and synonyms arbitrarily and randomly, and in fact provides erroneous data. The greatest difference with respect to the previous version lies in the apparent application of conversion and derivation engines which increase - at times excessively - the amount of data offered. This notwithstanding, the lack of control over the resulting product could lead to its being considered misleading by the user. Notes 1 To work toward the unity, integrity and growth of the common language' (see www.rae.es). 2 This 'tool* ('Synonyms: Spanish (Spain - international alphabet)') is identified by the programme itself as 'Motor de sinonimos ['Synonym Search Engine'] Copyright © 1998-2002 de SIGNUM Cia. Ltda. Quito, Ecuador'. 3 Such as the pair esposo - esposa or the invariables el/la consorte, el/la conyuge. 4 To summarize, the acceptable pairs should be: hombre - mujer, varan - mujer (in Spain, since 1993), macho - hembra (for animals and other organic species). Sadly, macho, referring to men, is one of the loan words from Spanish into other languages. Keep in mind that, based on stereotypical Hispanic sexism, in colloquial language and especially among boys, macho is often used as a laudatory appellation, i.e. connoting qualities such as 'bravery', 'strength', etc., which are traditionally attributed exclusively to men. 5 At the website of the Institute de la Mujer, under http://www.mtas.eS/mujer/3.pdf, corresponding to dossier 3 of La mujer en el franquismo y la transition espanola ('Women under Franco and the Spanish Transition'), some interesting information can be found on this topic. 6 For the definition of 'meaning' and 'synonym', see the Introduction. 7 In the following examples, the meanings are in bold, whereas the items given for each meaning ('synonyms') stand in brackets, in italics. 8 Remember that the term hembra is applied to both women and to animals of the female sex, while men, apart from the corresponding macho, have also a tenn exclusive to them, varon (see §2 and note 4). 9 Maritornes is the proper name of a maid in Don Quixote, defined lexicographically as criada tosca ('an uncouth maid') (Moliner 1996) and moza deservicio, ordinaria, feay hombruna ('serving girl, ordinary, plain, mannish") (DRAE 1995). 10 At least in the Peninsular European variety. 11 In Spain, for a woman to refer to her husband as hombre or mi hombre ('my man') carries with it a connotation of vulgarity. By contrast, as we have noted, the middle and upper classes typically use mujer or mi mujer ('my woman') for 'wife', the lower social classes using mi senora.
168
GENDER, LANGUAGE AND NEW LITERACY
12 Alluding, of course, to that of Adam, from which Eve was created. 13 La aparicion de terminos aparentemente inapropiados esfortuita. 14 It is true that criada no longer appears (although criado does), but we understand this to be a simple error of omission, as it figures as a 'synonym' for other entries, in the case of the alternative sirvienta. 15 'Beautiful woman which in the Islamic religion is promised to the blessed in Paradise'. Although the Arab influence on Spanish culture is unquestionable, the reader should bear in mind that the Arabs were expelled from Spain as long ago as 1492. 16 Also, in Mexico, 'mestizo* (DRAE). 17 See also the observations in §2 regarding the relationships of the pairs hembra vardn/hombre, mujer- varan/hombreand hembra- macho, as well as notes 4 and 8. 18 Observe the 'delicacy' of the metaphor on which these denominations are based; it seems to be widespread, however, as we find it in other languages as well. 19 This explained why the masculine term appeared first, as meaning, for the feminine terms searched for.
References Ambadiang, Theophile (1999), 'La flexion nominal. Genero y numero', in IgnacioIgnacio Bosque and Violeta Demonte (eds), Gramdtica descriptiva de la lengua espanola, Vol. 3. Madrid: Espasa, pp. 4843-913. Bazzanella, Carla; Camugli-Gallardo, Catherine; Guil, Pura; Manera, Manuela and Tejada, Paloma (2000), 'Categorizzazione del femminile e del maschile nelle nuove tecnologie: prime ricerche nel Thesaurus italiano, spagnolo, francese, inglese di Wonf. Cuadernos defilologia italiana, 7, pp. 193-245. DRAE (1995) = Real Academia Espanola, Diccionario de la lengua espanola (DRAE). Edicion electronica, version 21.1.0. Madrid: Espasa-Calpe. Garcia Meseguer, Alvaro (1994), ^Essexista la lengua espanola? Barcelona: Paidos. Garcia Mouton, Pilar (2003), Asi hablan las mujeres. Madrid: La Esfera de los Libros. Moliner, Maria ([1970] 1996), Diccionario de uso del espanoL Edicion electronica. Madrid: Gredos. Nissen, Uwe Kjaer (2003), 'Gender in Spanish: tradition and innovation', in Marlis Hellinger and Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 3. Amsterdam and Philadelphia: Benjamins. pp. 251-79. Real Academia Espanola ([1973] 1989), Esbozo de una nueva gramdtica de la lengua lebgyea espanola. Madrid: Espasa-Calpe.pe
11 Gender on-line in Hebrew: new technology, old language ZoharLivnat
I Introductio n According to Almog (2002), the gender-related qualities of Hebrew are noteworthy due to their intensity: Hebrew repeatedly, and even incessantly, marks the gender of addressees and selects the masculine signifier as the main signifier that also incorporates the feminine referent. Hebrew lays the groundwork for a Weltanschauung in which the constant need to distinguish between men and women plays a central role and at the same time the masculine acts as the model and the main signifier. One example of die nature of Hebrew can be found in the word adam 'man'. This is the first name of the first man, which in and of itself explains the source of the word's ambiguity. On the one hand, it refers to the human race as a whole (benadam= 'descendant of the first man'), while, on the other hand, it refers to the males of this group.1 This is also the term used in Israeli law to denote any human being. According to Ariel & Giora (1988), the word adam reflects the precedence the masculinity takes over the femininity. Despite the fact that humans come in two forms (male and female), the male form of existence is perceived as the standard, prototypical existence. This is in keeping with the tradition of the creation of man - the man as the prototype from which woman is created2 - and this continues to be the case in usage of Modern Hebrew. Ariel & Giora (ibid.) point to the differences between the following sentences: A. B.
Ha 'adam ha 'modemi kvar yachol litrom mizar'o lema 'an acherim Modern man can already donate his sperm for others. Ha 'adam ha 'modemi kvar lo noheg lehanik kmo badarot hakodmim Modern man does not breastfeed to the same extent as in previous generations.
The first sentence is acceptable because it presents a solely male experience as if it were universal. The second sentence sounds odd, as it presents a purely female experience as a universal one. It seems that it is easier for us to perceive the male experience, and not the female experience, as representative of all humanity.
170
GENDER, LANGUAGE AND NEW LITERACY
In the words of Elior (2001), Hebrew speakers are 'consciously or unconsciously inheritors of a religious, patriarchal and gender-oriented world of thought'. What, then, will be the fate of Hebrew in an evolving society? How can an ancient language born, shaped and preserved over centuries in a religious, conservative and closed society adapt to the needs of the modern Israeli society? And how will this be expressed in an innovative electronic tool such as a computerized thesaurus? The next section will provide a brief and schematic overview of the morphology of feminine inflection in nouns and of gender as a lexical phenomenon in Hebrew. Section 3 presents evidence of social change, while sections 4-5 concentrate on the findings of the study of the Thesaurus in the Hebrew-enabled version of MS Word (XP 2002), which is the focus of the current volume. 2 Gende r construction in Hebrew Hebrew is a 'gender language' which has two gender classes - feminine and masculine. The syntactic rules, which require agreement between nouns and words of other categories - verbs, adjectives, participles, pronouns and numerals - demand that every noun, animate or inanimate, has a grammatical gender, either 'necessary' or arbitrary. Most of the nouns in Hebrew belong to one class of grammatical gender. Very few nouns belong to both classes, and when this happens, it is mainly the result of changes over the long history of the Hebrew language. Examples are sakin 'knife', matbe'a 'coin', ruach 'wind', derech 'way', shemesh 'sun', and kos 'glass'. Very few nouns have two completely unconnected meanings, and each meaning is marked by a different gender, e.g. etzem (f.) 'bone' - etzem (m.) 'substance'. In this chapter, we are primarily interested in the category of animate nouns, and we will leave out the fascinating psycholinguistic question of gender of inanimate nouns (Guiora et al 1980, 1982). Inanimate masculine nouns are frequently used as the base for derivation of feminine nouns (e.g. me'il (m.) 'coat' - me'ila (f.) 'embezzlement'; gav (m.) 'back' - gaba (f.) 'eyebrow'), though the semantic relations between the two words are often unpredictable, far from automatic, and generally should be viewed as words that were each derived independently (Schwarzwald 2002: 39). Theoretically, 'every masculine [+ animate] noun, adjective, or participle can be inflected for gender, and the feminine form is created by the suffixation of +a, +t, +it, or +et /+at' (Schwarzwald 1991: 1410). Examples include: ikar 'farmer' (m.) - ikara 'farmer' (f.), safran 'librarian' (m.) - safranit 'librarian' (f.), zamar'singer' (m.) - zameret 'singer' (f.). Schwarzwald also illustrates the morphophonological mechanisms that determine which suffix is chosen in any given case. For example, in the case of monosyllabic nouns and nouns that end with the closed syllable U, the feminine suffix is +a; loan words are always inflected with the suffix +it, and so on.
HEBREW LJVNAT
171
While there are exceptions to some of the rules, they still allow native speakers to derive the feminine form from the masculine form freely and, most of the time, without hesitation. As stated, Hebrew is a language with 'maximum gender loading' (Guiora 1984). Grammatical gender exists in Hebrew in noun categories; participle forms (shomer 'keeping' (m.) - shomeret 'keeping' (f.)); adjectives (gadol 'large' (m.) - gdola 'large' (f.), ishi 'personal' (m.) - ishit 'personal' (f.)); verbs (in the past tense: katav 'wrote' (m.) - katva 'wrote' (f.), in the future tense: yichtav 'will write' (m.) - tichtov 'will write' (f.)) and numbers (shalosh 'three' (f.) - shlosha/shloshet 'three' (m.)3). Lexical gender which does not correspond to grammatical gender exists in Hebrew in categories of pronouns (in second and third person, singular and plural) and demonstrative pronouns (in the singular) (Table 11.1). In all of the cases listed in Table 11.1, the feminine form is not achieved through regular inflection of the masculine form. Lexical gender can be also found to a small extent in nouns. In some cases, the feminine form is not derived from the masculine form, but has a complementary distribution. These cases will not be listed in full here, though we will raise a number of the most important: 1 One case is the pair zachar - nekeva 'male' - 'female'. These words are biblical from the story of the creation of the world, and in Modern Hebrew they are used to objectively relate to the biological differences between the sexes among human beings as well as animals. However, it is important to note that the word nekeva is derived from the root nekev ('aperture'). According to Brenner (1997: 12), this indicates that from a social perspective there is no difference between a women's biological and social functions. Due to this emphasis of the biological aspect, in low register colloquial speech, the word nekeua also carries a crude sexual and scornful connotation. 2 Another pair is chatan - kala 'groom' - 'bride'. In this case, the form kola is related through the root to 'completeness' or 'wholesomeness', and in biblical Hebrew it denotes beauty and is used as a term for a beloved woman. 3 The forms of address used for men and women are also realized through lexical gender. The form of address for men in Modern Hebrew is through the tides of mar'Mr' or adon 'sir'. The word adon is a biblical word
Table 11. 1 Masculine and feminine pronouns in Hebrew ata 'you' (m. singular) atem 'you' (m. plural) hu 'he' /im'them' (m.) ztf'this' (m.)
at* 'you' (f. singular) aten 'you' (f. plural) hi 'she' hen 'them' (f.) zot, zo'this' (f.)
172
GENDER, LANGUAGE AND NEW LITERACY
meaning 'ruler', from which even the term used to refer to God is derived - adonai 'God'. The word mar is borrowed from the Aramaic, and the noun marut 'authority1 is derived from it. The feminine forms adona 'master' (f.) and adonit 'master' (f.) can be found in some Jewish sources, but are not accepted in Modern Hebrew. Instead, the common forms are moral 'Mrs' or gveret 'Mrs'. Morphologically, gueretis the feminine form of gever'man' (see §5 below). The common forms of polite address are adoni 'my master' (m.) and guirti 'my master' (f.) used, for example, in court settings. 4 Complementary distribution also exists in the designations of some professions. The pair chayat 'tailor' - toferet 'seamstress' is one such example. The masculine form chayat does not have a corresponding feminine form, and the feminine form toferet does not have a corresponding masculine form, although morphologically it would not be difficult to derive them. Another example relates to the cooking profession, a profession whose status has changed markedly in recent years. Women, who traditionally filled this role, were called mevashelet 'cook' (f.). This form does not have a corresponding masculine form, though along with it we have tabach 'cook' (m.), which has a corresponding feminine form, tabachit 'cook' (f.). The word tabach is primarily used in Modern Hebrew to denote someone who works in a military kitchen - a place where cooking is done solely by men. In contrast, men who cook at restaurants serving haute cuisine are referred to with the word s/^/'chef (m.) - a loan word from which the feminine form shefit 'chef (f.) is derived. 3 Socia l gender and social chang e As stated above, Hebrew allows automatic derivation of the feminine form from the masculine in animate terms, though this hypothetical potential is not necessarilv realized in any given case. Reflecting the social reality, many nouns exist solely in the masculine. The words abir 'knight', afifior 'Pope', matzbi 'military leader' (m.) and numerous others do not have corresponding feminine forms, for obvious reasons. In contrast, there are also words that exist only in the feminine form. There are professions which in the reality of Israel today are filled exclusively by women. These include shinanit 'dental hygienist' (f.), saiyaat 'dental assistant' (f.), kosmetikayit 'cosmetician' (f.), manikuristit 'manicurist' (f.), katvanit 'typist' (f.) and omenet 'nanny' (f.).5 All of these words end with the feminine suffix, which morphologically allows for corresponding masculine forms to be created, but it does not appear that these forms exist in Modern Hebrew. Yet, Israeli society is undergoing change. The status of professions or trades which in the past were the exclusive domain of women or men has changed and they have opened up to members of the other sex. Some evidence will be provided for social changes which have given rise to the need for linguistic change as well.
HEBREW LIVNAT
173
1 The term akeret bayit 'housewife' denotes a woman who does not work outside the home, and it never had a corresponding masculine form. However, in recent years there are more couples in which the woman is the partner who works outside the home and the man does the housework. As a result, the term akar bayit 'homemaker' (m.) has come into use. 2 The crucial pair of ganan - ganenet. The feminine form, ganenet, denotes a woman who teaches children of kindergarten age. In contrast, the masculine form ganan signifies a 'gardener', a person who works in a garden. When only several years ago men began working in childcare, they had a problem using the masculine term ganan as a corresponding term for ganenet, because it was already established with another meaning. Therefore the term ganen was proposed. The Academy of the Hebrew Language was called in, and in 1998 it discussed the issue and decided on ganan or the term ganan yeladim (literally: 'children's teacher'). In popular usage one can also find the term ganenet mimin zachar 'a male kindergarten teacher (f.)'. 3 Another example is the profession mozeg 'bartender' (m.). Despite the hypothetical existence of the feminine form, the word mozeget 'bartender' (f.) was never part of the mental lexicon of Hebrew speakers. Women have entered the profession in recent years, though they are referred to by the word bannenit 'barman' (f.) - the feminine form of the loan word barmen 'barman' (m.). 4 Titles and professions from the religious world are worthy of special note. In religious social frameworks, many roles are filled only by men, not only by tradition, but also in accordance with more strict guidelines. Professions such as gabbai 'synagogue treasurer' (m.), kavran 'undertaker' (m.), and shochet 'ritual slaughterer' (m.) are not open to women, and there are, therefore, no corresponding feminine terms. Likewise, the words avrech, hesdernik or mekubal, all terms from the religious or ultra-religious world, do not have corresponding feminine forms. The word avrech in Modern Hebrew denotes a student in a Yeshiva, an institution open only to men; hesdernik is a student in a hesder yeshiva (men who combine yeshiva studies with military service); mekubal is a man who studies the wisdom of the Kabbalah. These are all areas that are completely closed to women. However, the religious Jewish world is not uniform, and it is also divided into different groups. In some of them, there is currently a struggle for greater equality between men and women, and the relative success of this struggle, at least within a limited group, is starting to make an impact in the linguistic sphere. The Israel Movement for Progressive Judaism currently trains women as rabbis, cantors and ritual circumcisers. The terms used to refer to women that fulfil the role of cantor and ritual circumciser are the feminine forms derived automatically from the masculine forms: chazan 'cantor' (m.) - chazanit 'cantor' (f.); moheVritual circumciser' (m.) - mohelet 'ritual circumciser' (f.). The case of raw 'rabbi' (m.) is somewhat more complex. The feminine form rabanit exists in Hebrew, but refers to the wife of a rabbi. Though for a female rabbi the Academy of the Hebrew
174
GENDER, LANGUAGE AND NEW LITERACY
Language decided on the feminine form raba, in fact, women who serve in this role decide whether they want to be called rav or raba on the basis of their personal preference. 5 A preference for being addressed with the masculine form is by no means unusual. Even in the secular world, many women in executive positions prefer to use the masculine form nasi 'president' (m.), mankal 'CEO' (m.), yoshev rosh 'chairman' (m.) or dekan 'dean1 (m.), instead of the feminine forms nesia 'president' (f.), mankalit 'CEO' (f.), yoshevet rosh 'chairwoman' (f.)6 or dekanit 'dean' (f.) - despite the fact that all of the latter are common and accepted terms (Schwarzwald 2002:12). 6 Another interesting case is the pair mazkir 'secretary' (m.) - mazkira 'secretary'' (f.). While the words nasi 'president' (m.) and nesia 'president' (f.) are similar in all respects aside from gender, there is a world of difference in the meanings of the words mazkir and mazkira. The feminine form, mazkira, is generally related to a simple clerical position, without authority and which does not require any special skills.7 In contrast, the masculine form mazkir generally denotes a key role in an organization; thus women in this position frequently prefer the masculine title. 4 Som e characteristics of the MS Word Thesaurus in Hebrew In order to see if and how feminine forms are found in the thesaurus, we first had to determine which words exist in the language both in the masculine form and in the feminine form derived from it. Since feminine forms are rarely found in Hebrew dictionaries, this determination was primarily based on intuition. The previous section listed examples of words in the masculine form that do not have corresponding feminine forms for social reasons. Despite the several changes discussed, even in secular Israeli society many professions and positions are still not very open to women. Consequently, while professions such as chovel 'sailor' (m.), kore 'miner' (m.), me'avteach 'security guard' (m.), kablan 'contractor' (m.), kvarnit 'captain' (m.), sharat 'janitor' (in.), chablan 'sapper' (m.) and numerous others have hypothetical feminine forms, it is doubtful that they exist in the mental lexicon of native speakers of Modern Hebrew. Other instances are nouns that are for the most pan used genetically. Nouns used genetically will always appear in the masculine form. A case in point is the word ezrach 'citizen' (m.). In Hebrew, generic sentences are structured using the singular or plural masculine form. Consequently, while the masculine form is very common, the feminine form is rare. Although all speakers know how to derive the form, they may have a difficult time recalling whether they have heard it recently and may think that it sounds odd. As no psycholinguistic research of this type has been done on Hebrew, we have included these words in our study despite the reservations and created a list of words that have a minimal masculine-feminine pair. The Thesaurus in Word contains close to 1,000 such words, and it was found that in only 3 per
HEBREW LTVNAT
175
cent of the cases does the corresponding feminine word also appear in the thesaurus. The feminine form appears in two main cases: 1 Kinship terms, such as av 'father' - em 'mother', dodan 'cousin' (m.) dodanit 'cousin' (f.), achyan 'nephew' - achyanit 'niece', choten 'father-inlaw' - chotenet 'mother-in-law', cham 'father-in-law' - chamot 'mother-inlaw'. 2 In cases where the meaning (or meanings) suggested for the feminine form is different from that suggested for the masculine form. The following are several examples: (1)
(2)
(3)
megarash - megareshet. Both forms mark a person who has been exiled, except that according to the Thesaurus, the masculine form megorash means a person who has been exiled from his country for political reasons, and the feminine form megvueshet is a synonym for grusha 'divorcee' (f.), meaning a woman who has been divorced by her husband.8 metapel - metapelet. The feminine form metapelet generally refers to childcare, and that is how it is found in the Thesaurus, the synonyms are o per 'au pair' and omenet 'nursemaid'. In contrast, the masculine form metapel is presented in the Thesaurus as a synonym for terapist 'therapist' (m.) and psichoterapist 'psychotherapist' (m.) - synonyms not suggested for metapelet. mechashef- machshefa (or mechashefa). Originally, these two forms denoted someone who practises magic. The Thesaurus suggests five synonyms for this meaning of the masculine form mechashef. However, this meaning is not found for the feminine form machshefa, despite the fact that this is the biblical meaning of mechashefa and the fact that in the history of the Western world it was women who were suspected of practising magic. Instead, for the feminine form the Thesaurus suggests the meaning mirsha'at 'shrew' and the synonym klavta 'bitch' - a negative meaning which was added to the original meaning of the word due to the negative attitude towards magic.9
Worthy of mention are the areas where feminine forms are heavily represented and corresponding masculine forms are lacking: The entry zona 'prostitute' offers six synonyms, some high register and some low register: prutza 'whore', zanzonet 'little slut', najkanit 'harlot', yatzanit 'prostitute', sharmuta 'slut', kdesha 'temple prostitute'. The words pilegesh 'mistress' (f.) and me'ahevet 'lover' (f.) are also included, though the masculine form me'ahev 'lover' (m.) is not included.10 Furthermore, there is a strong presence of words and phrases used to describe women physically or their status as women who are loved by men. These are words from all registers, beginning with the biblical and poetic style: yefat chen 'charming beauty', yafa kalevana 'beautiful as the moon', tzviya
176
GENDER, LANGUAGE AND NEW LITERACY
gazelle', ya'alat chen 'graceful gazelle', ayelet ahavim 'beloved gazelle', yona tama 'beautiful dove', and through low register: chaticha 'babe' (f.) (the masculine form chatich 'hunk' does not appear, although it is quite common in speech), shafa 'hot chick', ptzatza 'bomb' (f.), ptzatzat min 'sex bomb' (f.), chatulat min 'sex kitten' (f.). 5 Crucia l pairs in Word Thesaurus This section presents some of the minimal pairs that appear in the Thesaurus. The most interesting comparison is between the word isha 'woman' and its masculine equivalent. The word isha creates a minimal pair with three different words: 1 ishn - this is a word belonging to the masculine class of grammatical gender that is used to denote male humans as well as to denote a person without denoting gender, for example in sentences like ein sham ish 'there's no one there', ish lo hitlonen 'nobody complained'. 2 gever- this word denotes only human males. It is accepted as an 'objective' word in denoting a human being, for example in scientific contexts. However, it should be noted, that by its root it is related to physical strength and braverv (gibor chayil 'military hero', gvura 'braverv'), and in certain contexts may have a sexual connotation, for example if a woman says ani tzncha gever 'I need a man' (Lavi et al. 1982). 3 ba'al- this word is the common word in Modem Hebrew to denote a man who is married to a specific woman ('husband'). However, it also has the meaning of ownership of property (ba'al mechonit 'car owner') or spiritual assets (ba'al ekronot 'holder of principles'). In terms of the root (and also in terms of Jewish law), it is also connected to the verb denoting sexual intercourse (liv'ot), when it is described from the perspective of the man towards the woman.12 Let us now look at how these entries are presented in the Tfasaurus as compared to the entry isha. The Thesaurus provides two meanings for the word w/w:13 (4)
isha: re'aya (re'aya 'wife', estiet cheiko 'wife of one's bosom', eshet ne'urim 'wife of his youth'); nekev a (nekeva 'female1, bat adam 'Adam's daughter', bat chava 'Eve's daughter').
The first meaning is presented as the equivalent of ba'al 'husband'. The second meaning is the more general meaning, emphasizing both gender and the attribute of being 'part of the human race', similar to ish. The word ish 'man' also appears in the Thesaurus with two meanings, though the differences between the two entries are significant: (5)
ish: adam (adam 'man', enosh 'person', benadam 'Adam's son', nezer
HEBREW LTVNAT
177
habr'ia 'lords of creation', ben adam 'Adam's son', ben-tmuta 'mortal', basar vadam 'flesh and blood', nefesh chaya 'living soul', krutz chomer 'formed out of clay', holchey al shtayim 'biped', ben enosh 'human being'); barnas h (barnash 'guy', zachar 'male', gever 'man', adam 'man', briya'creature'). The first meaning is the more general, denoting the human race and presenting 12 synonyms, some high register, such as nezer habri'a 'lords of creation'. The second meaning seemingly emphasizes the masculine aspect of the word. It suggests five synonyms, three of which indeed express this aspect: zachar, gever and barnash. However both of the other synonyms, adam and briya, are words that again denote a more general meaning, without any masculine marking. The word ish appears with a total of 17 synonyms, almost all of which denote the general attribute of belonging to the human race. The word isha, in contrast, appears with only six synonyms, three of which, that are presented first, denote the woman's role as wife and partner to man. The comparison with the entry gever is also interesting: (6)
gever. gever (ish 'man', zachar 'male', adam 'man', briya 'creature', barnash 'guy').
Here only a single meaning is presented, and the synonyms are the same as those suggested for the second meaning of ish - a mix between words that denote the human race and words that emphasize the male base without any distinction between them and without any clear priority. As stated earlier, the feminine form gueret 'Mrs' is morphologically derived from the word gever, though the word is only used as a tide and does not denote the female sex. The synonyms for gueret appearing in the Thesaurus are titles in other languages: madam, lady, Mrs and senora. Another interesting comparison is between ben 'boy, son' and bat 'girl, daughter'.14 The word ben is polysemous and denotes both kinship relations (each person relative to his/her parents), either a young person or a male child, and other shades of meaning related to various kinds of relation. This can be seen in phrases like ben tarbut 'cultured person' (m.), ben mavet 'mortal' (m.) and numerous others. The Thesaurus suggests two meanings for the word ben: (7)
ben: yeled (yeled 'boy', yaldon 'little boy', taf 'children'); tze'etz a (tze'etza 'descendant', zero 'sperm', netzer 'scion', choter 'scion', pri beten 'fruit of the womb').
The first meaning emphasizes both the feature of 'youth' and the genderrelated aspect, although the word to/is generally used to denote a group of children without respect to gender. The second meaning suggested here has to do with kinship. The words netzer and choter arc metaphorical expressions
178
GENDER, LANGUAGE AND NEW LITERACY
borrowed from the semantic field of plants. They are not only considered flowery language, but also have a special connotation that suggests descent from an especially respected family. This meaning is based on the biblical verse that refers to the future coming of the messiah, who will be a descendant of the family of King David.15 In contrast, the Thesaurus entry bat offers a single meaning and ignores both the issue of kinship and the additional meanings that appear in other dictionaries, for example in the phrase chevrat bat 'subsidiary'. As opposed to the nine synonyms listed for ben, the entry for bat offers only two synonvms: (8)
bat. tinoket 'baby' (f.), yalda 'girl'.
The differences in amount of detail of the entries in other examples from this semantic field can also be seen here: (9a)
tinok 'baby' (m.): olal 'infant' (m.), yonek 'suckling' (m.), yanuka 'baby' (m.), yelod 'newborn' (m.), harach hanolad 'tender newborn baby' (m.). (9b) tinoket 'baby' (f.): bat 'daughter'. (lOa) sav 'grandfather': sab a (saba 'grandfather'); yashish (yashish 'elderly man', sav 'grandfather', zaken 'old man', kashish 'elderly man', ba bayamim 'man along in years', svayamim 'man getting on in years'). (lOb) sava 'grandmother': savta 'grandmother'. Other types of differences can be seen in a group of words from an entirely different semantic field - the relations between Israel and other nations. The word goy has two meanings. Its biblical meaning is 'nation', and the meaning it took on later, and which is the common meaning in Modern Hebrew, is 'gentile, a nonjewish person'. Both of these meanings can be found in the Thesaurus, even though the division is not distinct: (lla) goy: am (am'nation', team'nation', uma'nation'); umothaolam (umot haolam 'nations of the world', nochri 'a man from a foreign place', arel 'uncircumcised male', ben nechar'foreigner' (m.)). (lib) goya: shikse 'noriJewish woman', arelit (literally:) 'uncircumcised female', shiktza 'abomination' (f.). Of the synonyms suggested for the second meaning, the words nochri and ben nechar can also be used to denote anyone foreign, born in a distant land. Only the word arel, meaning 'a man who is uncircumcised' is a traditional Jewish way of referring to a gentile, meaning that emphasized the 'non-Jewish' character and has a negative connotation. In contrast, for the feminine form goya only one meaning is provided, that of a 'nonJewish woman'. The Thesaurus suggests three synonvms: shiktza originally meaning 'abomination', a word with a very strong negative connotation. The word shikse is actually the same word with a Yiddish
HEBREW UVNAT
179
pronunciation. The word arelit is the feminine form of the arel 'uncircumcised man', and as the word arel was originally used to refer to a physical quality that is not relevant to women, the word arelit, then, can only refer to the 'non-Jewish' quality and carry the negative connotation. The following is a comparison of the Thesaurus entries for arel and arelit (12a) arel goy (goy 'gentile' (m.), nochri 'man from a foreign place', umot haolam 'nations of the world', ben nechar 'foreigner' (m.)); goyi (goyi 'non-Jewish', nochri 'man from a foreign place'). (12b) arelit goya 'gentile' (f.), shikse 'non-Jewish woman1, shiktza 'abomination' (f.). While the Thesaurus offers a list of neutral synonyms for arel, the entry arelit is filled with synonyms that have a negative connotation and can even be considered cursing. 6 Summar y of findings This section provides a brief summary of the findings of the study of the MS Word Thesaurus in Hebrew: 1 Presence. There is a marked prominence of male forms. It was explained above that this is attributable to the morphological structure of feminine inflection in Hebrew, where the feminine form is almost always automatically derived from the masculine form, meaning there is no justification for it to appear in the dictionary. However, the Thesaurus only provides the masculine form, even when this is odd. For example, for rakdan 'dancer' (m.), the Thesaurus offers the synonym balerina 'ballerina', a word which in fact can only be a synonym for the feminine form rakdanit 'dancer' (f.). On the other hand, feminine forms are prevalent in semantic fields related to physical characteristics of women or to sexual conduct. 2 Amount of detail When the Thesaurus contains both the masculine and feminine forms, systematically more detail is provided for the masculine form, from the point of view of both the number of meanings listed and the number of synonyms that appear for each meaning. Not a single example to the contrary was found. 3 Connotations. There is a tendency for the feminine forms to be marked by a negative connotation as compared to the corresponding masculine forms.16 While there are not many examples of this type in the Thesaurus (machshefa, goya and arelit), not a single example to the contrary was found. Moreover, the Thesaurus contains a not insignificant number of words that are actually curse words directed at women. Curse words directed at men also exist in spoken Hebrew, though these do not appear in the Thesaurus. Take, for example, the word shmuck which was borrowed from Yiddish and literally means 'male sex organ', but is rarely used with this meaning. This word is most commonly used as a negative term to refer to a
180
GENDER, LANGUAGE AND NEW LITERACY
man, in order to indicate stupidity, dishonesty or evilness, and this is how it appears in several dictionaries. The Thesaurus, though, presents this word as if it had only a single meaning, the original meaning, and completely ignores its negative connotation. 4 Prestige. When professions are marked by a varying level of prestige, the masculine form appears with the more prestigious meaning (metapel metapelet). Here, too, we did not find numerous examples, but did not find a single example to the contrary. Notes 1 Numerous expressions in Hebrew reflect this, for example shlosha dvarim marchivin da'ato shel adam, elu hen: dira naa, isha naa vekelim naim (Three things enhance a man's self esteem: a beautiful dwelling, a beautiful wife, and beautiful accessories') (Babylonian Talmud, Berakoth 57b, between the third and sixth centuries AD). 2 A discussion of the biblical story of the creation of the world told from this perspective is provided by the attorney Shulamit Almog as background for her proposal to refrain from use of the term adam ('man') in the laws of the State of Israel. 3 With numbers, it is actually the feminine form which is shorter, and the masculine form which has a seemingly feminine suffix. See Schwarzwald (2002). 4 This case appears to be the opposite of the others, as the masculine forms ends in +a, however historically this issue is more complex. 5 Although the masculine form omen appears in the Bible with the meaning of 'adoptive father'. 6 To maintain the focus of this study, we will not get into the specific morphological problems of acronyms (such as mankal) and of complex noun phrases (such as yoshevrosh). 7 When they first came out, answering machines were called mazkira elektronit "electronic secretary' (f.), a term which angered many speakers. In 1991, the Academy of the Hebrew Language decided on the term meshivon, which belongs to the male class of grammatical gender. 8 According to Jewish law, still applicable today and still current practice among most Jews in Israel, during the act of divorce, it is the man who divorces the woman. 9 In the study conducted by Lavi et al (1982), 90 per cent of the respondents defined the word machshefa as denoting a 'wicked woman', while none of the respondents understood the word mechashefas denoting an 'evil man'. 10 Similarly, see the absence from the language of the words grushah ('her divorcee') and almanah ('her widower'), in contrast to the accepted forms grushato ('his divorcee') and almenato ('his widow'). According to Ariel (1982), this means that the regular and basic status of being the 'divorcee of or 'the widow of is being a woman. That is to say that it is natural for a woman to be described as the divorcee or widow of someone, while this is not the case for a man. 11 The word isha looks like an automatic derivation of ish, and it appears that many speakers perceive it as such. This perception is reinforced in the biblical verse lezot yikare isha ki me'ish lukacha zot ('she shall be called Woman, because she was taken out of Man') (Genesis 2:23), which hints that the word isha was derived from ish. However, in etymological terms, isha is derived from a different root
HEBREW IJVNAT
181
12 For these very reasons there are women who abstain from using the common form ba'aU 'my husband' and prefer the form ishi 'my man', though they are a minority. 13 In the following examples, the meanings are in bold, whereas the items given for each meaning stand in brackets, in italics. 14 The word bat does not look like a derivation from ben, although that is the case historically. 15 Vayetze choter migeza yishay venetzer mishorashav yifre ('And there shall come forth a scion out of the stem of Jesse, and a scion shall grow out of his roots') (Isaiah 11:1). 16 In the study conducted by Lavi et ai (1982), respondents had a greater tendency to define the feminine forms para 'cow' and kalba 'bitch' as derogatory terms for women than the masculine forms par'bull' and kelev 'dog' (m.) as derogatory terms for men.
References Almog, Shulamit (2002),' "And these names are eternal" - Hebrew, Gender and Law'. Bar-Han Law Studies, 18,373-98. Ariel, Mira (1982), Three aspects of issues related to women and language'. Noga, 4, 10-13. Ariel, Mira & Giora, Rachel (1988), 'Sticking your tongue out'. Noga, 15,39. Brenner, Athalya (1997), The Intercourse of Knowledge: On Gendering Desire and >Sexualit\' in the Hebrew Bible. Leiden: Brill. Elior, Rachel (2001), '"Presence absence", "silent nature" and "a lovely maiden with no eyes": Addressing the question of the presence and absence of women in the holy tongue, Jewish religion and Israeli reality'. Alpaim, 20,214-70. Guiora, Alexander Z. (1984), 'The dialectic of language acquisition', in Alexander Z. Guiora (ed.), An Epistemology for the Language Sciences. Detroit: Wayne State University Press, pp. 3-12. Guiora, Alexander Z.; Beit-Hallahmi, Benjamin & Sagi, Abraham (1980), 'A crosscultural study of symbolic meaning'. Balshanut Shimushit, 2, 27-40. Guiora, Alexander Z.; Beit-Hallahmi, Benjamin; Fried, Risto & Yoder, Cecelia (1982), 'Language environment and gender identity attainment'. Language Learning, 32. 289-304. Lavi, Orit.; Porat, Orna & Feir, Pnina (1982), The attitude of Israeli society towards women as seen in language'. Noga, 4. 13-15. Schwarzwald, Ora (1991), 'Lexical weight in Hebrew inflectional feminine formation', in Alan S. Kaye (ed.), Semitic Studies in Honor of Wolf Leslau, Vol. II. Wiesbaden: Harrassowitz, pp. 1409-25. Schwarzwald (Rodrigue), Ora (2002), Studies in Hebrew Morphology. Tel Aviv. The Open University of Israel.
12 Lexical gender in Hungarian: reflections of social structures in an electronic corpus Marcell Nagy and Viviana Patti
1 Gende r features in Hungarian In Hungarian, just as in all Finno-Ugric languages, the concept of grammatical gender is unknown. Hungarian nouns and their modifiers cannot be classified as masculine, feminine or neuter and do not change according to the natural gender. Although the 'meaning' of a word like ferfi 'man' is naturally masculine and nd"woman' is feminine, the selection of the definite article does not depend on this, and the form of an adjective like boldog 'happy' is the same with both words: a ferfi 'the man' - a no"'the woman'; boldog ferfi 'happy man' - boldog no"'happy woman'. Nor does Hungarian use different personal pronouns for male and female people: (la) (1 b)
6 csendben lepett be a raktdrba. S/he entered the store silently. 6k meg tudjdk, hoi lakik edesapdd. They still know where your father lives.
Since Hungarian grammar contains no inherent distinction of natural genders, all differences between lexemes indicating female and male people should be analysed on the level of lexical gender (Pete 2000). Lexical gender appears mostly in word groups denoting semantic categories like 'animals' or, in the case of human beings, 'professions', 'roles' and 'social status'. In Hungarian, the expression of lexical gender is realized through both lexical and morphological processes. Hungarian animal names show the phenomenon of lexical gender sporadically; only some of the most frequent domestic animals have their own male and female correspondents: bika 'bull' - tehen 'cow', kan 'male dog' - szuka 'bitch dog'. Therefore, from a linguistic - and social - perspective it is more profitable to analyse the lexical gender of words referring to human beings. Though basic words like ferfi 'man' - no 'woman', ur 'sir', 'man' - holgy 'madam', 'woman' undergo a lexical substitution process (like the most frequent slang expressions: srdc 'guy' - csaj 'chick', hapsi 'fellow' - spine 'chick, bird'), the majority of female profession names are formed by a morphological process.
HUNGARIAN NAGY AND PATTI
183
Our starting point is that Hungarian contains both marked and unmarked profession nouns. An unmarked profession noun (UPN) can indicate a person of any gender; a marked profession noun (MPN) can only refer to a female person (Table 12.1). Table 12.1 demonstrates that in Hungarian morphological and various lexical processes can 'feminize' a word. The production of a marked profession noun can be fulfilled by four methods: 1 The UPN is substituted by a more or less different MPN form: this is typical in the case of animal nouns, basic words indicating men and women, and names of those professions which were traditionally done by women: nover- dpolo 'nurse' - '(male) nurse', mosono- mosodds 'washerwoman' - 'laundry man'. 2 The adjective not 'female' produces the MPN: noi^c^ + [unmarked profession name] —» [markedprofession name]: noi pilota 'female pilot'. 3 The noun asszony 'woman, lady' produces the MPN: [unmarked profession name] + asszonynoun —> [marked profession name]: miniszterelnok asszony 'female prime minister'.1 4 In the morphological production of an MPN the UPN gets compounded with the word wo"woman': [unmarked profession name] + nonoun —» [marked profession name]: for example eladono'saleswoman'. In a few isolated and non-productive cases the lexeme no is interchangeable with the allomorph ne. szakdcsno-szakdcsne' (female) cook'.2 Although the two are related, the lexeme no is not to be identified with the morpheme ne. The latter is called traditionally 'married name creating suffix', and it serves to create the wife's name (or denomination) from the husband's name or profession: (2)
Karoos Elek -» KardosElekne 'Mr Elek Kardos' -> 'Mrs Elek Kardos'. Kardos —> Kardosne'Mr Kardos' —» 'Mrs Kardos'. erdokerulo-* erdokerulohe 'forest guard' —> 'the forest guard's wife'.
Table 12.1 Unmarked and marked profession nouns in Hungarian 1. unmarked profession nouns elado 'salesman' sportolo 'sportsman' kolto 'poet'
marked profession nouns (formed by a morphological process) eladono 'saleswoman' sportolono 'sportswoman' kb'Uono ' (female) poet*
2. unmarked profession nouns
marked profession nouns (formed by lexical processes)
pilota 'pilot' miniszterelnok 'prime minister' publidsta 'publicist'
not pilota'(female) pilot' miniszterelnok asszony '(female) prime minister' not pubtirista' (female) publicist'
184
GENDER, LANGUAGE AND NEW LITERACY
This suffix can be attached to any male proper name and to many profession names as well. The married women's names produced this way are official female names in Hungarian, and the professional names serve as an 'identification tag' for females, for example, gondnoknemeans 'the woman known as the warden's wife'. This method is no longer really productive in the creation of married women's names. Nevertheless, the relatively high frequency of lexicalized words compounded with ne allows the making of a distinction unknown in some other languages: kirdlyno 'queen (monarch)' * kirdlyne 'queen (the king's wife)'. The morphologically produced profession names are supposed to seem 'more natural' to the speakers; thus, if an unmarked profession noun has a marked correspondent produced by a morphological process, it tends to mean a male person (though a female interpretation cannot be excluded at all). In the case of a lexically produced MPN, the UPN can refer to a female and a male person as well. One of the difficulties in the examination of MPN-creating processes in Hungarian is the fact that the lexical selection of a certain form (e.g. tandrno '[female] teacher' instead of *tandr asszony or nm pilota '[female] pilot' instead of *pilota asszony) depends on non-linguistic facts (such as the speaker's age, education level and social status, the grade of elevation of the speech act or the dialect used: Lanstyak and Szabomihaly 2002). As regards the preference of certain forms over others, we can only indicate some general tendencies rather than fixed rules. The morphological and the first lexical MPN-creation methods are presumably more typical in the case of jobs traditionally done by women (as well). The second lexical MPN creation method appears in nouns of professions done only recently by women, whose social prestige is not particularly high; the phrases created by the third lexical MPN more often denote the highly prestigious professions. In many cases it is possible to form more than one MPN from a UPN. Although the situation is complicated, it is possible to observe a correlation between the possible forms: for example, if a noun admits the second lexical MPN form, the third method tends to be blocked. On the basis of these correspondences we can establish four characteristic MPN classes. However, the borders between these groups seem sometimes blurred; some nouns tend to be classified in more than one group. Whenever possible, we will note the relative occurrence frequency of some examples found with Google () in the entire corpus of Hungarian websites. 1 type tonar'teacher' This group consists of job nouns traditionally done by both men and women. The results of the second lexical method are often accepted but the third one rarely gives a grammatical result. This is one of the most frequent and natural MPN-forming methods in Hungarian (Table 12.2 and Table 12.3).
185
HUNGARIAN NAGY AND PATT1
Table 12.2 Occurrence of MPN-creating methods in the nouns of the type tandr 'teacher' UPN no MPN ?no* UPN MPN ?UPN asszony MPN
Table 12.3 Frequency of various female profession nouns at http://www.google.co.hu (%) 'female teacher'
'librarian'
'cleaning woman'
'saleswoman'
'female writer*
tanarno 99.94 noi tanar 0.06 tanar asszony 0
konyvtarosno 99.06 noi konyvtaros 0.94 konyvtaros asszony 0
takaritono 99.75 noi takarito 0.16 takarito asszony 0.09
eladono 95.37 noi elado 4.53 elado asszony 0.01
irono 99.54 noi iro 0.44 fro asszonv 0.02
Other members of this group are kolto'poet', kalauz 'conductor', sportolo 'sportsman', riporter 'reporter', biro 'judge'. In some cases the MPN created with not is not acceptable because of its alternative meaning: noifodrdsz does not mean '*female hairdresser' but 'ladies hairdresser'. 2 type miniszterelnok 'prime minister' In most cases highly prestigious and relatively new profession denominations are found in this group. The feminine forms can always be produced with the word asszony, but the compounding with no is not alwavs accepted; even if the form exists, it tends to be only a 'describing noun' and cannot be vocative (Table 12.4 and Table 12.5). 3 type pilota 'pilot' This group contains words denoting professions done mosdy by men. In this group the use of the UPN to indicate a female referent is typical; however, it is always possible to create an MPN with the adjective nm. The compound with no is often accepted as well, whereas the type asszony is
Table 12.4 Occurrence of MPN-creating methods in the nouns of the type miniszterelnok 'prime minister' ?UPN no ?noi UPN
UPN asszony
MPN
MPN MPN
186
GENDER, LANGUAGE AND NEW LITERACY
Table 12.5 Frequency of various female profession nouns at http://www.google.co.hu (%) 'female head of department (at university)'
'female prime minister'
tanszekvezetono 0.38 not tanszekvezeto 0 tanszekvezeto asszony 99.62
miniszterelnokno dllamtitkdrno 14.78 5.1 not miniszterelnok not dllamtitkdr 16.04 0 miniszterelnok dllamtitkdr asszony 69. 18 asszony 94.9
'female secretary of state'
'female chief executive'
'female president'
vezerigazgatono 28.57 nm vezerigazgato 8.16 vezerigazgato asszony 73.27
elnoknd 23.99 not elnok 0.97 elnok asszony 75.04
practically unused. In Table 12.5 we do not show the relative occurrence because the most frequendy used form is obviously the UPN, and though an MPN formation with nm is always possible, sometimes it is hard to find an example because of the fact that these professions are not frequendy done by women (Table 12.6). Words of this group are pilota 'pilot', bordsz 'wine-grower', esztergdlyos 'turner', villanyszerelo 'electrician', traktoros 'tractor-driver', publicista 'publicist', menedzser'manager', haldsz'fisherman'. 4 type ovono'kindergarten teacher' In this group there are found professions traditionally done by women; these words often do not have a male correspondent. Once a female profession has been compounded with no, it is not possible to create a male profession name with backformation; the only possibility is substitution by another expression. Examples for these nouns are wosono'washmg woman' - mosodds 'laundry man', naver - dpolo'(female) nurse' - 'male nurse', stewardess - legiutaskisero 'stewardess' - 'steward'. Some words like fejono'girl who milks the cow' do not have any male correspondent at all. 2 Women' s socio-economic roles and gender equality policies in Hungary Hungarian women enjoy a legacy of high education, high labour force participation and both die expectation and the means to combine working and
Table 12.6 Occurrence of MPN-creating methods in the nouns of the type pilota 'pilot' UPN no- MPN *nm UPN MPN UPN asszony MPN
HUNGARIAN NAGY AND PATT1
187
raising a family: indeed, the 'worker mother' model was intrinsic to the labour policy of the so-called Communist regime before 1990. In recent decades structural socio-economic changes have impacted on conditions of work and quality of life. After 1990, Hungary experienced a recession due to the introduction of capitalism in the economy, and employment declined for both women and men, although since 1995 a slight recovery has been registered. The new market economy has done little for the progress of Hungarian women; a recent report (Pollert & Fodor 2005) indicates the following negative trends: female labour force participation is decreasing, women have dropped out of the labour market to work in the household or in informal economies, and equal pay for work of equal value is still a dream. However, two key findings of this report concerning enduring positive legacies should also be mentioned. First, the tradition of a strong female labour market attachment persists from the past and statistics show that Hungarian women have found their way into more professions and more management positions than women in Western Europe. Second, Hungarian women continue to reach high educational standards: in 2001 58 per cent of graduate employees in Hungary were women. As regards the 'worker mother' model, before 1990 it was supported by relatively generous maternity and family policies (Haney 2002). After the end of the Communist regime, the system of benefits was overhauled. The alternating left-wing and right-wing governments promoted family policies looking to different family models. The role of women encouraged by the right-wing governments seems to be the traditional one of women as caretakers of children, staying at home and raising children according to the single-earner family model in which men are not required to compensate for their wives' difficulty in performing household duties when they are in full-time employment. On the contrary, the family policy proposals of the socialist governments seem to be sensitive to gender equality and more open to family models alternative to the traditional one. In fact, the aim of state action is not to encourage one predefined family model, but instead 'the promotion of free choice for everybody in defining the family model' (Horvath et al. 2004). Full-time motherhood is not seen as the most desirable female role, sharing tasks within the family is encouraged and 'keeping the labour market position of mothers is seen to be in the interest of the woman, the child and also of society' (Horvath et al. 2004). Sharing work and the care of children within the family has been supported independently of the colour of the government. Moreover, promoting sharing policies is the target of a set of documents prepared during the 'Joining the EU' process, which provided an impetus for steps forward in the Hungarian debate on gender equality. In this context actions were mainly devoted to benefiting 'combiner-women': promoting reconciliation of work and family life remains in focus, the clear message being that women should move away from traditional roles in the family and that increasing the number of active, working women is in the interest of society, of
188
GENDER, LANGUAGE AND NEW LITERACY
children, and is at the end of the day in line with equal opportunities. However, this does not seem to go along with the assumption that traditional male roles should also change. (Horvath et aL 2004)
Finally, the placement of the institutions responsible for monitoring and implementing gender equality mechanisms within the state structure reflects the different perspective on this issue held by different governments: in 1996-98, under the socialist-liberal government, the Secretariat was under the Labour Ministry; in 1998-2002, during the conservative government, it was under the Ministry of Social and Family Affairs; under the new left-wing government, it was under the Labour Ministry until May 2003, when equal opportunities were raised to ministerial status. Currently, the Minister for Equal Opportunity has merged again into the Ministry for Youth, Family, Social and Health Affairs. In summary, with the change of political regime in 1990, Hungary started a slow but steady process towards the development of gender equality. Hungarian gender equality policies are weak in many respects (Krizsan and Zentai 2004). However, the statistics concerning the high education and labour force participation of women testify to the endurance of the positive legacy of the past, and good quality academic research in gender studies - e.g. a Hungarian team is active in the European project on gender mainstreaming strategy (MAGEEQ) - makes expectations for improvement in gender equality in Hungary high. 3 The MS Word Thesaurus in Hungarian: technical limits and flaws Our analysis was carried out by examining the Thesaurus tool in MSfi Word 2000, Office 2000 and Windows 2000.3 There are some general linguistic limits and flaws that emerged during the inspection of the Hungarian version of the tool. The MS Word Thesaurus is in many respects an unreliable lexical tool. In spite of the fact that it is supposed to be a 'dictionary of synonyms', it often gives false synonymous terms which in the best case are connected by hypernvmy and hyponymy with the entry, but often do not fit easily into any semantic affinity such as hypernymy, hyponymy or synonymy. Furthermore, it is not uncommon to find the lemma itself proposed as the first synonym of the list, which is at best a useless repetition. In order to highlight these limits, in the following we will sometimes mark the terms offered by the Thesaurus for a given entry by using the symbol ^ for hyponyms and the symbol ^ for hvpernyms: e.g given the entry ember 'human being' we will write ^feifi 'man' and ^ eloleny 'living being'. A significant linguistic shortcoming concerns the treatment of the Hungarian female profession names. As we saw in §1, the processes for creating female profession nouns form a multilateral and complicated system. In a labvrinth of such complexity any user of Microsoft* Word is en tided to expect
HUNGARIAN NAGY AND PATTI
189
some help from the Thesaurus or from the spelling function of the program. However, the lexically formed marked profession nouns are not filtered out by Word (even if they are not actually in use: "tandr asszony 'teacher woman' would seem correct) and die morphologically produced forms are not controlled, either. Word accepts the compound of almost every noun + no and, surprisingly, the majority of compound adjectives -I- no. Thus, not only does the program accept rarely used forms like "haldszno 'fisher woman' or ' publidstano 'publicist woman', but it also accepts absolutely meaningless forms like *borsno'pepper woman', *palacsintand"pancake woman', *buszno 'bus woman' or *sdrgano 'female yellow'. Moreover, some dictionary editing flaws are reflected in the Thesaurus, for example, the expression meg van oriilve utdna 'to be crazy about her/ him' is offered as a synonym of szerelmes 'in love'. Besides the syntactic problems of diis substitution, the phrase is also incorrect: the correct postposition should be erte, and not utdna. 4 Th e MS Word Thesaurus in Hungarian: a gender analysis The shortcomings described above highlight the weakness of a poorly accomplished endeavour to simplify the system. Furthermore, an analysis of die meanings and synonyms of several pairs relevant for the linguistic categorization of women and men reveals the 'socio-cultural' unreliability of the Thesaurus, in the sense dial die socio-cultural picture conveyed by die Thesaurus is traditional in a way which does not correspond to Hungarian reality. The gender analysis of significant pairs clearlv shows a qualitative imbalance in favour of the male gender. Male entries are often more wide-ranging with regard to semantic fields, bodi in meanings and in synonyms, and die semantic fields do not correspond. Also die register of information (formal informal; humorous; obsolete - modern; domestic; tender - neutral pejorative) provided for feminine and male lexemes, the related stereotypes, the order in which the information is listed and the use of metaphorical reference terms vary. All diese aspects have been considered by analysing a corpus of about 300 lexemes. 4.1 The crucial pairs
Let us start with an analysis of the, crucial pair no 'woman' -ferfi 'man'.4 (3a)
(3b)
no: n o (no, fehernep, feherszemely, fehercseled, ddma, asszony, noszemely, amazon, tyuk, spine, baba, nocske, noci, nosteny); lan y (Idny, no); harpia (hdrpia, furia, hdzisdrkdny, noszemely, nember, perszona, no); holgy (holgy, holgyemeny, holgy ike, no); asszony (asszony, no, asszony sag, urno). ferfi: ferf i (ferfi, ember); ferf i (ferfi, pasas, pasi, pali, alak, miokus, muki, ipse, hapsi, krapek, mdndro, manusz, muksi, mukso, pacdk, pok, azivar, tag, urge,figura, pofa); hi m (him, bak, kandur, kan, ferfi).
190
GENDER, LANGUAGE AND NEW LITERACY
Though with regard to the quantity of meanings and synonyms, in the couple no-ferfi the asymmetry is on the side of woman, when we analyse the main semantic fields a qualitative imbalance towards the male is evident. Universality is exclusively stressed in the male entry: Member 'human being' is a synonym only of the male lexeme ferfi and has a very visible position, being the only non-redundant synonym of ferfi offered to the user. Instead the woman is preferably reduced to subordinate roles or to traditional stereotypes, and synonyms of no often belong to an obsolete and pejorative register. In the most visible positions of synonyms for no we find highly obsolete and pejorative terms like fehernep, feherszemely 'white people'6 or ^-fehercseled 'maidservant, wench' - which are practically never used nowadays and relegate women to domestic chores - preceding the more neutral asszony 'Mrs (married woman), lady' - which mark marital status. Equivalent terms relegating men to marital status or to subordinate professional roles do not appear in the ferfi entry. Most of the remaining synonyms in the first list (tyuk 'chicken', spine 'chick, bird', baba 'baby', nocske, nod) are slang and often pejorative expressions typical of men's language, mainly stressing sexual appeal and frivolous attitudes. In this respect we do find a correspondence in the male entry, considering the synonyms of ferfi, belonging to the second group, where we find on the one hand the slang expressions pasas 'guy', past 'fellow/potential boyfriend', hapsi 'fellow', typical of women's language and belonging chiefly to a humorous or familiar register, and on the other hand terms like figura, ^ pali, urge, pofa, belonging to an informal pejorative register. Among the other meanings of no we find ^ Idny 'girl' (which unfortunately in this context is not provided with any significant list of synonyms) and then ^hdrpia 'harpy', followed by holgy 'lady' and asszony. Surprisingly, the meaning hdrpia precedes holgy and asszony, and many terms like ^•amazon 'amazon', ^nosteny 'doe' - provided as synonyms of no- or furia 'fury', hdzisdrkdny 'dragon of the house', noszemely 'female', nember, perszona 'turbulent woman'- provided as synonyms of hdrpia and all belonging to a humorous/pejorative register - connotate a negative stereotype of women as an unpleasant animal-like creature, showing irrational, brutish and turbulent behaviour. It is true that man is also characterized as an animal-like creature, but with the difference that in this case the metaphorical terms are not used chiefly to suggest an ugly or nasty nature, but rather to put the accent on the male as sexual partner. Among the meanings of ferfi we find -^ htm 'male', which in this context has a sexual connotation, as confirmed by a list of synonyms that includes metaphorical terms like bak 'he-goat', kandur 'tomcat', kan 'male-hog', all slang expressions supporting the stereotype of a macho man. No equivalent term referring to woman as active partner is present under no. A similar asymmetry results by analysing the pair Idny 'girl/daughter' - fiu 'boy', if we consider active and passive roles in courting:
191 Idny: lany (Idny, csaj, pipi, spinko, bige); lany (Idny, no); lany (Idny, ledny, kisldny, lednyzo, lednygyermek); bakfi s (bakfis, serdiiloldny, kamasz Idny, Idny); szu z (szuz, Idny); hajado n (hajadon, Idny). (4b) fiu: fiatalembe r (fiatalember, ifiu, fiatalur, fiu); udvarl o (udvarlo, gavaller, lovag, szeptevo, fiu); kisfi u (kisfiu, fiucska, legenyke, fiu); kamas z (kamasz, srdc, siheder, legeny,fiu, ijju).
(4a)
In fact, in the first group of Idny synonyms we find pipi, an expression from male language denoting a fashionable girl that is the object of courtship, while among the meaning of fiu (in a more visible position) we find udvarlo 'suitor', stressing the active role of the boy in courtship. Both pipi and udvarlo, like gavaller 'gentleman' and lovag 'knight' (provided under udvarlo), belong to a very obsolete register: they evoke a world that does not exist anymore and sound humorous nowadays. In the Idny entry the sexual side is emphasized by the adjective szuz. 'virgin', provided as meaning, with no equivalent in the male entry.6 Young men and women are also characterized as single, unmarried people, by using terms from an obsolete register: the synonym hajadon 'unmarried woman' is provided under Idny, while the synonym legeny, which refers to the bachelor condition, is provided under kamasz 'adolescent', a meaning of fiu. But the kinship term lednygyermek 'daughter' is provided among the synonyms of Idny, while equivalent terms reducing a young unmarried man to the category of son are not provided for fiu. Regarding the family sphere, a substantially complete quantitative and qualitative symmetry can be found in the pair anya 'mother' - apa 'father': (5a)
(5b)
anya. anya (anya, edesanya, mama, muter, mami, anyu, anyuka, anyuci, anyus, anyuska, mamus, mamuska, csalddanya, sziiloanya, edes); szulo (sziilo, az oregek, edesanya, edesapa, apa, anya). apa: apa (apa, edesapa, papa, atya, tata, tdti, tatus, apu, apuka, apuci, papus, papuska, az oregem, fater, csalddapa); szul o (sziilo, az oregek, edesanya, edesapa, apa, anya).
For this pair the Thesaurus behaves correctly. In fact, this is symmetrically the area of tenderness and intimacy, as is shown by the dominance of terms of endearment and diminutives, like edesanya, mami, anyu, anyuka, anyuci, anyuska, mamuska - for anya -, or like edesapa, tdti, Status, apu, apuka, apuci, papuska, az oregem - for apa: they are all synonyms belonging to a domestic register that encodes slightly different ways of endearing the mother or the father. The formal terms for referring to mother and father (csalddanya and csalddapa, respectively) are at the end of the lists. Finally, both father and mother have szulo 'parent' as second meaning. One negative symmetry that needs to be mentioned is the absence of any entry related to the pair nonemu 'feminine' - himnemu ('masculine').
192
GENDER, LANGUAGE AND NEW LITERACY
4.2 Social roles: the pair asszony/ ur and marriage-related terms
A strong asymmetry is made evident by analysing the pair asszony 'Mrs/mistress, lady' - ur 'Mr/mister, gentleman, proprietor', which can both be used as terms of address: (6a)
(6b)
asszony: (asszony, no, asszonysdg, urno); feleseg (feleseg, nej, hitves, hitvestdrs. hdzastdrs, asszony, oldalborda); n o (no, feheirnep, feherszemely, fehercseled, vdszoncseled, ddma, asszony, noszemely, amazon, tyuk, spine, baba, nocske, nod, nosteny). ur. u r (ur, urasdg, zsarnok); ur (ur, uriember); birtoko s (birtokos, gazda, ur); uriste n (uristen, ur, atya, mennyei atya, teremto, orokkevalo, mindenhato, magassdgbeli, isten); tulajdono s (tulajdonos, ur); hatalmassa g (hatalmassdg, kisisten, ur).
Although it is true that wrin Hungarian is used both as a term of address and as 'gentleman' and 'proprietor/monarch', the terms offered under wrtend almost exclusively to emphasize the power of man; the very least we can say is that the use of wras 'proprietor' is over-represented. In fact, no term relating ur to marital status is provided, and the sole term covering the meaning of 'gendeman1 is uriember. Instead, many meanings and synonyms of ur refer to the semantic field of possession, power and dominance, e.g. the most visible synonyms for i/rare urasdg 'proprietor' and zsarnok 'despot', and among the meanings we have birtokos 'landowner', tulajdonos 'owner', hatalmassdg 'mightiness/ possession of power'. On the contrary, asszony is related to the subordinate role of feleseg 'wife' and to the generic gender term no, while the correspondent y«7' 'husband' andferfi are not provided for ur. The ranking of meanings within the asszony main entry also stresses dependency on man: thus feleseg 'wife' precedes no 'woman'. No term related to dominance is provided among the meaning of asszony. One explanation could be that asszow) corresponds to ur when the latter is used as a term of address or means 'gendeman', while when ur means 'proprietor' the female equivalent is urno 'mistress': (7)
urna asszony (asszony, no, asszonysdg, urno).
However, analysing the meanings and synonyms given under urno reveals that die semantic field of dominance and power, which should allow us to distinguish the use of urno and asszony, is not evidenced at all; indeed, the meaning of urno is reduced to asszony. Women's dependence on men is also underlined in the feleseg' wife' entry, which has among its meanings oldaborda 'rib' in allusion to Eve, made out of Adam's rib. (8)
fefaeg. felese oldalborda) .
g
(feleseg,
nej, hitves, hitvestdrs,
hdzastdrs, asszony,
HUNGARIAN tiAGY AND PATTI
193
An analysis of marriage terms reveals a quantitative asymmetry on the side of woman: both ferj 'husband' and volegeny 'bridegroom' are missing, while the correspondent feleseg'wife' and menyasszony 'bride' are present.' In contrast, unmarried man and woman are symmetrically represented by the presence both of the neutral pair notlen 'unmarried man' - hajadon 'unmarried woman' and their pejorative versions volegeny 'old maid' - menyasszony 'confirmed bachelor'. (9a) (9b) (10)
hajadon: hajado n (hajadon, fiatal Idny, szuz); hajadon (hajadon, Idny). notlen: legenyember (legenyember, notlen, aggkgeny). venkisasszony. venlan y (venldny, oregldny, oggszwz, venkisasszony); leg enyember (legenyember, notlen, agglegeny).
Even if the entries above are not very rich, a sexual reference to virginity is present under hajadon, where szuz 'virgin' is provided as synonym. Age considerations do not make any difference in the case of female entries (volegeny is related to aggszuz - literally 'old virgin'), while the male entries notlen and menyasszony are free from sexual references. This suggests that virginity outside marriage is a female prerogative. 5 Conclusio n The Word Thesaurus is in many respects an unreliable text-editing tool for the Hungarian language. Some serious shortcomings concern general linguistic flaws. At the semantic level, the terms provided as synonyms are at best connected by hypernymy and hyponymy with the entry, and there are no contextual or usage indications - such as register - either for meanings or for synonyms. This weakness is made worse by the fact that most of the lexemes of the corpus belong to an antiquated vocabulary, even as regards slang expressions: when they are provided,, they mostly belong to a 1950s-1960s, or in the best case 1970s vocabulary, while up-to-date commonly used slang expressions, like the ones included in the slang dictionary of Kovecses (1997), are absent. This often makes it hard for the user to choose the appropriate word without making mistakes or coming up with outmoded terms that sound ridiculous, e.g. among the synonyms provided for Idny 'girl' we find terms like pipi, which is no longer part of current usage, unless it is used ironically; or terms like spinko which, besides being pejorative, is typically used informally; or, finally, terms like lednygyermek, which belong to a very formal register. Even the presence of metaphorical terms, often evoking obsolete cultural models now little used except ironically, may cause confusion, since this characteristic is not indicated, e.g. under feleseg'wife' we have the synonym oldalborda 'rib', in allusion to Eve, made out of Adam's rib, but contexts in which felesegcan be appropriately replaced by this expression are very rare. Moreover, the lack of clarity regarding morphological categories can lead a careless user to generate grammatically incorrect phrases. For instance,
194
GENDER, LANGUAGE AND NEW LITERACY
legenyember, which is a noun, is listed as a synonym of notlen 'unmarried man', which is instead an adjective: thus the substitution legenyember -> notlen produces incorrect phrases. Similarly, the antiquated adjective szeptevo (synonym of fiu 'boy' under udvarlo 'suitor') can substitute nouns only in sporadic cases; an automatic substitution would thus be risky. While lexical unreliability seems to be a cross-language problem of the Ward Thesaurus that is confirmed with respect to Hungarian, a significant linguistic shortcoming specific to Hungarian concerns the treatment of female profession nouns. The program accepts the compound of almost every noun + no, the majority of compound adjectives + no, and names like Wtandrasszony 'teacher woman', which are not used but sound correct. Finally, our analysis of the meanings and synonyms of several pairs relevant for the linguistic categorization of women and men reveals the 'socioculturaT unreliability of the Thesaurus. The socio-cultural picture conveyed by this tool does not reflect the recent social changes; rather, it reflects a way of speaking about Hungarian women and men characteristic of the 1950s. What this leads us to suspect is that the source of most of the synonyms is a 'dictionary of synonyms' (Magyar szinonimaszotdr, see O. Nagy & Ruzsiczky 1978), whose author - O. Nagy Gabor - lived from 1915 to 1973! This obviously makes the Word Thesaurus a tool that is useless for linguistically skilful users and mislead for others. Moreover, a qualitative imbalance toward the male gender also resulted from our analysis: male entries are often more wide-ranging with respect to the semantic fields, both in meanings and in synonyms, and in many cases the semantic fields do not correspond. In general both the universality and power of man are emphasized, while command-related features are never ascribed to women; this is particularly evident for terms concerning social roles. Woman is preferably associated with her subordinate roles or stereotyped as aggressive by nature: here we need only recall that hdrpia 'harpy' is provided as a meaning of no. The only exception to the prevalent qualitative asymmetry is represented by the pair anya 'mother' - apa 'father', where terms of endearment and diminutives are symmetrically dominant. Acknowledgements We would like to thank Andras Horvath for the inspiring discussions we shared on the subject of this chapter. Notes 1 In children's language this kind of process is extremly frequent and asszony is substituted by neni (an affective word to indicate [older] women), for ex. tandr Tieni'(female) teacher', eladoneni 'saleswoman'. 2 The second element of the compound can be changed into Idny 'girl', or in informal speech into csaj (a slang word to denote girls). 3 See Microsoftfiword2000; Thesaurus© 1994-2000.
HUNGARIAN NAGY AND PATTI
195
4 In the following examples, the meanings are in bold, whereas the items given for each meaning ('synonyms') stand in brackets, in italics. 5 Women in Hungary were called fehernep or feherszemely 'white people' because they dressed in white linen. However, today nobody wears white folk costumes and these terms have become pejorative and obsolete. 6 The same asymmetry occurs in the pair notlen 'unmarried man' - hajadon 'unmarried woman', one of the meanings of hajadon (but not of notlen) being szuz. 1 It is surprising that ^hitvestdrs and -^hdzastdrs 'consort', which are provided as synonyms of felesegbut denote both wife and husband, are not used to create an entry for ferj.
References Haney, Lynne (2002), Inventing the Needy: Gender and the Politics of Welfare in Hungary. Berkeley, CA: University of California Press. Horvath, Anna; Krizsan, Andrea & Zentai, Violetta (2004), Policies Frames and Implementation Problems: The Case of Gender Mainsteaming (MAGEEQ). Hungarian Report on Family Policy. Vienna: IWM Institute for Human Science. Kovecses, Zoltan (1997), Szleng-szotdr. Budapest: Akademiai Kiado. Krizsan, Andrea & Zentai, Violetta (2004), 'Gender Policies in Hungary on the Road to an Enlarged Europe'. Paper prepared for the 34th Annual Conference and 9th Research Conference The European Union: New Neighbours, New Challenges (UACES 2004). Birmingham (draft version), http://www.mageeq.net/docs/magpap03.pdf Lanstyak, Istvan & Szabomihaly, Gizella (2002), 'A magyar nyelv szlovakiai valtozatainakjellemzoi' ('Characteristics of Hungarian language spoken in Slovakia'), in Magyar nyelvtervezes Szlovdkidban. Tanulmdnyok es dokumentumok. Pozsony (Bratislava): Kalligram Kiado, pp. 84-108. Nagy, O. Gabor & Ruzsiczky, Eva (1978), Magyar szinonimaszotdr ('Synonymous Dictionary of Hungarian Language'). Budapest: Akademiai Kiado. Pete, Istvan (2000), 'Ferfinyelv-e a magyar?' ('Is Hungarian a male language?'). Magyar Nyelvor, 124, 108-15. Pollert, Anna & Fodor, Eva (2005), Working Conditions and Gender in an Enlarged Europe, European Foundation for the Improvement of Living and Working Conditions. Luxembourg: Office for the Official Publications of the European Communities.
13 A world of difference: gender disparity in the Turkish Word Thesaurus Marina Castagneto and Rosita D'Amora
1 Gende r in Turkish Turkish is one of the very few languages in the world totally devoid of morphological gender. More specifically, Turkish does not require a noun to be identified as masculine or feminine, and of course there is no syntactic agreement for gender inside a syntagmatic structure. As for articles, the problem is avoided by the complete lack of a definite article which possibly had to conform to the noun.1 Nor do we find a gender distinction in the pronoun class, as often happens in other languages where gender is now a weak linguistic category (see Bettoni): in Turkish the third-person pronoun o2 does duty for 'he'/'she' ' 'it', so that any anaphoric agreement for gender in the discourse is excluded as well.3 As a consequence, in a sort of 'naturalness scale' it is obvious that the verb does not show agreement in gender either. 2 Lexica l gender
Yet, in every language - and Turkish is no exception - on cognitive grounds, lexical gender does exist and nouns like anne 'mother', baba 'father', ktz 'girl' and so on, because of the 'sex of the referent' (Hellinger & Bufimann 2001), are semantically specified as carrying among their specifications [+ male] or [+ female] features. So, how does Turkish deal with semantic gender distinctions? 2.1 Lexical suppletion
The Turkish language system is inclined to use totally different words for [+ male] and [+ female] terms in the same pair, even when the referent is not a human being: see aygir 'stallion' vs kisrak 'mare' or boga 'bull' vs inek 'cow' (on this problem, see Lewis 2000: 23). As the commercial value of a domestic animals varies depending on its sex, a bull or a cow are actually conceived as totally different referents, as are the terms for 'mother' and 'father', for different reasons, in human relations.
TURKISH CASTAGNETO AND D'AMORA
197
2.2 Suffixes In some rare cases there is 'morpheme induction'4 from loanwords for morphemes denoting feminine. So, from three Serb-Croat loanwords, ktralife 'queen' (< kraljica), imparatorife 'empress' (< imperatorica), farife 'tsarina' (< carica) the suffix -i& of feminine gender has been introduced. This suffix has proved to be reusable and productive to a certain degree: see for example the noun tanrtfa 'goddess', a neologism based on tann 'god', a very old preTurkish word traceable back to the third century BCE, if not earlier, to the language of Hsiung-nu, with the meaning 'Heaven' intended as a deity. As it is easy to see, in tannfa the morpheme -i$e (> -ifa) has undergone a phonological process of integration to the rule of vowel harmony.5 2.3 Loanwords Even though Turkish has no morphological gender marking, the languages from which it has borrowed many loanwords do have it. This is particularly the case with Arabic, from which Turkish imported loanwords up until the end of the nineteenth century, and with French, especially in the last days of the Ottoman Empire. So in Turkish there are quite a lot of noun pairs marked [+ female] or [+ male] directly derived from Arabic with a slight phonological integration, particularly in vocalism, as in mudiir 'director' mudire 'directress', muattim ' male teacher' - muallime 'woman teacher', and several others, but even if these loanwords are easy to segment, the feminine morpheme -e has never undergone induction, and no feminine Turkish neologism has been built through it. A powerful process of phonological integration has taken place by which French language phonemes have been replaced with the nearest Turkish system phonemes.6 This is the case for noun pairs such as Cantor 'male singer' (< chanteur) - santoz 'female singer' (< chanteuse), prens 'prince' (< prince) prenses 'princess' (< princesse), masor 'masseur' - masoz 'masseuse' and the more generic term mosyo'mister' (< monsieur) - madam 'madam', 'lady', matmazel'mademoiselle', 'miss' (< mademoiselle). As often happens in the case of borrowings, the nouns mosyo, madam and matmazel underwent a semantic narrowing, which resulted in their functions being specified as titles or address terms for non-Muslim people, respectively 'men'/'married women'/'unmarried women'. Even a cursory glance at how the Thesaurus treats these terms reveals all its limits: it seems not to be sensitive to the structure of the semantic fields or to gender symmetries, providing as the only meaning for mosyo the proper Turkish tide of respect beyefendi 'sir', and for matmazel die more generic term kadin 'woman' (madam has been provided with two meanings: hanim 'lady' and kadin' woman').
198
GENDER, LANGUAGE AND NEW LITERACY
2.4 Addition of a whole explicit word
But in most cases, as we said, Turkish nouns are neutral as regards gender, and can equally well denote [+ female] or [+ male] referents: it is scarcely possible to know, for instance, on hearing the word kedi 'cat', if the referent is a she-cat or a he-cat, and there is certainly no guarantee that the unmarked term will refer to a male. In such cases Turkish applies the strategy of the addition of a whole explicit word. In the quoted instance a she-cat is referred to as disi kedi 'female cat' while a he-cat is an erkek kedi 'male cat'. 3 Th e first crucial semantic field Does this strategy apply to human beings as well? Of course it does. But here the word erkek 'male' is not in direct opposition with difi 'female', as it is for animals; rather it creates two crucial pairs with two different words: ktz 'girl, virgin' and kadin 'woman', as in ktz ognenri 'girl student' or in kadm asa 'woman cook'. The choice between ktz and kadin depends, in our opinion, on the age of the referent but also on the related parameter of her presumed sex-appeal. That is, the word kiz, denoting 'girl', also has a quite significant sexual connotation (clearly shown by its second meaning 'virgin', or by colloquial expressions like kiz gibi, lit. 'like a girl', whose meaning is 'new, untouched, beautiful'), while the word kadm, far from having a bad or lowgrade sexual connotation, is entirely free from this. A Turkish woman is considered a kadin not only when she grows old but also when she loses her virginity or gets married.7 In addition, a woman socially acquires the title of kadin as soon as she becomes a mother (see Saracgil 2001: 24-5). On the other hand, a male is always referred to by the word erkek, whether we are talking about a male animal (e.g. erkek kedi 'male cat'), a boy (erkek focuk 'male child'), a man (plainly erkek, or used to specify in professions that a referent is a male, e.g. erkek oyuncu 'actor'), or even an object, which can be termed as 'male' because of its shape (e.g. erkek anahtar 'solid stemmed key' or erkek kopfa 'a hook for an eye') and because of its hardness, as is the case for minerals (e.g. erkek bakir 'hard copper' or erkek demir 'hard iron').8 The word erkek carries a strong sexual connotation, as is suggested by its secondary meanings 'manly, courageous, virile' (furthermore: 'honest and true')9 and by the related word erkeklik 'masculinity7', 'manliness', 'virility', 'courage', 'male sex', 'sexual potency'.10 So men's sexual identity is always recognized, even in general discourse (as in expressions in which the term erkek is parallel to the term kadm like kadmh erkekli toplanti 'mixed company', 'a company with men and women'11), while to women this is denied. We will now proceed to examine the treatment of these words in the Thesaurus (see Table 13.1 in the on-line Appendix, available at www.lingue.unibo.it/wordthesaurus). Obviously there is no trace of this intricate semantic field structure, as the Thesaurus is not a dictionary. But here the first meaning of erkek is just herif'A guy', 'a fellow1 (whose first provided synonym is adam 'human being, male' -
TURKISH CASTAGNETO AND D'AMORA
199
see below - and the second one is, surprisingly, kavalye 'lady's escort', a semantically narrowed loanword from French): this is a clear mistake, because not a single attested sentence in Turkish can be found to justify such a meaning for erkek. The second meaning to be provided is yigit 'young man', 'youngster', 'hero', with nine synonyms (out of 11) that show similar meanings related to courage, bravery, boldness and so on (a gem: among the related words provided for erkek, we find erkek itself). What about the 'female' side? As for difi, 'female' referring to objects and animals, the first unbelievable meaning provided is erkek 'male'! Nor do things improve with the remaining meanings: 11 synonyms correspond to the second meaning, gevsek 'loose', 'slack', 'lax', 'flabby', 'flaccid', 'lacking in backbone' (see this entry in Redhouse 2000), most are similar (see kagfak 'decrepit'), except for some which denote 'softness' (e.g. yumufaak 'very soft' or lop 'big, soft and round', a dimension often stereotypically connected to the feminine in general). The following third and last meaning for difi is kadm (where the deprecatory connotation comes by being provided as a synonym of difi), and the choice of the synonyms of kadm is quite humorous in its stylistic and semantic heterogeneity: matmazel 'mademoiselle', used to address non-Muslim unmarried girls; dudu, a colloquial term denoting an old Armenian woman; madam, another loanword from French for nonMuslim married women; hamm, a tide of respect used to address a woman; amazon 'Amazon'; bakire 'virgin', 'spinster'; kiz 'girl'; kocakan, an impolite slang term to denote 'old woman', not to mention kokana 'overdressed and excessively made-up woman', a term formerly denoting all kinds of non-Muslim women. As for kiz 'girl', the only meanings provided by the Thesaurus are (ocuk 'child' (which opens the semantic field of 'smallness' to the synonyms yavru 'pet', bebek 'baby', 'doll', and so on), the loanword term dam 'lady partner in dance'12 (but its synonyms are related to its homonym 'roof), and the syntagm iskambil kagidt 'playing card'. The last meaning provided is again kadm 'woman', where the bulk of its heterogeneous synonyms, compared to the ones provided as a meaning of the headword difi, appears in a chance order: some words are missing, some others are new entries,13 while the very negatively connoted word kokana 'overdressed and excessively made-up woman' is always present. Turning to the word kadm 'woman', the most frequent and crucial word in this whole semantic field, it is a true surprise to discover that the Turkish Thesaurus offers just one single meaning, and a very unexpected one: hanimefendi, that is literally 'lady gentleman'. A title of stronger respect than the non-compounded word hamm 'lady', this last one may be slightly more appropriate to serve as a meaning for the word kadm. The synonyms sloppily provided for the meaning hanimefendi are nearly the same group of words which are given as synonyms of kadm 'woman' when it was considered a meaning of other headwords (difi, kiz), never excluding the offensive kokana, the 'overdressed and excessively made-up woman', in the lexicographer's opinion an unavoidable word when speaking about women. But our
200
GENDER, LANGUAGE AND NEW LITERACY
lexicographer did (her?)/his worst in the choice of related words, among which there are aybast 'menstruation', gebe 'pregnant', cilveti 'coquettish', the very same provided as related words of difi, 'female', too! Moreover, the first related word is kadi 'judge of Islamic canon law', probably as a result of its phonic similarity with kadtn, another proof of the Thesaurus's carelessness; the second related word is e$ 'one of a pair', 'one of the married couple', just to underline (as if there were any need to) that this is the obvious role for a woman. Unfortunately, the negative semantic connotation of women is rooted in Turkish society, and is not an artifact of the Word Thesaurus, or something specifically limited to the word kadin 'woman'. If we look at proverbs, we immediately find, as happens in other cultures too, the expected stereotypes for women: unreliability (e.g. in hadinla denize guven olmaz, lit. 'it's not possible to rely on sea and woman',14 or kadtn mevsimler gibi, lit 'woman is like the seasons', two proverbs with a non-opposition structure based on 'identificational features'),10 and dangerousness (e.g. kadinm fendi erkegi yendi, lit. 'woman's cunning has defeated man', a classic proverb with quadripartite structure, each half having a thematic and a rhematic part).16 As for die terms erkeklik 'masculinity', 'manliness' and kadmhk 'womanhood', 'ladyship', two derived abstract nouns from the roots discussed above, the Thesaurus provides them with just a single meaning each, but with a striking difference: according to the Thesaurus the only possible meaning for kadtnlik 'womanhood'-is 'virginity', while erkeklik 'masculinity', 'manliness' evokes cesaret 'courage', with 12 synonyms and three related words belonging to the same semantic field. Let us check the word adam, 'human being, man'. This very well-known loanword from Hebrew refers, as happens in Hebrew itself (see also Livnat, this volume), to the human race as a whole, but sometimes its referent is perceived as a male in sentences like arabada bir adamla iki (ocuk vardt 'in the car there were two children with a man' (the example is in Puskulluoglu 1995). In most cases, however, adam is a very generic term designating an individual natural person for both men and women, in contrast to words such as insan 'human being', 'mankind', felt, in our opinion, as having a more abstract and philosophical referent. See the frame by the Thesaurus (Table 13.2, on-line Appendix). For our lexicographer this term mainly refers to 'man', having as its first meaning erkek 'male'. The second meaning is es 'one of a pair', 'mate', 'one of the married couple (husband/wife)', probably because, on a colloquial and familiar linguistic register, adam can also have the meaning 'husband' as used by a wife when referring to her husband:17 in this case too the Thesaurus fails to distinguish between linguistic registers, and thus risks suggesting an unsuitable meaning or synonym. As for es, the second meaning offered for adam, this term means 'one of the married couple', so it can be used to indicate both wife and husband. But under the headword adam the husband-like synonyms always precede the wife-like ones: thus *r'male, brave man' and koca, the most frequent word to refer to 'husband' (second and third synonym respectively) precede kart
TURKISH CASTAGNETO AND D'AMORA
201
'wife', the word in pair with koca (fifth synonym), while the sixth synonym, zeuf, is again the old loanword from Arabic denoting 'husband' (but the corresponding feminine loanword, zevfe, is missing). The word heldl 'lawful wife' comes even further down this list. The complete frame of es as headword in the Thesaurus is in Table 13.3 (on-line Appendix). 4 The Thesaurus and the linguistic reform: Bay 'Mr* and Bayan 'Miss'/'Mrs' The words Bay 'Mr' and Bayan 'Miss'/'Mrs' are very important indicators of lexical accuracy or inaccuracy in the Thesaurus. These two loan translations of 'Mr' and 'Miss'/ 'Mrs', to be put before the name18 (following the English model), were introduced19 in modern Turkish by the language reform of 26/11/1934 with the law No. 5290 intended to abolish tides like efendi, bey, pasa and to move towards an approximately similar status for men and women. As a consequence, all Turkish dictionaries are very symmetrical in definitions, meanings and synonyms provided for the Bay and Bayan headwords. What about the Thesaurus?20 (la)
(Ib)
Bay 'Mr.': baskan 'president, chairman' (bajvekil 'premier', melike 'queen', hukumdar 'ruler, monarch', far 'tsar', prens 'prince', padisah 'padishah, sultan', sah 'shah', hakan 'khan', firavun 'pharaoh', han 'khan', reis 'head, chief, president'); beyefendi 'sir' (mosyo'monsieur, foreign gentleman', sinyor 'mister', felebi 'gentleman'); erkek 'male' (oglan 'boy, youth', herif 'fellow, guy', er 'male, brave man', kavalye 'lady's escort', muzekker 'masculine (gramm.)'); zengin 'rich' (yapkin 'wealthy', mureffeh 'comfortable, luxurious', esraf 'notables of a town or village', kapitalist 'capitalist', variyetli 'well-to-do, wealthy1. sosyete 'the upper classes', parah 'having money', milyoner 'millionaire', parasi bol 'her/his money is abundant', banker 'banker'). Related words: . . . . Bayan 'Miss'/'Mrs': kadin 'woman' (ktz 'girl', dudu 'old Armenian woman', madam 'madam, lady', bolide 'Jewish woman', avrat '(prov.) woman, wife', hanim 'lady, woman', bakire 'virgin, maiden', dam 'lady partner in a dance', matmazel 'mademoiselle, miss', kokana 'overdressed and excessively made-up woman'); unvan 'title of rank or dignity' (far'tsar', carife'tsarina', begum 'lady', kayser 'emperor', fasar 'emperor of the Holy Roman Empire', beybabam, beybaba 'father (used also for other respected elderly person)', fagfur 'made of Chinese porcelain'). Related words:....
Look at examples (la) and (Ib) (for a full survey comprehending also the Related Words, see Table 13.4, on-line Appendix): even with these two terms which were specifically introduced in order to be symmetric, the Thesaurus shows no symmetry at all. The word Bayan 'Miss'/'Mrs' has just two
202
GENDER, LANGUAGE AND NEW LITERACY
meanings, the first one being kadin 'woman', while Bay 'Mr' has four. The word erkek 'male', with an explicit reference to the sex of the referent, is only Bay's third meaning, the first one being the word bafkan 'president, chairman' (with all its panoply of synonyms related to men of power: bafvekil 'premier', hiikumdar 'ruler', far'tsar', prens'prince\firavun 'pharaoh' and so on, with the only female exception of melike 'queen'), and the second one being the word beyefendi 'sir', an address term of respect for men (providing as unlikely synonyms some loanwords: mosyo 'mister', 'monsieur', 'foreign gendeman', from French, and sinyor 'mister', from Italian). The choice of related words is grotesque: Bay is provided with synonyms that are mosdy power titles (among the others: diktatdr 'dictator', komutan 'commander', yonetici 'manager', politikaa 'politician') while Bayan has among its related words kinship terms (as teyze 'maternal aunt' or ktzkardef 'sister') or terms such as aybafi 'menstruation', gebe 'pregnant', cilvett 'coquettish woman'. 5 Socia l and economic role s The fact that there is no gender distinction in Turkish does not mean that any given name can refer to both women and men. Reflecting the social roles women and men had, or still have, within society, many nouns are used exclusively either with a feminine or masculine value. Words such as sultan 'sultan',21 yeniferi 'Janissary', zampara 'womanizer' do not obviously have a feminine form. Similarly, other words such as eu hanimi 'housewife', sutanne 'wet nurse',22 orospu 'prostitute', do not have masculine equivalents. Therefore, even though names denoting occupations are grammatically ungendered, this does not prevent some of them referring to an occupation traditionally performed by a certain sex; as such they are perceived more as masculine than feminine or vice versa.23 To denote occupations Ottoman Turkish had always used a certain number of loanwords, and whenever in the language from which Turkish was borrowing there were two distinctive forms for the masculine and the feminine and a similar distinction needed to be made in Turkish, both these forms were adopted.24 As a result of the Turkish Language Reform - initiated in the middle of the nineteenth century and then strongly enforced by Mustafa Kemal Ataturk beginning in the late 1920s - Turkish acquired a huge number of neologisms that enlarged the wide lexical patrimony inherited from Ottoman Turkish. In order to replace foreign words, perceived as extraneous to the Turkish language, a large number of new words were coined using the possibilities offered by Turkish for forming new words through suffixation. In this way many words of non-Turkish origin were, and often are still, used together with or instead of neologisms or loans from other languages. In modern Turkish, a medical doctor, doktor, is also called hekim (Arabic loanword) or tabip (Arabic loanword),just as a cerrah 'surgeon' (Arabic loanword) can also be called operator (French loanword). Words such as hakim 'judge', menajer 'manager' and artist 'artist1 have their Turkish equivalents in the neologisms
TURKISH CASTAGNETO AND D'AMORA
203
yargLq (< yargi 'decision' < yar- 'to split'), yonetici or yonetmen (< yonet- 'to manage' < yon 'direction'), and sanatgi or sanatkdr (< sanat 'art'). The majority of words denoting occupations are obtained in modern Turkish by adding the suffix -ci or the suffix -men. The suffix -«, also used in Ottoman Turkish, is added to the singular of nouns and occasionally to adjectives and adverbs: is, 'work', i$$i 'workman', eski 'old', eskici 'old-clothes dealer'.25 The suffix -men, used to form agent-nouns denoting occupations, was, instead, 'invented' by the language reformers in the mid-1930s and it was used mostly to form administrative and military terms.26 Terms coined in this way were, for example, ogretmen (< ogret- 'to teach') 'teacher', seamen (< sef- 'to choose') 'elector', yeoman (< yaz- 'to write') 'secretary'. It is interesting to notice how, even though the suffix -men occurred in a number of timehonoured words, mostly 'with intensive significance' (see Lewis 2000: 21920) (e.g. koca 'big', kocaman 'huge'), this suffix also recalls the English -man (as in postman) and the German -mann (as in Wahlmann 'elector') in compounds.27 It is clear that while borrowing words from other languages such as Arabic or French, which often have two separate forms for feminine and masculine, involved an 'ideological choice' regarding the form that was to be adopted, the Turkish system of forming nouns to denote occupations contains, in principle, no gender orientation. The word oyuncu (< oyun 'game', 'play', 'dance') can refer to an actor and to an actress and dansci to both a female and a male dancer. The same could be said for all the other terms formed through recourse to the suffix -ci Analogously, the suffix -men, despite its supposed origin, does not have any masculine connotation in Turkish. In case of ambiguity it is necessary to specify whether the person in question is a man or a woman and usually the word erkek 'male' or kadm 'woman' is put before the name: erkek oyuncu 'actor' and kadin oyuncu 'actress' (see also §2.4 above). These ungendered suffixes do not conceal, though, the social and cultural stereotypes regarding the roles occupied by men and women within society and Word Turkish Thesaurus reflects many of these fossilized stereotypes. 6 Socia l gender and crucial pairs in Word's Thesaurus The first crucial pair analysed here is yonetici 'manager, administrator' vs hizmetfi' se rvan t': (2a)
yonetici 'manager, administrator': $ef 'chief, 'leader' (yonetici 'manager, administrator'); mudur 'director' (amir 'superior, commander', idareci 'manager, administrator', isletmed 'manager, administrator', vasi 'guardian, executor', reis 'leader, chief, patron 'boss, employer', menajer 'manager', vezir 'vizier', mudire 'directress'). Related words: yonetmek 'to manage', garson 'waiter', bakan 'minister', baskan 'president', diktator 'dictator', komutan 'commander', onder 'leader, chief, vali 'governor of a province', patron 'boss, employer', hukumet
204
(2b)
GENDER, LANGUAGE AND NEW LITERACY
'government', politikaci 'politician', yonetim 'management', kahya 'steward, majordomo', menajer* manager'. hizmetfi 'servant': gundelikci 'day labourer' (temizlikfi 'cleaner', yamok 'male assistant', azap 'male servant', ispir 'stableman', beslemelik 'servant girl brought up as a member of a household', geysa 'geisha', ayakfi 'servant who runs errands', seyis 'stableman', usak 'manservant', kalfa 'journeyman', 'ranking female servant'). Related words: hizmet 'service', garson 'waiter', hole 'male slave', yardima 'helper', haremagasi 'black eunuch in the sultan's palace'.
Both nouns are formed with the suffix -a and they should therefore have an ungendered reference. However, the subordinate role that is usually given to a servant in the hierarchical organization of society is echoed in the Thesaurus with the inclusion of the female sphere in the references given for this term, while the word yonetici seems to refer almost exclusively to men. The meanings proposed for yonetici are sef 'chief, leader' and mudiir 'director', two loanwords both originally masculine. While as a synonym of sefihe term yonetici reappears, the corresponding feminine form of mudur, mudire 'directress' is relegated to the end of the list of synonyms for this word. Significantly, this latter group of synonyms includes many other loanwords from Arabic originally marked in this language as masculine forms. Furthermore, historically, some of these words have been used almost exclusively with a masculine connotation, as for example vezir 'vizier'.28 Among the related words it is worth noticing the lexicographer's choice of the word garson 'waiter' placed even before words such as bakan 'minister' and baskan 'president'. Quite different is the approach to the word hizmetfi. According to Puskulluoglu (1995), this word denotes a servant but more specifically a female servant, as opposed to the archaic hizmetkdr, or more common usak 'manservant'. The onlv meaning proposed by the Thesaurus is gundelikfi, literally 'a labourer who works on the basis of a daily wage (gundelik)'. Puskulluoglu (1995) gives for the headword gundelikfi the definition 'day labourer' too, but he relates this word mostly to women, quoting, as an example of its use, only the compound gundelikfi kadin 'cleaning woman paid by the day'. Among the synonyms here it is possible to find, along with words denoting different kinds of male servants, words referring exclusively to women such as beslemelik 'servant girl brought up as a member of a household1, geysa 'geisha' and kalfa. This last word, which in modern Turkish translates as 'experienced apprentice workman', has been included, probably with the meaning it had in Ottoman Turkish, where it designated a ranking female servant of the Imperial Palace or of a konak (mansion).29 Among the related words, inexplicably included, it is worth mentioning the choice of haremagasi 'black eunuch in the sultan's palace', a peculiar choice considering the countless options offered by the rich ranking terminology of the Ottoman Court. In Turkish society not only professional, but also economic success has
TURKISH CASTAGNETO AND D'AMORA
205
traditionally been perceived as a purely male prerogative. A woman is considered unable to produce wealth and her possessions can only come from her family or from her husband.30 Significandy, the Thesaurus lists, almost regularly, feyiz 'trousseau' among the words denoting the female sphere. Analysing a word such as milyoner 'millionaire' it is not surprising, then, to find not even a single reference to women (see Table 13.5, on-line Appendix). A sufficient degree of symmetry between male and female elements, if not actual prominence given to the female, emerges from the analysis of the words sanatfi 'artist' and dansfi 'dancer' (see Table 13.6, on-line Appendix). The meanings that the Thesaurus offers for sanatfi 'artist' are: oyuncu 'actor', 'actress', dadaist 'dadaist' and sarktci 'singer'. Beside the inexplicable preference given to dadaism, these terms, at least, can be applied indifferently to both women and men. Examining the synonyms it is worth noticing two of those proposed for dadaist. balerin 'ballerina' and balet 'male ballet dancer'. Although some effort has clearly been made to represent both sexes fairly, once again the lexicographer's choice seems to be inadequate. As regards oyuncu 'actor', 'actress', the first of the svnonyms offered is aktris 'actress', closely followed by ba$aktris 'leading actress', while the corresponding masculine form aktor is omitted. The only term that clearly refers to the male sphere is a rather specific one: jon, which translates as 'a young male actor playing leading roles'. Analogously, the only meaning proposed for damp 'dancer' is dansoz 'female dancer', while its masculine form dansaris not even listed among the synonyms.31 Furthermore, in this group it is possible to find three more words used exclusively to denote different kinds of female dancers (balerin 'ballerina', fengi 'girl dancing accompanied by a musical instrument' and rakkase 'belly dancer') and two masculine terms, namely balet 'male ballet dancer' and ko$ek, which is used to denote a youth who performed erotic dances in women's garb. The richness of synonyms offered bv the Thesaurus and the symmetry between male and female elements in this semantic field reflect the process of enrichment of Turkish, which, over time, acquired words denoting different artistic traditions in which the female sphere was better represented. Many of the words analysed here entered Turkish in the nineteenth century, when the Ottomans, willing to emulate European culture, introduced new literary genres and performing arts extraneous to their tradition. Along with them came an entire new vocabulary to accommodate the consequent need to define new roles. The word tiyatro 'theatre' (< Italian teatro) was followed at that time by other words such as aktris 'actress' or aktor 'actor'. The word zenne (literally 'woman'), used in the traditional Ottoman Orta oyurnc® to indicate the character of a 'lady' performed by a male actor in women's dress, must surely have been perceived, for obvious reasons, as completely inadequate to describe the role played bv a female actress in a theatrical play. Similarly balerin 'ballerina' and balet 'male ballet dancer' were two necessary loans to describe a western kind of dance that had no connection whatsoever with the Ottoman tradition.
206
GENDER, LANGUAGE AND NEW LITERACY
7 Famil y roles and crucial pairs in Word's Thesaurus A family-based society, such as the Muslim-Turkish one, where setting up a family is crucial for the entire community life, has naturally evolved a rich vocabulary to define the different positions and roles each member of a family has within it. In Turkish great importance is given to all the family ties and the terminology used for the in-laws is also very articulated. On the other hand, while Turkish has two different words to denote a sonin-law (damat) and a groom (guvey, also meaning 'son-in-law'), the word gelin which translates as 'bride' also means 'daughter-in-law' or 'wife of a younger member of the family'.38 If we look at how the Thesaurus deals with these last two words (Table 13.7, on-line Appendix), we find that the only meaning proposed for gelin 'bride' is es 'one of a pair, one of the married couple', better defined by the synonyms attached to it that also recall concepts concerning legal wedlock such as zeuce 'wife', helal 'lawful wife', nikdhh 'married', zeu$ and koca 'husband' or terms denoting relationships acquired by marriage: eloglu 'son-in-law', 'husband', damat 'son-in-law', 'groom'. While adam 'human being', 'man', is listed as a synonym of gelin, the word kadin 'woman' is 'banished' among the related words. Furthermore, it is only in this last group that it is possible to find the only term relating to love: sevgiti 'dear'. The way the Thesaurus deals with gfwey 'groom' is completely different, where seugili 'dear' is presented, instead, as first meaning while e$ 'one of a pair', 'one of the married couple' comes as the second one. Even more interesting is that, as synonyms of seugili 'dear', alongside the words related to formalized relationships leading to marriage (namzet, yavuklu, sozlu all meaning 'fiance', 'fiancee') and those used in addressing a beloved person (dftktaf 'sweetheart', yar 'beloved' and gozbebegi, literally 'pupil' but also used as 'apple of the eye', 'darling'), there is a striking number of terms denoting extramarital affairs or illicit lovers: cariye 'female slave', 'concubine', odalik 'odalisque', oynaf 'illicit lover' and kapatma 'mistress', 'concubine'. Finally, it is also worth emphasizing the bizarre inclusion of the word poliandri 'polyandry' in both groups of related words. For the words anne 'mother' and baba 'father' (see Table 13.8, on-line Appendix) the Thesaurus, besides offering some synonyms such as ana and validefor mother and pederand beybaba for father, presents, curiously enough, a random list of terms denoting family ties mostly referring to grandparents and parents-in-law: (3a)
(3b)
anne 'mother': haminne 'grandma' (valide 'mother', kaynana 'motherin-law', kayinvalide 'mother-in-law', kocakan 'old woman, hag', ana 'mother', buyukanne 'grandmother'). Related words: kadin 'woman', gebe 'pregnant', anaerkil 'matriarchal', anneanne 'mother's mother', babaanne'father's mother', dadi 'nanny'. baba** 'father': ded e 'grandfather, male ancestor, old man' (babahk 'fatherhood, paternity', agababa 'grandfather', kaynata 'father-in-law', kayinbaba 'father-in-law', ihtiyar 'old person', sutbaba 'husband of a wet
TURKISH CASTAGNETO AND D'AMORA
207
nurse', peder 'father', beybaba 'father', buyukbaba 'grandfather'). Related words: bap 'gate', yash 'aged', erkek 'male', ataerki 'patriarchy', babaanne 'father's mother', kaym 'brother-in-law'. Worth noticing, though, is the powerful asymmetry between the word haminne 'grandma', the only meaning proposed for anne 'mother', and the word dede 'grandfather', the only meaning offered for baba 'father'. In the first case, haminne (< hanim nine, lit. 'lady grandmother') is a colloquial form of address used for an old woman, especially within a family context. Dede instead is a term of a much higher register and it is, in fact, also used to designate a male ancestor or a senior dervish. Along the same lines, among the synonyms of haminne we find kocakan, a derogatory term for an old woman, which could be translated as 'hag', while the group of words given for dede includes the adjectives ihtiyar 'old person' and yash 'aged' that do not carry, at least in principle, any negative connotation. Most important, however, is that none of the words proposed, except for babalik 'fatherhood, paternity' (listed among the synonyms of baba 'father'), seems to have any real connection to the roles a father and a mother have within the family. Finally, it is worth analysing how the Thesaurus presents the words bekdr 'single, umarried', kalik 'old maid', bakir 'virginal, untouched' and bakire 'virgin, maiden' (see Table 13.9, on-line Appendix). The word bekdr can refer to both women and men. It is generally used with the meaning of 'unmarried' but it can also refer to a married person living apart from his/her spouse and to a person who is separated, divorced or widowed.35 The adjective bakir, of Arabic origin, when used in relation to human beings, refers specifically to the sexual sphere with the meaning of 'virginal', 'untouched' and it can be used both for women and for men. The feminine form of this adjective has the same meaning but can only be used for women. Kalik, which literally means 'remaining', 'left over', from kal- 'to remain', 'to be left', as it can be easily assumed, is a derogatory term used for a woman regarded as being beyond the age of marriage, or, as it is said in Turkish, a woman who has been 'left at home', evde kalmts, Despite the negative connotation usually attached to this last word when referred to a woman, the Thesaurus seems to consider it an equivalent of bekdr 'single', 'umarried', bakir'virginal', 'untouched' and bakire'virgin', 'maiden' and provides the same sequence of meanings for all of them. The symmetry displayed in this case conceals the different way a virgin or unmarried man and a virgin or unmarried woman are traditionally perceived by the society. 8 Conclusio n The main problem with Word's Turkish Thesaurus is the way in which the lexical material is presented: the Thesaurus offers for each searched word one or more meanings and, in most cases, for each single meaning a list of synonyms is provided. These synonyms refer to both the searched word and the meaning offered for it. Furthermore, there is a list of related words that can
208
GENDER, LANGUAGE AND NEW LITERACY
refer either to the searched word or to any one of the meanings and synonyms listed previously. Searching for the word gelin 'bride', for example, we are offered as a meaning es 'one of a married couple' and, as synonyms, words strictly referring to bride, as well as words such as bir 'one', bironek 'identical' and yeknesak 'similar', which are, instead, only synonyms of es, which also means 'one of a pair, equal, like'. Furthermore, among the related words we find the words ayni 'the same' and etene 'placenta' that are again connected to es, which also means 'afterbirth, placenta'. From this point of view the Thesaurus is far from being an easy and reliable text-editing tool and most of the time its use presupposes a very high level of linguistic competence. In the Word Thesaurus linguistic registers are mixed up so that standard words appear side by side with archaisms (e.g. muattim 'teacher'), provincial terms (e.g. dados 'brother, young man', used in Eastern Anatolia), colloquial terms (e.g. dudu 'old Armenian woman') and slang terms (e.g. Inzdik 'little child'). Often the choice of the meanings is quite inexplicable. As a first and only meaning for ktz kardef 'sister' the Thesaurus proposes haber 'news, information' and as first meaning of disi 'female' there is erkek 'male'. Things are no better with the synonyms (e.g. dadaist 'dadaist', given as second meaning for the word sanatp 'artist', has as synonyms baletin 'ballerina' and balet 'male ballet dancer'). Furthermore, some words are listed onlv because of their partial homography (e.g. among the related words of kan 'wife' we find kdr 'profit, benefit' and kar'snow'). There is a systematic lack of symmetry between words denoting male reference and female reference. Even neologisms built up to be symmetric, as Bay 'Mr' and Bayan 'Miss'/'Mrs', are not dealt with as such by the Thesaurus. to the word Bay the Thesaurus devotes twice as many meanings and synonyms as to the equivalent female term. Acknowledgements Sections 1-4 were written by Marina Castagneto; Sections 5-7 were written by Rosita D'Amora. Notes 1 The category of 'definiteness' is expressed in Turkish by demonstrative adjectives and by the accusative suffix of the definite object. 2 Third person plural pronoun: onlar, where -far is the generic suffix denoting the plural, while -n- before the plural suffix still has a controversial etymology. The circumstance that 'personal pronouns have no relation to gender either', with its consequences about gender, is underlined by Hayasi (1998). 3 As is well known, the complete absence of a gender distinction in Turkish has been profitably used in the Ottoman erotic-mystic poetry pervaded by a profound
TURKISH CASTAGNETO AND D'AMORA
4 5
6 7 8
9 10 11 12
13 14 15 16
17 18
19
20 21
209
gender ambiguity regarding the identity of the beloved. On this problem, see Bombaci (1969:79-80). On the notion of 'morpheme induction', see Gusmani (1986: 137-64 and 1987: 105). Another probable suffix -im in hamm 'lady' (originally: 'wife of a Khan') has been disputed as to whether it was an old feminine suffix or not. It probably is not, but is the possessive suffix of the first person. See, among the other proposals for an etymology, the one of Saadet Cagatay (1962), in which hamm derives from Persian khanuman 'household'. On all this question, see Lewis (2000: 23). Gusmani (1987: 98) calls this kind of phonological integration process 'sostituzione per approssimazione', 'replacement by approximation'. A woman considered old enough not to get married is called in Turkish kiz kuntsu 'dry girl'. Symmetrically difi anahtar, lit. 'female key' denotes a hollow key, difi bakir/demiris 'soft copper/iron' and, as a confirmation of the widespread stereotype, difi sea, lit. 'female word', means 'soft words'. So in Redhouse (1968), while Redhouse (2000) adds the meaning 'macho' too. How far we are from ktzlik 'girlhood', 'virginity', morphologically symmetric to the analysed word erktklikl See also the following expression: bin erkeklere ve digeri kadmlara mahsustur 'one is reserved for men, the other for women'. As it often happens to borrowings, this loanword from French dame 'lady' has undergone a process of semantic narrowing leading it to the very specialized meaning of 'lady partner in dance'. E.g. dudu 'old Armenian woman' is a synonym of kadm in the difi but not under the headword ktz, where we find instead bolife 'Jewish woman'. Here the aorist form gives to the proverb an everlasting, sapiential value, in the words of Greimas (1974). In Dundes's (1975) terminology. On Turkish proverb structures in general, see also Castagneto (2002). Another significant proverb is kadin erkegi isterse vezir isterse rezil eder 'the woman, if she wants, makes the man a vizier, if she wants, put him in disgrace', in which the standard Turkish syntax is endangered by the raising to focal and thematic position of the object of the verb, erkek. For this kind of proverb structure, see Milner i1969). But this is a very marginal signification: e.g. in Puskulluoglu (1995) 'husband* is die eleventh meaning provided for adam. Bay and bayan are thus not to be put after the name as is obvious in the Turkish language, and as happens in original Turkish titles (for instance, Bay Mehmed, but Mehmed Efendf). As a matter of fact, bayan is a reform neologism, but bay is an old Turkish term, which was already found in a paper manuscript in the Runic Alphabet (Irk Bitig, a book of divination dating from the ninth century) and that survives in all Turkish modern languages (see Clauson 1972: 384). But its historical meaning was 'rich', while the meaning 'mister' has been tailor-made (Heyd 1954: 107 and Iz s.v. 'Bav', Islam Ansiklopesi, vol. 2, pp. 355-7). In the following examples, the meanings are in bold, whereas the items given for each meaning ('synonyms') stand in brackets, in italics. In the Ottoman Court, though, this title would refer to the sultan only when it was placed before his name (e.g Sultan Mehmet). This same title, after a female name,
210
GENDER, LANGUAGE AND NEW LITERACY
would refer to a woman of the ruling family (e.g. Hurrem Sultan). The sultan's mother had, in fact, the title of Vatide Sultan (valide 'mother'). See this entry in Angelico da Smirne (1965). 22 The masculine form sutbaba is used to refer to the husband of the sutanne. regarded as a sort of godfather. 23 Stimulating research about gender connotation in words referring to social and economic words in Italian is due to Calabro (2005). 24 The vast majority of these words were borrowed from Arabic (e.g. hakim 'judge') but also from Persian (e.g. hemsire 'sister', 'nurse'). In later times, as the Ottomans increasingly came into contact with European culture, many other loans passed into Ottoman mainly through French (e.g. artist 'artist'). As we already noted with some Arabic loans, in certain cases both the masculine and the feminine forms were acquired (e.g. aktor 'actor', aktris 'actress'). See also §2.3. above. 25 Following the rules of the consonant-alternations and the vowel harmony of the Turkish language this suffix has eight possible forms: -ci as in kahveci (kahve 'coffee') 'coffee-maker', -cii as in gozcu (got 'eye') 'oculist', ~ci as in kapici (kapi 'door') 'janitor', -cu as in sporcu (spar 'sport') 'sportsman', -fi as in diffi (dif 'tooth') 'dentist', -fii as in sutfii (sut 'milk') 'milkman', -ft as in sanatfi (sanat 'art') 'artist', (u as in topcu (top 'cannon', 'artillery piece') 'artilleryman'. This suffix can also be used to denote persons who are habitually concerned with, or devoted to, the object, person, or quality denoted by the basic word: Ataturk> Ataturkfu 'Ataturkist', yalan 'lie' > yalanci 'liar', evet efendim 'yes', 'sir' > evet efendimd 'yes-man', see Lewis (2000: 55-7). 26 This suffix, following the rules of the vowel harmony, can also have the form -man. 27 Lewis, for example, argues that the English -man in compounds must have been familiar to the Turks in four borrowings from French (vatman 'tram-driver', sportmen 'sportsman', barmen 'barman', rekortmen 'record-holder'). See Lewis (2000: 219-20) and Rossi (1964, vol. 2: 11). Heyd (1954: 91 and 106) noticed how this suffix at first met with strong criticism, but more recently a number of such words have become increasingly used. 28 Strangely enough, in Turkish, the queen in chess is called vezir. 29 The current meaning of this term comes as well from Ottoman Turkish where kalfa was also used to designate, in the hierarchy of an artisan's guild, an intermediate position between that of apprentice (ftrak) and that of the master (usta). 30 A Turkish proverb significantly says: Kadinin samdam alttn olsa mumu dikecek erkektir, which literally means 'even though the candlestick of a woman is made out of gold, it is the man who will put the candle on it' and it is used to mean that however plentiful and valuable is the trousseau of a woman, it is the man who will provide the livelihood. 31 In this case dansorznd dansoz are two real loanwords, while dansft derives from a segmentation of the two borrowings with the consequent morphological integration in Turkish through the suffix -ci. 32 The Orta oyunu (lit. 'entertainment staged in the middle place') is a form of Turkish popular entertainment so called because it takes place in an open air space around which the spectators form a circle with the actors playing in the centre. The word zenne is also used to denote the characters of the 'ladies' of various types and ages in another form of Turkish popular entertainment, the shadow theatre called Karagoz (lit. 'black-eyed') from the name of its main figure. 33 This implies the connection of the bride to a whole family rather than simply to
TURKISH CASTAGNETO AND D'AMORA
211
the groom and it is explained by the fact that, under the traditional wedding arrangements, a gelin did not choose her future husband and therefore the attention was on her role within the husband's family more than on her relationship with her husband. See Duben & Behar (1991:104-5). 34 The word baba appears in the Thesaurus also with its meaning of 'bollard', 'bit'. For the sake of clarity, we omitted in the example and in Table 13.8 this meaning and all the synonyms referring to it. 35 In the Ottoman Empire this word was also used to designate those unmarried men who, beginning in the sixteenth century, started moving from the provinces of the empire to Istanbul in search of employment. These young bachelors with no family ties were considered a potential danger for society and were, in fact, encouraged to live in special hostels termed bekdr odalan 'bachelors' rooms' to keep them away from the mahatte 'quarter', where only married households were permitted. See Saracgil (2001: 30 n. 63).
References Angelico da Smirne (1965), Nuovo Dizionario Turco-Italiano. Reggio Emilia: Libreria Editrice Trate Francesco'. Bombaci, Alessio (1969), La letteratura turca. Florence: Sansoni. Calabro, Denise (2005), 'La semantica del genere. Un nuovo approccio ai gender studies.' Unpublished MA thesis, Universita di Cagliari. Castagneto, Marina (2002), 'La alimentazione nella paremiologia turca', in Domenico Silvestri, Antonietta Marra & Immacolata Pinto (eds), Atti del Convegno Internationale 'Saperi e Sapori Mediterranei', Napoti 13-16 ottobre 1999. Quaderni di AIQN, NS, 3, III, 995-1021. Clauson, Gerard (1972), An Etymological Dictionary of Pre-Thirteenth Century Turkish. Oxford: Clarendon Press. Cagatay, Saadet (1962), 'Turkgede "Kadm" icin Kullamlan Sozler'. Turk Dili Arastirmalan Yilhgi Betteten, 130,13-49. Duben, Alan & Behar, Gem (1991), Istanbul Households: Marriage, Family and Fertility, 1880-1940. Cambridge: Cambridge University Press. Dundes, Alan (1975), 'On the structure of the proverb'. Proverbium, 25, 961-73. Greimas, Algirdsjulien (1974), 'Proverbi e detti', in DelSenso. Milan: Bompiani. Gusmani, Roberto (1986), Saggi sull'interferenza Unguistica. Florence: Le Lettere. Gusmani, Roberto (1987), 'Interlinguistica', in Romano Lazzeroni (ed.), Linguistica storica. Rome: La Nuova Italia Scientifica, pp. 87-114. Hayasi, Tooru (1998), 'Gender differences in modern Turkish discourse'. InternationalJournal of the Sociology of language, 129, 117-26. Hellinger, Marlis £ BuBmann, Hadumod (2001), 'Gender across languages: the linguistic representation of Women and Men', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages, Vol. I. Amsterdam and Philadelphia: Benjamins, pp. 1-25. Heyd, Uriel (1954), Language Reform in Modern Turkey. Jerusalem: The Israel Oriental Society. lz, Fahir (1940-), s.v. 'Bay', Islam Ansiklopesi, vol. II, 355-7. Lewis, Geoffrey (20002), Turkish Grammar. Oxford: Oxford University Press. Milner, George-Bertram (1969), 'Quadripartite structure'. Proverbium, 14,379-83. Puskulluoglu, Ali (1995), Turkey Sozluk. Istanbul: Yapi Kredi Yaymlari. Redhouse, Walter (1968), Yeni Turkfe-Ingilzce Sozluk. Istanbul: Redhouse Yaymevi.
212
GENDER, LANGUAGE AND NEW LITERACY
Redhouse, Walter (2000), The Larger Redhouse Portable Dictionary, Istanbul: Matbaacihk Egitim Ticaret A.§.
Rossi, Ettore (1964), Manuale diLingua Turca, Vol. 2. Rome: Istituto per 1'Oriente. Saracgil, Ayse (2001), R maschio camaleonte: Strutture patriarcali nell'Impero ottomano e netta Turchia moderna. Milan: Bruno Mondadori.
14 Gender in Chinese and new writing technologies Antonella Ceccagno
1 Introductio n This chapter undertakes a preliminary analysis of the linguistic representation of gender in standard Chinese as it emerges from the analysis of two different tools: a modern writing environment, consisting of the most widely used computer writing programme for writing in Chinese; and a corpusbased database containing person reference names from which information on gender can be inferred. The examination of standard (commercial) word processing software can provide significant information on gender representation considering that the visual aspect of Chinese characters (or logographs) and the shift from standard Chinese phonetic transcription (pinyiri) to characters in writing a text in Chinese entail a totally novel approach to word processing: in fact the software makes suggestions and choices at the word processing level which are not necessary for alphabetic scripts. This chapter will therefore highlight the role of widely used word processing input methods in contributing to perpetuate or even strengthen gender-biased uses of the (written) language. Moreover, person reference entries in the corpus-based database, combined with equivalent entries in the Chinese dictionary of the most widely used word-processor software, enable us to analyse person reference nouns in Chinese from a gender perspective. This article therefore analyses pronouns and person reference nouns with the aim of verifying how gender manifests itself in Chinese. It describes the main tendencies in the use of genuinely indefinite nouns, false generics and nouns with covert gender, that is expressions that have an inherent gender bias even though they are neutral structurally (on generics, see Hellinger Be BuBmann 2002, on covert gender, see Braun 2001). The main focus of the chapter will be on covert male gender. In fact, revealing the existence and the widespread use of nouns with covert gender proves to be of paramount importance in challenging the traditional assumption that Chinese, being a language almost without derivation, is a genuinely gender-neutral language. The issues discussed and the findings presented here therefore go beyond the original scope of this research. A study originally started at an applied level - with the aim of highlighting the categorization of gender in new
214
GENDER, LANGUAGE AND NEW LITERACY
electronic devices - also provides new data and new insights that cast a new light on gender theory in Chinese itself. 2 Visua l level of gender bias 2.1 The visual impact of Chinese characters
One distinguishing feature of written Chinese is the fact that gender issues arise in this language earlier than in most other languages, in that the visual aspect of the Chinese characters already provides room for an analysis of the gendered representation of social models. Most of the characters in Chinese are phonetic/semantic (DeFrancis 1984): these characters are composed of at least two elements each providing the reader with vague information on the sound or on the meaning of the character as a whole. While it is not possible to infer the exact meaning of a character from its features, the visual impact of Chinese characters may convey gender-biased messages. This issue is analysed by Ettner (2002), who found that out of 500 dictionary entries relating to compound logographs containing the nu 'woman/female' element, approximately 20 per cent were derogatory or had negative connotauons. A clear example of this phenomenon is the compound character /*' meaning 'be jealous , hate where the element on the left, nu 'woman/ female', is intended as the part contributing hints as to the meaning and the element on the right, «', provides hints as to the sound of the whole. However, the semantic element nu is also found in (a limited number of) characters that appear not to be female-specific but have instead a wider lexical and referential potential. Examples of these characters - which may well be remnants of an earlier matrilineal society - are xlng' surname', shV ancestor' and xu 'husband'. Unfortunately, it is not easy to contrast the Chinese logograph nu 'woman/female' with the logograph nan 'man/male', because the latter is in itself not a simple graph but is composed of tidn 'field' and II 'force'. 2.2 Gendered characters and the ward processor Due to the properties of written Chinese, the process of writing using word processors can in itself be analysed from a gender perspective. Software makers have tried to meet the requirements of Chinese language users by developing a series of writing tools. One of these is known as IME (Input Method), which allows a standard keyboard to be used for typing in non-alphabetic writing systems. For the purpose of this chapter, I have used and analysed Microsoft Word 2003 with Microsoft Pinyin IME 2003. In order to produce one of the several thousand characters, often homophonous, of the Chinese language using a standard western keyboard (with about 100 keys) the Pinyin IME requires the following steps: • the user types standard pinyin phonetic transcription of words/phrases;
CHINESE CECCAGNO
215
• as the pmyin is typed, the IME offers a range of possible homophonous characters which could correspond to the intended input; the IME also tries to anticipate what the following characters will be. The first option is presented by the program as the default, while the following ones need to be explicitly selected by the user; • the user selects the option that best suits her/him and then moves on. In order to provide options among different characters, compounds and phrases, the IME must have a large database of lexical and phrasal entries. Unfortunately, since IMEs are commercial products, no precise information is available as to how these databases have been built up or how they are structured. It is, however, possible to analyse how these tools actually work, focusing our attention on both the lexicon contained in the database and the criteria according to which the database suggests a compound or phrase, for example, which gender of personal pronoun it chooses to combine with a given verb of action or with a verb of quality. The analysis of the rationale underlying the functioning of the Chinese IME as far as gender is concerned appears of fundamental importance given the fact that it is this tool - and its implicit gender-biased choices - that users writing in Chinese are regularly confronted with. 2.3 The visual aspect of Chinese pronouns
Chinese is a left-branching language with modifiers preceding syntactic heads. Typologically speaking, Chinese is known as a typically 'genderless' language. To be more precise, Chinese is a classifier language, where grammatical gender is reduced. While some issues central to the description of gender in other languages such as nominal classes and their gender-variable satellites, agreement and gender-resolution are not relevant to Chinese, other issues, such as morphological/lexical processes of gender specification and the so-called covert male generics and false generics are quite relevant also for this language. Contemporary Chinese displays a gender distinction in third person pronoun forms: td/tamen 'she-her'/'they-them' are pronouns for women; td/tdmen 'he-him'/'thev-them' are for men; finally, td/tamen it / they-them for things. In the spoken language, however, the difference is not perceived, due to the fact that the pronunciation of the three pairs of pronouns td/tdmen is exactly the same. Gender bias, as far as pronouns are concerned, emerges when the plural is formed. While female personal pronouns can be used only to refer to women, male pronouns have a referential potential which is wider: tdmen 'they/them' generally refers to men, but may also include women. The examples in (la) and (Ib) are correct while the examples in (Ic) and (Id) are incorrect:
(la)
216
GENDER, LANGUAGE AND NEW LITERACY
Tdmen (jiejie he didi) hut Idi le Theym p| (older sister and younger brother) come back CRS1 They (older sisters and younger brothers) came back.
(lb) Wang laoshijido tdmen (nan tongxue he nu tongxue) Wang professor teach themm pl (male student and female student) Professor Wang teaches them (male students and female students).
(10 Tdmen (jiejie he didi) hut Idi le TheVf.pi. (older sister and younger brother) come back CRS They (older sisters and vouneer brothers) came back.
(Id) Wang Idoshijido tdmen (nan tongxue he nu tongxue) Wang professor teach them,p, (male student and female student) Professor Wang teaches them (male students and female students). Even though this is probably due to the fact that, historically, the written pronoun $J td was devoid of gender distinctions, now, one century after the adoption of a gender distinction for third person pronouns, a hierarchy which considers the masculine as the privileged gender is evident. 2.4 Pronouns and gender categorization in Chinese word processors
As we have seen, the structure of the Chinese written language makes it possible to apply a gender-centred analysis also to the way in which word processing software has been programmed, as far as gender is concerned. Consider the following example. The pronoun td in its spoken form and in its frinyin transcription form means both 'she/her' and 'he/him' (and also the neuter 'it'). When shifting from the phonetic transcription to characters, the IME of the word processor suggests one default gendered logographic pronoun that the user can accept or refuse. Randomly typing some sentences in Chinese we can detect a pattern followed by the word processor: in most of the sentences including a third person pronoun and a verb written in ptnytn the suggested pronoun form is the male one. In (2) a random selection of ten such sentences is shown:
(2)
td bdijdng td kdi che td jidng ke tdhdn tdkdn td zuojdn td xiwdng td shuijido tdhim td hud chudn
'he pays a visit' 'he drives' 'he teaches' 'he shouts' 'he watches' 'he prepares food' 'he hopes' 'he sleeps' 'he breathes' 'he rows a boat'
CHINESE CECCAGNO
217
By contrast, with other verbs the word processor suggests the third person female pronoun, as illustrated in (3): (3)
td chdngge td tido wu tdxido tdzoulu tdxiyifu td chdo cdi tdku td titie td wenqing td sdo di
'she sings' 'she dances' 'she smiles' 'she walks' 'she washes clothes' 'she deep-fries vegetables' 'she weeps' 'she looks after with great care' 'she is gentle and kind-hearted' 'she sweeps the floor'
These examples clearly show that the functioning of this database is based on the identification of humankind with male gender. The male pronoun is considered the default choice and therefore actions or attitudes which are considered common for all human beings are generally proposed with the male pronoun td. The same rationale of treating the male pronoun as unmarked is to be found in the previous version of Microsoft Input Method, IME 3.0, though here the system actually works in a slightly different way. In fact, the word processor invariably offers the female pronoun as the first option for the user to select. If the user does not select it, the word processor immediately changes to the male pronoun, which is evidently intended as unmarked. Here too, therefore, the male pronoun is considered the default. The semantics of the characters following the pronoun is irrelevant for IME 3.0. But Microsoft Pinyin IME 2003 goes a step further and proposes actions or attitudes that are traditionally believed to be typically feminine, such as 'to take care of people', 'to sweep the floor', etc., with the female pronoun ta and therefore implicitly as gender-specific. We can therefore conclude that both Microsoft IME 3.0 and Microsoft Pinyin IME 2003 introduce a degree of gender bias; and Microsoft Pinyin IME 2003 seems to strengthen it. On the basis of some of the examples quoted above with the female pronoun and the male one, an anthropologist might be induced to think that in a Chinese environment while a man teaches, a woman smiles, while a man hopes, a woman weeps. This pattern is also followed for apparently contradictory suggestions as in the examples in (4) where both female and male pronouns are proposed with regard to 'food preparing'-actions: (4a) (4b)
td zuo fan td chdo cdi td shdofdn
'he prepares the food' 'she deep-fries vegetables' 'she cooks rice'
What actually happens is that the verb
zuo 'to do' in
to zuo fan 'he
218
GENDER, LANGUAGE AND NEW LITERACY
prepares the food' in (4a) is in itself generic enough ('to prepare') to be proposed with the male pronoun (evidently the software is not designed to also recognize the following part of the phrase/sentence). By contrast, the verb chao 'to deep-fry', as well as shoo 'to cook', specifically refer to cooking food, and are recognized by the software as typically feminine activities and therefore the female pronoun is suggested as the default. The reverse is also true. Usually the choice of the female pronoun character is evident before the sentence is completely written in pinyin transcription. For example, the female pronoun is proposed in the following incomplete sentences:
(5)
to chao... to wen... to ca...
'she cooks ...' 'she tender/submissive/meek 'she rubs ...'
Presumably the automatic choice of the female pronoun is what explains curious cases such as the following (6): (6)
td chao cat 'she deep-fries vegetables' td chao dl pi lit. 'she deep-fries land', i.e. 'she speculates illegally in real estate'
The female pronoun is proposed both with 'she deep-fries vegetables', an action traditionally regarded as female specific, and with 'she speculates illegally in real estate', an action diat is not specifically associated with women. These are side effects of the automatic attribution of actions or qualities to the gender that usually is not regarded as unmarked! 3 Lexica l gender in Chinese, word processor input methods and corpusbased data-base 3.1 Gender and person reference nouns
As a general rule, Chinese does not possess a system of nominal classification and nouns do not require gender to be made specific, therefore Chinese can be perceived as a comparatively gender-neutral language. For instance 'terrace' and 'garden' in (7) show no gender: (7)
ydngtdi 'terrace'
huayudn 'garden'
lidngci Besides, articles do not exist in Chinese. Numeral classifiers 'measure words' in Chinese) are used instead; classifiers have no role in determining gender.
CHINESE CECCAGNO
219
Explicit gender marking can be achieved through the combination of X nu gender lexemes with other terms for person reference: the nouns 'woman/female' or nan 'man/male' acting as modifiers may precede the noun, in an appositive position, as shown in (8): (8)
pengyou friend 'friend'
nu pengyou woman/female-friend 'female friend'
nan pengyou man/male-friend 'male friend'
ren person 'person'
nuren woman/female-person 'woman'
nan ren man/male-person 'man'
ren 'person' as a genuinely gender indefinite noun will be The status of discussed below. Other words containing the broad meaning of 'male' or mil 'mother , po 'elderly woman', etc., which are referred to 'female' females; gong 'elderly man', 'father', fu 'father', etc., which are referred to males) can be used in final position to distinguish gender as in (9a), and are used in compounds indicating kinship relationships as in (9b):
(9a) shipo 'witch1
shigong 'wizard'
(9b) bofu bomu father's older brother - mother father's older brother - father 'wife of father's older brother', 'aunt' 'father's older brother', 'uncle' It is also possible to signal sex for animals in the same way:
(10) muji mother - chicken 'hen'
gongji male - chicken 'rooster'
Since gender is usually lexically constructed, in Chinese we do not find those derivational suffixes expressing gender such as English -ess and -ette as in 'sculptress', 'usherette' -which are often considered as 'trivialisingwomen's occupations and undermining their professional status' (Holmes 2001: 117). 3.2 Gender-indefinite nouns At first glance, most terms for person reference seem to be genuinely genderindefinite. We can single out a group of Chinese words for which three forms
220
GENDER, LANGUAGE AND NEW LITERACY
exist, a gender-neutral one, one referring to women and one referring to men .In (11), a list of such words is given: (lla)
\isheng doctor'
nu ylsheng 'female/woman doctor'
nan yisheng 'male/man doctor'
(lib)
nufuwuyudn female/woman-servicemember Janitor/attendant' 'female/woman janitor/ attendant'
'male/man janitor/ attendant'
bingyudn army member 'soldier'
nu (ring female/ woman anny 'female/woman soldier'
nan tnng male/man army 'male/man soldier'
keren guest person 'guest'
nuke
female/woman guest 'female/woman guest'
ndnke male/man guest 'male/man guest'
fuwityudn service-member
ndnfuwuyudn male/man-service-member
(He)
(lid)
ylsheng 'doctor' in (1 la) is a grammatically neutral term. Both 'female/ woman doctor' and 'male/man doctor' exist in the Chinese language and are used to specify the gender of a doctor. The same is true for the subsequent examples in (11). Both fuuniyudn 'janitor/ attendant' and bingyudn 'soldier' are compounds based on yuan 'member'; yuan 'member* can either remain there when gender is specified explicitly as in (1 Ib) or be eliminated as in (lie). 1 he same happens for ren, 'person', in (lid). On the basis of the examples in (11) we could conclude that, from the point of view of the subject of this article, Chinese could be a model of biasfree human nouns, where men are riot the norm and where female and male specifications are made explicit onlv when necessary. If this were indeed the norm for gender formation in Chinese, then nouns such as jiaoshou, 'professor', would be perceived as non-gender-biased, and, on the model of the aforementioned examples, we might be induced to expect that an additional female or male specification would be possible so that we could nu jiaoshou 'woman/female professor' and nan jiaoshou have 'man/male professor'. However, this assumption would be wrong. nan jiaoshou is not ungrammatical; only it is perceived as 'unnatural' in conditions where no emphasis on gender would be required. The same is not nu jiaoshou 'woman/female professor'. true for
CHINESE CECCAGNO
221
3.3 Asymmetries in human references shown by the word processor
The fact that in Chinese human nouns are not bias-free as far as gender is concerned is evident if, again, we turn to Microsoft Pinyin IME 2003 and to the choices it recommends. For a group of professions I have tried to write the three forms: the gender-free one, the female one and the male one. The nu 'woman/female' word processor has shown that while the option noun of profession is available in an automatic way (that is, if written in pinyin transcription the word processor provides the characters), the male equivanan 'man/male' + noun of profession often is not available, and the lent user needs to construct it by separately choosing the characters for male and for the desired profession. In most cases when trying to write nan + noun of profession, what the word processor automatically proposes is the profession nan 'man/male', viz. nan, preceded by another nan, a homophone of which means 'south'. In (12) a list of the professions for which Microsoft Pinyin IME 2003 has provided such choices is given: (12)
jiaoshou 'professor' nu jiaoshou 'female/woman professor' nan jiaoshou 'south-for-male professor; ddngshizhdng 'director' nu dongshhhdng 'female/woman director' nan ddngshizhdng ' south-for-male director': niwagong 'bricklaver' nu niwagong 'female/woman bricklayer' nan niwagong'south-for-male bricklayer'; lushi 'lawyer' nu lushi 'female/woman lawyer' nan lushi 'south-for-male lawyer'; xiaozhdng 'school president' nu xiaozhdng 'female/woman school president' nan xiaozhdng 'south-for-rnale school president'; zhuxi 'president' nu zAiixf'female/woman president' nan zhuxi''south-for-male president'; xiejiang shoe-maker nil xiejiang 'female/woman shoe-maker' nan xiejiang 'south-for-male shoe-maker'.
These examples show that jiaoshou 'professor' - as well as all the nouns listed in (12) - is not a genuine generic noun including in its meaning both female and male professors but rather a covert male generic, which means •man/male professor' and which, on this basis, is able to acquire a wider lexical and referential potential to include even female referents. On the other hand, nu jiaoshou is devoid of such potential and only refers to 'woman/female professor' (we will discuss male generics and female nouns with gender-indefinite reference later).
222
GENDER, LANGUAGE AND NEW LITERACY
3.4 Female and male occupational activities in the Hawnet database
In order to test these preliminary findings, I extracted a list of occupational nouns containing the character for woman nu from the corpus-based lexical database Hownet (Dong & Dong 2000) Hownet 2000 contains 183 entries including the character nu. I listed these entries under different groups, and signalled the number of entries for each group as shown in Table 14.1. Translations in English are those provided by the database Hownet. Out of a total of 183 entries, I analysed in detail the list of 68 occupational activities containing the character nu with the twofold aim of verifying how gender is constructed in Chinese and highlighting the frequency of false generics and female specifics in the Chinese lexicon. What emerges from this is that by far the most productive pattern is the one already described, where the character nu is in an appositive position, as in (13):
(13) nu jdnwenjizhe nujingK woman/female news journalist woman/female manager 'newspaperwoman' 'manageress'
Only a limited number of entries - listed in (14) - are formed in a different way, that is & nu 'woman' is the head of the noun compound word and therefore in the right-hand position (on head position of Chinese compounds, see Ceccagno & Scalise 2005):
(14)
cdinu 'gifted female scholar' mityangnu 'shepherdess' zhlnu 'weaving woman' bdnii 'barmaid' jiuba nushi 'barmaid' shinu 'chambermaid' gpngnu 'a maid in an imperial palace' shinu 'maid in an imperial palace' shinu 'maid' Irinu 'servant girl' shuniii 'fair maiden' wunu 'dance-hostess' xiimu 'nun of Roman Catholic and Greek Orthodox Church'
Most of the entries listed in (14) are words originating from classical Chinese muydngnu 'shepherdess' and sonenu 'a maid in an (for example imperial palace') and names of mythological figures (for example zhmu, which means both 'weaving woman' and 'Vega'). Others, however, are recent loan translations from English, such as bdnu 'barmaid', where
CHINESE CECCAGNO
223
Table 14.1 Hownet entries including nu Subject
N of entries Example
Occupational entries Kinship entries Entries for 'woman' Entries containing both genders
68 28 15 14
Entries on physical characteristics 9 Derogatory entries 7 Marriage and marital status entries 6 Other entries 36
nugaoyin 'soprano' waishengnu 'niece' niiJang'girl' zhong nan qing nu 'regard men as superior to women' chunu 'maiden' nuliu 'the weaker sex' nu/ang'the bride's side' nuxiichengren 'heiress'
ba is the phonetic transcription for the English word 'bar' and jiuba nushi 'barmaid', where jiiiba includes the semantic character jiu 'wine' and the phonetic loan translation from English ba 'bar' (on phonetic/semantic loans in Chinese, see Abbiati 1992:106.)3 Subsequently I compared occupational terms including the character nu with occupational terms including the character nan 'man/male' in the Hownet database. The Hownet database provides only 54 entries for nan. Of these, only five entries - listed in (15) - refer to occupational activities:4
(15)
ndndiyln 'bass' ndngdoym 'tenor' ndnzhongyin 'baritone' ndnsheng 'male voice' ndnwu 'gigolo'
Furthermore, all of the entries in (15) are limited to performing activities (essentially music) and this is possibly explained by the importance of specifying gender in this field. The disparity in the number of occupations found for women and for men in the Hownet database can be a trustworthy indicator of the high frequency in Chinese of generic nouns that are understood to refer to men, and which must have a specification if they are to be understood as referring to women. In order to confirm these findings, I cross-checked the woman-specific occupational nouns found in the Hownet database with the eauivalerit occupations containing nan 'male/man' in Microsoft Pinym IME 2003. As a result, out of a total of 68 occupational entries referring to women, for the large majority - 39 entries - Microsoft Pinyin IME 2003 did not provide any male equivalent; while for 26 entries the male equivalent was provided.* It should be noted that the majority (14) of these 26 entries concerned perfonning artists or sport practitioners. And even for the limited number of
224
GENDER, LANGUAGE AND NEW LITERACY
occupational names - only 12 if those referring to performing arts and sport are not included - evidence still needs to be found for a genuinely neutral use when gender is not specified. The analysis of a corpus-based database therefore enables us to see the prevailing patterns of perception and use of person reference nouns in Chinese in a completely different perspective: traditionally described as gender neutral, in actual fact their frequent covert male gender overrides the neutral nature of the words. Our findings confirm those of previous investigations of covert male bias conducted by researchers analysing languages with limited or no grammatical gender such as Finnish and Turkish. These researchers have found that morphologically gender-neutral human nouns can have a male connotation (Braun 2001; Engelbere 2002; see also Castagneto & D'Amora). 3.5
mu 'mother' and
fu 'father'
The empirical analysis on gender bias can be further pursued by analysing other nouns connected with gender, for example the use of mil 'mother' and fu 'father'. A telling example is provided by the two compound words shlmu and shifu, which at first glance could be considered as a morphological word pair - since thev both indicate a profession followed by the terms usually used for 'mother' and 'father' - but they are not in a fair word pair. shimii 'teacher' 'mother' means 'the wife of one's teacher or master' and is used when formallv addressing one's teacher's wife; a more informal shiway of addressing the same person is the equivalent compound nidng is a form of address for an elderly married woman, niang, where besides also meaning 'mother', However, the corresponding form of address for 'the husband of one's fu - which teacher or master' does not exist in Chinese. When the term besides meaning 'father' is also a term of respect for an elderlv man - is used. the meaning is different from what we could expect, on the basis of shifu 'teacher' 'father', as well as the shlmu. In fact the compounds shifu, according to the dictionary (Dictionary Departhomophonous ment 2002) is a polite utle for one with accomplished skill. These two terms suggest that in the Chinese tradition, while for males the condition of teacher was or could be a straightforward apposition, as in the shl is in the appositive position, for shifu, where compound word women the condition of teacher seems simply not to exist: that is, a woman could at most give birth to or marry a teacher, not actually be a teacher herself. 3.6 Person or man ?
In spite of a structure where gender-neutral nouns should be the norm, Chinese is no exception in being a language permeated with socio-cultural
225
CHINESE CECCAGNO
values which heavily modify its bias-free potential. As in other languages, so in Chinese the male is perceived as the default human being, and this special status accorded to one gender overshadows the presence of the other. This is further evident if we single out the group of occupational words in the nu 'female/woman' for which the generic Hownet database containing ren, 'person', and look for the male equivalent in expression includes Microsoft Pinyin IME 2003. As a result, only for a limited number of nouns do the three forms (generic, female specific and male specific) exist, while for others male expressions and generics coincide. Examples of the first type are shown in (16a) and of the second type in (16b). Again, the English translation for female specifics is taken from Hownet. (16a)
'servant'
&I8/&I8A nuyong/ nuyongren woman/ female servant woman/ female servant person 'maidservant'
gongren work person 'worker'
nu gong
nan gong
female/woman work 'female/woman worker'
man/male work 'man/male worker'
yongren servant person
ndnyong/ ndnydngren man/male servant man/male servant person 'male servant'
(16b)
cdiren talent person 'gifted scholar' and (covert) 'gifted male scholar'
cdi nu talent woman/female person 'gifted female scholar'
shiren
nu shlren
poem person 'poet' and (covert) 'male poet'
woman /female poem person 'poetess'
Keren
nu Keren woman/female hunt person 'huntress'
hunt person 'hunter' and (covert) 'man/male hunter' tidojieren mediate person 'mediator'and (covert) 'man/male mediator'
nutidojieren woman/female mediate person 'mediatress'
These examples show that in actual fact. renis not really gender-indefinite, since it does not only mean 'person' but also 'man'. Therefore, expressions
226
GENDER, LANGUAGE AND NEW LITERACY
with ren 'person' can be ascribed to the covert gender group of nouns, as they have an inherent gender bias even though structurally they are neutral (see Braun 2001 and Castagneto and D'Amora for the same phenomenon in Turkish). 3.7 Male gentries in the Chinese language The identification of male with people - and vice versa - is also evident in male generics. This phenomenon has been found in a large number of languages (see for example, Braun 2001, on Turkish; Romaine 2001, on English; Engelberg 2002, on Finnish; Marcato & Thune 2002, on Italian). Chinese is no exception. For example, hdn - with the meaning of hdnzi or hdnren 'man', can be used as a male generic as the two following examples show:
(17)
menwaihan lit: 'door outside male', i.e. 'layman' ddnshenghdn 'single male', 'bachelor or man who is not living with his wife
menwaihan has lost its male specific meaning and can be used of a woman, who will therefore describe herself as 'lavman'. Besides, hdn 'male' and ren 'person' can be interchangeable, as can be seen in (18): (18)
woshl menwaihan I be door outside male 'I am a layman' wo shi menwairen I be door outside person 'I am a layperson'
ddnshenghdn 'bachelor' was also originally specifically used for men but now - with the increasing number of singles of both genders in China can also be used of women. The slightly derogatory nuance perceived bv Chinese people in this expression is sometimes neutralized by replacing gulzu 'noble', giving ddnshen guizu 'single noblehdn with man . gulzu 'nobleman' has itself a covert male gender, as confirmed by nugulzu Microsoft Pinyin IME 2003, which automatically provides female/woman noble but does not provide the equivalent for nobleman 3.8 Female nouns with gender-indefinite reference
In general, female nouns tend to be female-specific. For instance iidtingfunu 'house' 'woman' 'housewife' can only be used to refer to women: iidtinendnren 'house' 'male' 'house husband' the male equivalent not only is not found in the dictionary, but is also not automatically suggested bv Microsoft Pinyin IME 2003 when the fnnyin transcription is written. A neologism now widespread in China is kongjie 'air' 'older-sister' 'woman air attendant'. The form of address jie'older sister' is a form of respect and therefore used also when the attendant is very young. But age of
CHINESE CECCAGNO
227
women attendants seldom goes unnouced and therefore another neologism, k&nesdo lit. 'air father's older sister', 'air aunt', where age is stressed, has been coined. Both terms are found in Pinyin IME 2003, while the equivalent kongge 'air older brother' is still only poised for expression for males, take-off, and in fact is not automatically proposed by Pinyin IME 2003 when written in ptnym. For 'male air attendant' the generic expression chengwuyudn 'air attendant' seems to be preferred. While Pinyin IME 2003 nuchengwuyudn 'woman air attendalso provides the compounds ndnchengwuyudn 'male air attendant', in the Hownet dataant' and nuchengwuyudn 'woman air attendant' is found. This base only shows, again, that male and generic nouns can often overlap. The use of female nouns with gender-indefinite reference is a rare exception in most languages (see Hellinger & BuBmann 2002). One such case in baomu 'caretaker', a baomu, also written Chinese is the expression mu 'mother', muor boo 'to take care of, and compound formed by whose meaning, according to the dictionary (Dictionary Department 2002) is 'housemaid', 'housekeeper', 'female servant employed to care for children baomu is attested with the or to do housework'. In the Hownet database five following meanings: caretaker , homemaker housekeeper , mistress of the house', 'nanny', 'servant*. Consequently, it is a woman-specific referential word, with female semantic specification in the character mu, which is the head of the compound. However, in the course of time and due to a probably less strict gender distinction in this field, bdomu has increasingly tended to be used as a gender-indefinite noun. This evolution is confirmed bv the fact that both nubdomu 'female caretaker' and nan baomu 'male caretaker' are attested and proposed by Pinyin IME 2003 when written in pinym. nubdomu 'female caretaker' is pleonastic, nu, the 'woman' character, being the first constituent of the compound and one of the two elements of the second constituent. The pleonastic nature of this confirms the wider referential potential that has been acquired by what was formerly a female-specific expression. 4 Conclusio n This chapter challenges the traditional and widely diffused assumption according to which Chinese, being a language without inflection and without derivation, is a genuinely gender-neutral language. To this aim, and using a corpus-based database, this chapter has undertaken to verify how gender manifests itself in Chinese and has illustrated the main existing patterns. The main finding is that covert male gender abounds in terms for person reference in Chinese, as the neutral meaning is often associated with male gender. In fact, only for a limited number of nouns are the three forms generic, female-specific and male-specific - used, while for most nouns male expressions and generics coincide. This is particularly evident when analysing occupational nouns. Examples extracted from the lexicon in Microsoft IME showed that for a wealth of occupational terms what at first glance could
228
GENDER, LANGUAGE AND NEW LITERACY
be perceived as a genuine generic noun, including in its meaning both female and male reference, is instead a covert male generic. This finding was confirmed by the analysis of occupational terms contained in the database Hownet, which shows 68 entries containing the lexeme nu 'woman/ female* and only five entries containing the lexeme nan 'man/male'. Furthermore, all the occupational nouns containing nan 'man/male' were limited to performing activities, that is to fields where specifying female and male gender could be particularly important. The disparity in the number of occupations found for women and for men can be considered a trustworthy indicator of the high frequency in Chinese of generic nouns that are understood to refer to men, and which must have a specification if they are to be understood as referring to women. This chapter has therefore highlighted the prevailing patterns of use of person reference nouns in Chinese from a perspective completely different from the traditional one: normally described as gender-neutral, in actual fact their frequent covert male gender overrides the neutral nature of the words. This phenomenon of neutral meaning associated with male gender while female gender needs to be explicitly marked has never been pointed out before. Further research should address the magnitude of this phenomenon. This chapter has thus shown that Chinese, a potentially gender-free language, is in actual fact governed by socio-cultural factors, i.e. by socio-cultural assumptions and expectations about the relationship between women and men, and therefore can be gender-biased. This situation is even more pervasive and dangerous as the inherent gender bias remains hidden at first glance. The widespread use of the gender lexeme nu 'woman' with terms for person reference does not carry in itself derogatory connotations; rather, it is the fact that the male equivalent nan 'man' is not as often used with terms for person reference that creates an imbalance and therefore suggests a world where men are the linguistic norm, the centre, and women are something peripheral. These patterns are even more striking as they undermine the potentially gender-free nature of nouns in Chinese. The linguistic and socio-cultural gender-bias is both reflected in and strengthened by the most widely used computer writing program for writing in Chinese. This chapter has also shown that the most widely used word processing input method for Chinese - Microsoft Pinyin IME 2003 - is far from being objective or neutral. It has highlighted how Microsoft Pinyin IME 2003 - with its male-centred proposal of personal pronouns and its arbitrary suggestions of personal pronouns according to the verb following the pronoun - contributes to, and even fosters, a strongly unbalanced categorization and representation of the roles of women and men in Chinese. The impact of word processor input methods built with such a gender bias is even more pervasive as their use is not confined to certain specific tasks but rather is absolutely indispensable as a way of getting Chinese characters on the screen.
CHINESE CECCAGNO
229
Acknowledgements This research was made possible thanks to a financial contribution from the University of Bologna, Italy ('60 percent'). I wish to thank Jing Li, lecturer in Chinese at the University of Bologna, for suggesting examples of female and male generics. My thanks go also to Emiliano Guevara for discussing parts of this chapter with me. Notes 1 CRS stands for Currently Relevant State (see Li & Thompson 1981: 296 ff.). shi gong 'wizard' are used in 2 The two expressions shi po 'witch' and southern China. banii 'barmaid' is the abbreviated form of jiubd nushi 'barmaid'. 3 4 The following examples confirm that in the formation of male gender too the nan 'male/man' in the appositive position, while main pattern sees the noun examples of nan with 'male/man' as the head of the noun compound and ndnwii 'gigolo' is formed therefore in the left position are limited. In wunu 'dance hostess', the two being female and exactly in the same way as male nouns for 'dancer'. 5 For two entries the quesuon was not pertinent. 6 The compound nugpngren 'female/woman' - 'work' - 'person' 'female/ woman worker' also exists in Chinese, while the compound ndneongren 'male/man' - 'work* - 'person' 'male/man worker' is not found in Pinyin IME 2003.
References Abbiati, Magda (1992), La lingua cinese. Venezia: Cafoscarina. Braun, Friederike (2001), "The communication of gender in Turkish', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 283310. Ceccagno, Antonella & Scalise, Sergio (2005), 'Composti del cinese: analisi delle strutture e identificazione della testa', in Annamaria Palermo (ed.), La Cina e I'aUro. Naples: il Torcoliere. DeFrancis.John (1984), The Chinese Language: Fact and Fantasy. Honolulu: University of Hawaii Press. Dictionary Department, Institute of Linguistics, Chinese Academy of Social Sciences (2002), The Contemporary Chinese Dictionary, Chinese English Edition. Beijing: Foreign Language Teaching and Research Press. Dong Zhendong & Dong Qiang (2000), HowNet Knowledge Database. www.keenage.com Engelberg, Mila (2002), The communication of gender in Finnish', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 10932. Ettner, Charles (2002), 'In Chinese men and women are equal - or - women and men are equal?', in Marlis Hellinger & Hadumod BuBmann (eds). Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 29-54.
230
GENDER, LANGUAGE AND NEW LITERACY
Hellinger, Marlis & BuBmann, Hadumod (2002), 'Gender across languages: the linguistic representation of women and men,' in Marlis Hellinger &: Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 1-25. Holmes, Janet (2001), 'A corpus-based view of gender in New Zealand English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 115-36. Li, Charles N. & Thompson, Sandra A. (1981), Mandarin Chinese: A Functional Reference Grammar. Berkeley, CA: University of California Press. Marcato, Gianna & Triune, Eva-Maria (2002), 'Gender and female visibility in Italian', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages. The Linguistic Representation of Women and Men, Vol. 2. Amsterdam and Philadelphia: Benjamins, pp. 187-217. Romaine, Suzanne (2001), 'A corpus-based view of gender in British and American English', in Marlis Hellinger & Hadumod BuBmann (eds), Gender across Languages: The Linguistic Representation of Women and Men, Vol. 1. Amsterdam and Philadelphia: Benjamins, pp. 153-74.
Index
Page numbers in italics refer to tables, n indicates endnote(s). Ambadiang, T. 154,157 American English sec English, US Anastasiadi-Symeonidi, A. & ChilaMarkopoulou, D. 95,96 Ariel, M. & Giora, R. 169 asymmetrical masculine and feminine terms Chinese 221-4 Czech 31-4 Hungarian 190-1, 192, 193 Italian 112-16,117-18,119-20n Spanish 163-5 Australian English 66-7, 75 Barbosa,J. S. 139, 150, 151 n, 152n Bazzanella, C. etal. 9, 12,71, 107, 110, 159, 165-6 Manera, M. & 9, 12, 14, 16-17, 19n Benson, M. et al. 66 Bergvall, V.L. 12 Bettoni, C. 15, 18, 20n Braun, F. 3, 4, 19-20n Brenner, A. 171 BuBmann, H. &: Hellinger, M. 81,91-2 Hellinger, M. & 5, 8, 12, 32, 196 Burr, E. 18 Camugli-Gallardo, C. 9 Carvalho, H. de 142,151-2n Castagneto, M. & D'Amora, R. 12,14,16 Ceccagno,A. 13,17-18 Chan,M.K.M. 17 Chila-Markopoulou, D. 105n Anastasiadi-Symeonidi, A. & 95,96
Chinese 213-14 asymmetry 221^t father and mother 224 female nouns with gender-indefinite reference 226-7 gender-indefinite nouns 219-20 Hownet database 221-4,225 masculine generics 224-6 Microsoft Pinyin (IME) 2,3,18, 214-18 passim, 229n Microsoft Word 2003 214, 217 occupational roles 220, 221, 225, 226-7 person reference nouns 218-19 pronouns 215-18 visual level of gender bias 214-18 cognitive linguistics 4 Collins Cobuild 2 compounds German 81,91-2 Portuguese 141, 151n computer technology and culture 1-3 computer writing programs research 9-12, 18-19 results 12-18 see also specific languages Corbett, G. 27, 38 covert gender 5-6 crucial pairs Czech 28-30 Dutch 47-56 German 82-8 Greek 102-4 Hebrew 176-9 Italian 110-12,117, 119n
232
INDEX
Polish 129-32, 136n Portuguese 145-50 Spanish 159-66, 167n Turkish 203-5, 206-7, 2 10-1 In Crystal, D. 65, 67 culture and computer technology 1-3 language and gender 3-9 Czech 24-6, 38-9 asymmetry 31-4 epicenes, usage in text and interpretation 34-8 father and mother 30 feminine epicenes 35-8 grammatical gender: formal vs semantic reading 27-8 grammatical gender-marking 24-5 lexical gender/crucial pairs 28-30 masculine generics/epicenes 25, 34-5 orthography 25 Dalewska-Greh, H. 123, 125 derogatory terms 19-20n Dong, Z. & Dong, Q. 221 Dressier, W. and Thornton, A.M. 1 18n Dutch 41, 56-8 crucial pairs, analysis of 47-56 grammatical gender 41-3 lexical gender 43-6 occupational roles 44-6, 53-6 sexual emancipation 46-7 social and family roles 51-3 urouw and man 48-51 Eisner, P. 24, 25-6, 37 emancipation of women see social change endearments 20n English 62-4, 75-7 family relationships 73, 74 father and mother 68, 69, 73 gender reform 64-5 grammatical gender 63-4 language thesauri 67-8, 69 lexical gender 63-4 occupational roles 64 US Word 6 68-70 Word 6 to Word 11 71-3 Word 11 73-5 varieties of 65-7
English as a second language (ESL) 66 epicenes see feminine epicenes; masculine generics/epicenes Ettner, C. 18 family relationships Dutch 51-3 English 73, 74 Greek 100 Hebrew 175,176,177-8 Hungarian 192-3 Italian 114 Portuguese 143,145,152n Spanish 154,155,156,161-2 Turkish 206-7, 210-1 In see also father and mother father and mother Chinese 224 Czech 30 English 68, 69, 73 German 87-8,90-1 Italian 114 feminine epicenes, Czech 35-8 'feminization' of language 8 Fodor. I. 27 Frangoudaki, A. 101 Garcia Meseguer, A. 155, 157 Garcia Mouton, P. 156,157 gender belief system 3,19n gender categorization 2,5-7. 19n gender neutralization (degendering)
8-9 German compounds 81,91-2 crucial pairs 82-8.93n Duden 81-2,93n father and mother 87-8,90-1 grammatical gender 79-82,93n lexical gender 79-82 Movierung/Motion 80,91-2 occupational roles 80,88-90,92,93n social gender 80 Gerritsen, M. 8-9,15 Goffman, E. 6 Gorlach, M. 66, 67 Graddol, D. & Swann, J. 7 grammatical gender 5-6 Czech 24-5, 27-8 Dutch 41-3
INDEX English 63-4 German 79-82,93n Greek 96-7 Hebrew 171 Italian 107-8,118-19n Polish 124,135n Portuguese 139-40,141,151n Greek crucial pairs 102-4 family relationships 100 gender construction 98-100,105n grammatical gender 96-7 lexical gender 97-8 morphological gender 95-7,105n occupational roles 100-2,104-5 Guil, P. 9,12,14 Guiro,A.Z. 171 Gunthner, S 6-7 Halpern, D.F. 4-5 Handke, K. 127,136n Haney, L. 187 Hebrew 169-70,180n crucial pairs 176-9, 180-ln family relationships 175,176,177-8 gender construction 170-2 grammatical gender 171 lexical gender 171-2 Microsoft Word Thesaurus 174-80, 180-lw occupational roles and social change 172-4, 180n Hellinger, M. 44, 64, 67, 81 & BuBmann, H. 5, 8, 12, 32, 196 BuBmann, H. & 81, 91-2 Holmes.J. 4, 7,20n65,219 Horvath.A. 187-8 Hungarian asymmetry 190-1, 192, 193 crucial pairs 189-91 family relationships 192-3 gender features 182-6 lexical gender 182 Microsoft Word 2000188-93,194-5n occupational roles (UPN & MPN) 182-6,194n Ibrahim, M.H. 27 insults 19-20n Irigaray, L. 27
233
Italian 107, 118n asymmetry 112-16,117-18, 119-20n crucial pairs 110-12,117,119w family relationships 114 father and mother 114 grammatical gender 107-8, 118-19n lexical gender 108-9 occupational roles 110,115,119n social gender 109-10,119n Word Thesaurus 110-16 2000110-11 2004111-12 Jakobson.R. 31,32 Katsoyiannou, M. & Goutsos, D. 7, 8,14. 17, 20w Key,R.8 Koniuszaniec, G. & Blaszkowska, H. 126, 132 Krizsan, A. & Zentai, V. 188 Kuntjara, E. 20r< language families 10 Lavi, O. etal 176 Lee, D. 4 lexical gender 5, 6 Czech 28-30 Dutch 43-6 English 63-4 German 79-82 Greek 97-8 Hebrew 171-2 Hungarian 182 Italian 108-9 Polish 124-6,135 n Turkish 196-8 Linke.A. 13-14,125 Ldvia, A. 38 Livnat, Z. 13,14 McArthur, T. 66 Manera, M. & Bazzanella, C. 9,12,14, 16-17,19n Marcato, G. and Thune, E.-M. 108-9, 118,119w Marques, MA 12,14,17 marriage see family relationships masculine generics/epicenes 4-5, 7 Chinese 224-6
234
INDEX
Czech 25, 34-5 Portuguese 141-2, 151-2n 'meanings' 9, //, 17,19n, 20n Microsoft Pinyin 2, 3,18,214-18 passim, 229n Microsoft Word Thesaurus 2,3,9,11 see also computer writing programs research; specific languages Miecznikowski.J. 14-15,16,17 Miemietz, B. 132,135 Moliner, M. 156 monarchy Czech 33-4,36 Dutch 54 Moore, B. 66 morphological gender Greek 95-7,105n Portuguese 140 mother see father and mother Ms 7-S, 65 multicultural approach 12 Nagy Gabor, O. 194 Nagy, M. & Patti, V. 12,14,15,16 Nikiforidou, K. 100 Nissen, U.K. 153,154,156 occupational roles 14—15 Czech 31,32,33 Dutch 44-6, 53-6 English 64, 72, 74, 75 German 80-1, 88-90,92 Hebrew 172 Hungarian 182-6 Italian 110, 115, 119n Polish 128, 129,132-4, 136n Turkish 202-5, 209-lOn Pauwels, A. 65 Polish 123, 135n crucial pairs 129-32,136n grammatical gender 124,135n lexical gender 124-6,135n Microsoft Word 2000127-35 occupational roles 128,129,132-4, 136n personal pronouns 127-9,136n women's movement 126-7, 135-6n Pollen, A. &Fodor,E. 187 Portuguese
compounds 141, 151n crucial pairs 145-50, 152n family relationships 143, 145, 152n masculine form functions 141-2, 151-2n natural and grammatical gender 139-40,141,151n noun gender-syntagmatic category 140-1 occupational roles 143,145,149-50, 152n social relationships 148-9 Thesaurus Word 9144-51 professional roles see occupational roles psycholinguistic tests 4-5 Puskulluoglu, A. 204 Romaine, S. 2,3,5,8, 18,19n64,65 Sapir, E. 3 Schwarzwald, O. 170 semantics 4 Silveira.J. 5 Sobol, E. 129 social change England 64-5 Holland 46-7 Hungary 186-8 Israel 172-4,180n Poland 126-7, 135-6n Spain 156-9, 167n social gender construction 5, 6-7 social roles see family relationships; father and mother; occupational roles Spanish 153, 167n asymmetry 163-5 crucial pairs 159-66,167n family relationships 154, 155,156, 161-2 Microsoft Word Thesaurus 159-67 6.0159-62, 166-7,167n 11.0162-6, 167-8n occupational roles 155,157-9,165-6 social roles and social change 156-9, 167n 'synonyms' 11, 15-16,19n see also crucial pairs Tarif, R. 68 Tejada, P. 9
235
INDEX
Teubert, W. 20n Thune, E.-M. & Leonard!, S. 13,16. 20n Triantafyllidis, M. 96,99 Turkish 'l96, 208-9n crucial pairs 203-5, 206-7, 210-1 In family roles 206-7, 210-1 In first crucial semantic field 198-201, 209n lexical gender 196-8 linguistic reform: 'Mr' and 'Miss'/'Mrs' 201-2, 209n loanwords 197, 209n Microsoft Word Thesaurus 198-208 occupational roles 202-5, 209-lOn
suffixes 197, 209n US English see English, US van Alpen, I. & Corda, A. 13, 15-16 Violi.P. 118n Vostantzoglou, T. 99-100, 102 West, C. & Zimmerman, D. 6 women linguistic strategies towards visibility of 7-9 reference terms for men and 7, 11, 13-18 women's movement see social change