Grammatical Gender in Interaction

Grammatical Gender in Interaction
Brill’s Studies in Language,

Cognition and Culture
Series Editors
Alexandra Y. Aikhenvald (Cairns Institute, James Cook University)

R.M.W. Dixon (Cairns Institute, James Cook University)
N.J. Enfield (Max Planck Institute for Psycholinguistics, Nijmegen)
Willem Adelaar (Leiden University)
Carol Genetti (University of California, Santa Barbara)
Bernd Heine (University of Cologne)
Rosita Henry ( James Cook University)
John Lucy (University of Chicago)
Lev Michael (University of California, Berkeley)
Ton Otto (Aarhus University/James Cook University)
Bambi Schieffelin (New York University)
Masayoshi Shibatani (Rice University / Kobe University)
Anne Storch (University of Cologne)
Peter Trudgill (University of Fribourg/University of East Anglia)
Anthony Woodbury (University of Texas, Austin)
Volume 9
The titles published in this series are listed at brill.com/bslc

Grammatical Gender
in Interaction
Cultural and Cognitive Aspects
By
Angeliki Alvanoudi
LEIDEN | BOSTON
On the cover: “Communication No. 1”, artwork by Michele Meister. Picture courtesy of Michele Meister.
Library of Congress Cataloging-in-Publication Data
Alvanoudi, Angeliki.
Grammatical gender in interaction : cultural and cognitive aspects / By Angeliki Alvanoudi.
p. cm. — (Brill’s Studies in Language, Cognition and Culture)
Includes bibliographical references and index.
ISBN 978-90-04-28314-5 (hardback : alk. paper) — ISBN 978-90-04-28315-2 (e-book) 1. Greek language.
Modern—Grammar. 2. Greek language, Modern—Noun. 3. Greek language, Modern—Noun phrase.
4. Greek language—Sex differences. 5. Greek language, Modern—Psychological aspects. I. Title.
PA1057.A48 2014
489’.3045—dc23
2014034959
This publication has been typeset in the multilingual ‘Brill’ typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities.
For more information, please see brill.com/brill-typeface.
issn 1879-5412
isbn 978-90-04-28314-5 (hardback)
isbn 978-90-04-28315-2 (e-book)
Copyright 2015 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Brill Nijhoff, Global Oriental and Hotei Publishing.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive,
Suite 910, Danvers, MA 01923, USA. Fees are subject to change.
This book is printed on acid-free paper.

Contents
Acknowledgements ix
Phonetic Symbols x
List of Abbreviations xi
Transcription Conventions xii
1 Introduction 1
1.1 Contextualizing the Problem 1
1.2 On Grammatical Gender and Culture 4
1.3 On Grammatical Gender and Cognition 6
1.4 Summary of the Argument to be Presented in the Book 8
2 Grammatical Gender 13
2.1 Introduction 13
2.2 Noun Classes or Genders 13
2.3 Grammatical Gender in Modern Greek 19
2.4 Grammatical Gender and Reference: Preliminary Remarks 26
2.5 Summary 31
3 Grammatical Gender and the Intersection of Sociocultural

Gender/Sex 32
3.1 Introduction 32
3.2 The Intersection of Gender/Sex in Feminist Theory 33
3.3 Cultural Aspects of Grammatical Gender: The Perspective
of Linguistics 36
3.3.1 Grammatical Gender and the Construction of Gender
Identities 37
3.3.2 Grammatical Gender and the Maintenance of Gender
Hierarchy 40
3.4 Cultural Aspects of Grammatical Gender: The Perspective of
Feminist Non-Linguistic Theories 47
3.5 Summary 50
4 Grammatical Gender and Speakers’ Cognition 52

4.1 Introduction 52
4.2 Cognitive Aspects of Grammatical Gender 52
4.2.1 Grammatical Gender and Conceptual Categorization 52
4.2.2 Grammatical Gender and the Linguistic Mediation of
Thought 59
vi Contents
4.3 Indications of the Cognitive Role of Grammatical Gender:

Empirical Studies 65
4.4 Summary 71
5 Grammatical Gender, Culture and Cognition in Interaction 72

5.1 The Research Question 72
5.2 Grammatical Gender and Person Reference in Interaction 74
5.2.1 Indexing Gender: The Role of Grammatical Gender 79
5.2.2 The Generic Use of the Masculine Gender and Default
Inferences 81
5.3 Approaching Grammatical Gender in Interaction through the Lens of
Conversation Analysis 84
5.3.1 Gender Membership Categories and Grammatical Gender 88
6 The Empirical Investigation of Grammatical Gender in

Interaction 92
6.1 Introduction 92
6.2 The Routine Achievement of Sociocultural Gender in Interaction:
The Role of Grammatical Gender 95
6.2.1 Self-reference and Grammatical Gender 95
6.2.2 Reference to Recipient(s) and Grammatical Gender 106
6.2.3 Reference to Third Person(s) and Grammatical Gender 111
6.2.4 Grammatical Gender and the Relevance of Sociocultural
Gender 116
6.2.5 The Non-match between Grammatical Gender and Referents’
Sex 123
6.2.6 Conclusions 128
6.3 Indications of the Cognitive Role of Grammatical Gender in
Interaction 129
6.3.1 Repair: Introductory Remarks 129
6.3.2 Grammatical Gender as an Item of Repair 133
6.3.3 Other Indications 154
6.3.4 Conclusions 164
7 What Can We Conclude? 166

7.1 ‘Thinking for Speaking’ in Interaction 166
7.2 Sexism in Interaction: Fact or Fiction? 168
Contents vii
7.3 Reconsidering Gender Performativity 171

7.4 Grammatical Gender in Interaction: A Summary 175
References 177
Index of Authors 195
Index of Subjects 198
Acknowledgements
The study presented in this book is based on my PhD dissertation at the

Aristotle University of Thessaloniki. Writing a PhD dissertation is a lonely pro-
cess, which involves many hours in front of books and computer screens. In
reality though, it is a collectively produced work, as most academic work is.
This section is about acknowledging the contribution of scholars and others
who influenced this study and made the journey less lonely and more joyful
for me.
I wish to express my warm thanks to the supervisory committee for their guid-
ance and support: Theodossia-Soula Pavlidou (supervisor), Dimitra Kati (co-
supervisor) and Savas Tsohatzidis (co-supervisor). Theodossia-Soula Pavlidou
has been more than just a thesis-supervisor for me. She has been my teacher
since the very early years of my undergraduate studies at the Department of
Linguistics, in the Aristotle University of Thessaloniki. She introduced me to
language and gender research through her academic work and teaching, and
inspired my work and thinking to a very large extent. Dimitra Kati informed my
perspective on language and cognition, by shedding light on various aspects of
this complex issue and by generously sharing her ideas with me.
I gratefully acknowledge the input of my colleagues at the Institute of
Modern Greek Studies: Lena Gialabouki, Charikleia Kapellidi, and Eleni
Karafoti. I thank two feminist scholars, Mia Liinason and Maria do Mar Pereira,
for passionate discussions on gender and interdisciplinarity, and for dreaming
together better European academic contexts.
I am grateful to Alexandra (Sasha) Aikhenvald for encouraging me to submit
this study to Brills and for giving me insightful comments on an earlier draft of
the manuscript. I thank the two anonymous reviewers for their comments and
suggestions that improved the manuscript. Thanks also go to Brigitta Flick and
Venetia Jaschke for their assistance with proofreading the text.
Finally, I wish to thank my friend Vaso Karagouni, my sister Nancy, and my
parents Anthouli and Nikos for their unconditional love and support.
Phonetic Symbols
In this book, I use broad transcriptions based on the International Phonetic

Alphabet (ipa).
c voiceless palatal stop

k voiceless velar stop
x voiceless velar fricative
ç voiceless palatal fricative
ʝ voiced palatal fricative
ɣ voiced velar fricative
g voiced velar stop
ɟ voiced palatal stop
f voiceless labiodental fricative
v voiced labiodental fricative
b voiced bilabial stop
p voiceless bilabial stop
t voiceless dental stop
θ voiceless dental fricative
ð voiced dental fricative
d voiced dental stop
l alveolar lateral approximant
ʎ palatal lateral approximant
m bilabial nasal
n alveolar nasal
ɲ palatal nasal
ŋ velar nasal
r alveolar trill
s voiceless alveolar fricative
z voiced alveolar fricative
t͡s voiceless alveolar affricate
d͡z voiced alveolar affricate
a low central unrounded vowel
e mid front unrounded vowel
i high front unrounded vowel
o mid back rounded vowel
u high back rounded vowel
List of Abbreviations
1 first person
2 second person
3 third person
acc accusative
adv adverb
conj conjunction
cop copula
dat dative
fem feminine
fut future
gen genitive
imperat imperative
imperf imperfect tense
indef indefinite
interj interjection
ipfv imperfective aspect
masc masculine
neg negation
neuter neuter
nom nominative
part participle
particle particle
past past tense
pfv perfective aspect
pl plural
poss possessive
prep preposition
pres present tense
pron pronoun
rational rational
rel relative
sg singular
subj subjunctive
voc vocative
Transcription Conventions
I follow the transcription conventions used in the Corpus of Spoken Greek of

the Institute of Modern Greek Studies. The Corpus of Spoken Greek adopts the
standard Conversation Analysis conventions with certain modifications.
1 Temporal and Sequential Relationships
[ left brackets: point of overlap onset between two or more

[ utterances (or segments of them)
] right brackets: point of overlap end between two or more

] utterances (or segments of them)
= The symbol is used either in pairs or on its own.
A pair of equals signs is used to indicate the following:

1. If the lines connected by the equals signs contain utter-
ances (or segments of them) by different speakers, then the
signs denote ‘latching’ (that is, the absence of discernible
silence between the utterances).
2. If the lines connected by the equals signs are by the same
speaker, then there was a single, continuous utterance with no
break or pause, which was broken up in two lines only in order
to accommodate the placement of overlapping talk.
The single equals sign is used to indicate latching between two

parts of the same speaker’s talk, where one might otherwise
expect a micro-pause, as, for instance, after a turn construc-
tional unit with a falling intonation contour.
(0.8) Numbers in parentheses indicate silence, represented in tenths
of a second. Silences may be marked either within the utterance
or between utterances.
(.) micro-pause (less than 0.5 second)
Transcription Conventions xiii
2 Symbols and Combinations of Symbols for Representing Various

Aspects of Speech Delivery
punctuation indication of intonation, more specifically,

marks
. the period indicates falling/final intonation,
? the question mark indicates rising intonation,
, the comma indicates continuing/non-final intonation,
¿ the inverted question mark indicates a rise stronger than a
comma but weaker than a question mark
: Colons are used to indicate the prolongation or stretching of
the sound just preceding them. The more colons, the longer
the stretching.
word Underlining is used to indicate some form of emphasis, either
WOrd by increased loudness or higher pitch. Especially loud talk
WOrd may be indicated by upper case, whereas, in extreme cases,
upper case may be underlined.
° The degree sign is used to indicate the onset of talk that is
markedly quiet or soft. When the end of such talk does not
coincide with the end of a line, then the symbol is used again
to mark its end. When there are two degree signs, the talk
°° between them is a lot softer than the talk around it.
- A hyphen after a word or part of a word indicates a cut-off or
interruption.
_: Combinations of underlining and colons are used to indi-
cate intonation contours. If the letter(s) preceding a colon is
underlined, then there is prolongation of the sound preceding
it and, at the same time, a falling intonation contour.
: If the colon itself is underlined, then there is prolongation of
the sound preceding it and, at the same time, a rising intona-
tion contour.
↑↓ The arrows mark sharp intonation contours. The upper arrow
indicates sharp intonation rises, whereas the down arrow
indicates sharp intonation falls.
>word< The combination of ‘more than’ and ‘less than’ symbols indi-
cates that the talk between them is compressed or rushed.
<word> The combination of ‘less than’ and ‘more than’ symbols
indicates that the talk between them is markedly slowed or
drawn out.
xiv Transcription Conventions
h Hearable aspiration is shown with the Latin letter h. Its

repetition indicates longer duration. The aspiration may
represent inhaling, exhaling, laughter, etc.
.h If the aspiration is an inhalation, then it is indicated with a
period before the letter h.
^ Sandhi.
3 Other Markings
((laughs)) Double parentheses and italics are used to mark meta-

linguistic, para-linguistic and non-conversational descrip-
tions of events by the transcriber.
(. . .) The parentheses indicate that something is being said, but no
hearing can be achieved.
(word) Words in parentheses represent a likely possibility of what
was said.
(word)/ Alternative hearings of the same strip of talk are displayed in
(ward) parentheses separated by a slash.
CHAPTER 1
Introduction
1.1 Contextualizing the Problem
In the beginning of his book We have never been Modern, Bruno Latour (1993)
wonders whether the ozone layer is an object of study for chemistry, meteorol-
ogy, politics or economics, or whether it constitutes a hybrid, that is, a sort of
cultural-natural network that transgresses disciplinary boundaries and com-
bines elements of knowledge from all different disciplines mentioned above.
Speaking about the ozone layer may be a bizarre way to start a book on linguis-
tics but it is not irrelevant. Latour’s rhetorical question is useful for linguists
working on the relation between language, cognition and culture, because it
opens a window to the intersections that characterize the language-cognition-
culture complex and sheds light on its interdisciplinary nature. Based on
Latour’s approach, the interplay between language, cognition and culture can
be conceptualized as a sort of hybrid that transgresses disciplinary boundar-
ies. For example, it can be explored by linguists, anthropologists or cognitive
scientists, and when it comes to linguistics, it can be examined by linguists
specializing in psycholinguistics, sociolinguistics, cognitive linguistics, prag-
matics or anthropological linguistics. Being a mental phenomenon grounded
in sociocultural practices, language creates a wide range of interconnections
with culture and cognition. In this book, I aim to examine how aspects of these
interconnections manifest themselves in interaction, by focusing on a specific
grammatical feature, gender.
Grammatical gender is a noun class system of two or three distinctions,
which always include the feminine and the masculine (Aikhenvald 2000).
It constitutes an inherent property of the noun, which controls grammati-
cal agreement between the noun and other elements in the noun phrase or
the predicate (Aikhenvald 2000; Corbett 1991; Hellinger and Bussmann 2001).
Nouns may be assigned to specific genders according to semantic, morphologi-
cal and phonological rules. In Modern Greek, the grammatical gender system
includes three distinctions, the masculine, the feminine and the neuter, and
it is inflected in a vast number of linguistic items, such as nouns, adjectives,
participles and pronouns (Pavlidou 2003). In Greek, grammatical gender is
considered to be semantically arbitrary, because gender assignment in nouns
denoting inanimate referents does not follow any straightforward semantic
rules. For instance, nouns denoting physical entities can be feminine (βροχή
© koninklijke brill nv, leiden, ��5 | doi ��.��63/9789004283152_002

2 chapter 1
[vroçí] ‘rain’), masculine (ωκεανός [οceanós] ‘ocean’) or neuter (δέντρο [ðéndro]

‘tree’). Yet, in reference to humans there is a semantic basis, because generally
nouns denoting male humans are grammatically masculine (αθλητής [aθlitís]
‘male athlete’) and nouns denoting female humans are grammatically femi-
nine (αθλήτρια [aθlítria] ‘female athlete’). Thus, to a certain extent grammatical
gender displays a relation to meaning and more specifically to the semantic
distinction of sex. Gender assignment is usually semantically motivated when
it comes to reference to human beings.
The term ‘gender’ is also used in the literature to denote the lexical marking
of sex, the biological distinction of sex, and the social categories of women and
men. Lexical gender refers to the lexical marking of nouns as female or male-
specific. For instance, in English the nouns mother, sister and father, brother
carry the semantic property of femaleness and maleness respectively and are
pronominalized as feminine (she) or masculine (he) (Hellinger and Bussmann
2001). This does not apply to nouns such as citizen, patient and individual
which are gender-indefinite.1 In languages with grammatical gender, the femi-
nine or the masculine gender of the noun corresponds to the lexical marking
of the noun as female or male (Ηellinger and Bussmann 2001, 5). For instance,
in Greek the grammatically feminine nouns θεία [θía] ‘aunt’, μητέρα [mitéra]
‘mother’ and the grammatically masculine nouns θείος [θíos] ‘uncle’, πατέρας
[patéras] ‘father’ are also lexically marked as female and male respectively.
Another term often associated with grammatical gender is natural gender.
Natural gender or sex refers to the anatomical/biological differences between
female and male humans (sex is the term to be employed in this book).
Although biology does not yield a perfect dichotomy of bodies into females
or males (e.g. Fausto-Sterling 2000), feminist theorists have shown that this
binary sexual split is taken as a given and then used as the basis on which socio-
cultural gender is built (the term gender will often be used as an abbreviation
for sociocultural gender). Sociocultural gender refers to “the many and com-
plex ways in which social differences between the sexes acquire a meaning and
become structural factors in the organization of social life”, in Braidotti’s (2000,
189) words. This means that gender is a cultural and historical product, rather
1 However, gender-indefinite nouns in English can be pronominalized by male-specific or

female-specific pronouns depending on their meaning (Hellinger and Bussmann 2001, 10–11;
McConnell-Ginet 2014, 27–28). For example, very often nouns denoting higher status occupa-
tions, such as surgeon or lawyer, are pronominalized by the male-specific pronoun he, while
nouns denoting lower status occupations, such as secretary or babysitter, are pronominalized
by the female-specific pronoun she, because these occupations are stereotypically associated
with men and women respectively.
introduction 3
than a static essence or a given attribute. As Widerberg (1998, 134) notes, the
English word gender was used primarily in grammatical and literary contexts
and was adopted by American feminists in the 1970s to define sex in a social
sense. Contrary to English, the biological vs. social distinction is not codified
in other languages. For example, in Norwegian, Danish and Swedish the words
kjønn, køn, and kön cover the meaning of both ‘sex’ and ‘gender’ (see Braidotti
2002a for a discussion of the translation of the English term ‘gender’ in vari-
ous European languages). A discussion about the equivalent terms in Greek
is found in Pavlidou (2000). The Greek equivalent term of the English term
‘gender’ is γένος [ʝénos]. This word refers to ‘a general concept in whose exten-
sion specific concepts are contained’, ‘a group of people with common descent’
or ‘ethnic group’, and in grammar contexts it refers to the grammatical category
of gender: γραμματικό γένος [ɣramatikó ʝénos] (Pavlidou 2000, 42). However, the
term γένος is not used for reference to social gender. Sex and social gender are
denoted in Greek via the term φύλο [fílo], which is the equivalent of the English
term ‘sex’. For example, the expressions αρσενικό φύλο [arsenikó fílo] ‘male sex’
and θηλυκό φύλο [θilikó fílo] ‘female sex’ refer to the biological classification of
humans as male or female, while the expression κοινωνικό φύλο [cinonikó fílo]
‘social gender’ refers to the social norms and ideologies associated with this
biological classification.
In this book, I examine the relation between grammatical gender and socio-
cultural gender through the lens of the relation between grammatical gender
and cognition. In the following, I present a ‘map’ of the intellectual trajectory
that has been followed in this book as well as the reasons that motivated my
interest in grammatical gender in the first place. To a very large extent, this
trajectory also forms a politics of location, in Braidotti’s (2002b) terms, that is,
a sort of cartography about the spatio-temporal territory that I share together
with other women and feminist scholars, and about my situatedness in
specific disciplinary contexts. My aim in undertaking this research project is
to study the ways in which language contributes to the construction of socio-
cultural gender and the maintenance of gender inequality. My investigation
started with the following questions. How does grammatical gender contribute
to the construction of sociocultural gender and the reproduction of gender
asymmetry? How is this aspect related to the role of language in mediating
speakers’ interpretation of experience more generally? What sorts of answers
can we give if we examine empirically the use of grammatical gender in inter-
action, that is, in the locus where speakers construct various aspects of their
daily social life? In the next two sections, I sketch out some interesting points
raised in the literature with respect to these questions. These points informed
the argument to be presented in this book.
4 chapter 1
1.2 On Grammatical Gender and Culture
The early writings of feminist linguists in the 1970s and the beginning of the
1980s focus on the role of language as a system in reflecting or mirroring socio-
cultural gender and sustaining gender inequality. I have two scholars in mind
here: Robin Lakoff (1975) and Dale Spender (1980). Both scholars addressed
issues of linguistic sexism in their work. Lakoff (1975) pointed to the negative
representation of women in the lexicon and argued that women experience
inequality in the way in which they are treated by language itself. Spender
(1980) uncovered the role of language in constructing and enforcing a man-
made worldview and sustaining men’s domination and women’s subordina-
tion. For these scholars, language contributes to legitimizing power at the
social level. Therefore, language reform is treated as a crucial element in the
feminist struggle for social change (Spender 1980).
One of the key issues that feminist linguists initially explored was the lexical
and grammatical marking of sex in language. Feminist linguists pointed out
that the semantic distinction of female or male sex can be grammaticized or
be part of the noun’s lexical meaning, and they explored the implications that
these lexico-grammatical features have for the way in which the sociocultural
world is represented and reproduced through language (e.g. Eakins and Eakins
1978; Graham 1975; Miller and Swift [1981] 1988). For example, they focused
on the generic use of the masculine gender, that is, the use of the masculine
gender for reference to female plus male referents or referents whose sex is
unknown. They criticized the generic use of the masculine as a prescriptive
practice, which sustains the social hierarchy between men and women, or the
“Patriarchal Universe of Discourse”, in Penelope’s (1990, xxvi) words.
Lakoff’s (1975) and Spender’s (1980) work inaugurated a long tradition of
research on language and gender which, according to Freed (2003, 701), has
been dominated by three major themes: deficit, dominance, and difference
theory. According to deficit theory, women’s language is ineffective in com-
parison to men’s, and reflects their socially inferior position (e.g. Lakoff 1975).
According to dominance theory, the ways in which women and men use lan-
guage reflect power and inequality at the social level (e.g. Thorne and Henley
1975). Finally, according to difference theory, women and men use different
verbal strategies because they have been socialized in same-sex childhood peer
groups (e.g. Maltz and Borker 1982; Tannen 1990). Freed (2003, 702) argues that
in general these theories have approached gender in essentialist ways. They
presuppose a binary way of thinking about gender, treating men and women
as static categories, while little variation is acknowledged within each category.
introduction 5
However, recently there has been a shift in language and gender research
from “essentialist and dichotomous conceptions of gender to a differentiated,
contextualized and performative model which questions generalized claims
about gender” (Holmes and Meyerhoff 2003, 7). This turn has been informed
by Butler’s ([1990] 1999) theory on gender performativity. Instead of treating
gender as a given attribute that exists prior to language and society, linguists
started examining the practices that continuously produce and maintain gen-
der as a fluid and diverse category, and the “linguistic resources” that speakers
“deploy to present themselves as certain kinds of women or men”, in Eckert and
McConnell-Ginet’s words (2003, 5). As Eckert and McConnell-Ginet (1992, 462)
point out, gender cannot be isolated from other aspects of social identity and
relations, such as age, class or ethnicity. The meaning of gender and the lin-
guistic manifestations of that meaning vary across communities, such as fam-
ily, school or work place, in which speakers may participate in different ways.
Linguists working in the context of this new ‘paradigm’ have shifted their focus
from language system to language use (Pavlidou 2011, 412) and from issues of
representation to issues of construction at the micro-level of interaction.
More specifically, interaction and gender is explored by feminist conversa-
tion analytic studies (e.g. Kitzinger 2000; Speer 2005; Speer and Stokoe 2011).
These studies examine gender as an accomplishment, a process that is con-
structed and maintained through the practices that participants employ in
interaction. Speer and Stokoe (2011, 14) summarize some of the questions that
have been explored by this strand of research as follows. What practices do
speakers employ in order to “make a world of two sexes appear natural and
inevitable”, and ‘pass’ as male or female in interaction? How is gender ‘done’ or
accomplished in interaction? What counts as ‘orientation to gender’ in inter-
action? How can analysts show that gender is interactionally relevant? How
does power manifest itself in interaction? If participants do not orient to their
talk or the talk of others as sexist, can analysts claim that sexism has actually
occurred in interaction? These questions will be highly relevant for the topic
examined in this book.
Overall, grammatical gender in interaction emerges in the space created in-
between the two broad tendencies in language and gender research that were
discussed above. Studying grammatical gender in interaction foregrounds both
language system and language use, and invites scholars to ask how the use of
this grammatical feature in interaction structures the practices that speakers
employ, and what are the social and cognitive aspects of this use. Traces of this
line of thinking can be found in the following extract written by McConnell-
Ginet ([1988] 2011, 39):
6 chapter 1
How do grammars, mental representations of linguistic systems, connect

to other modules of the mind (e.g., those involved in social cognition; in
person reception, in the planning of intentional action)? How do minds
connect to each other through language use? [. . .] How are social and
linguistic change connected to one another? What role does language use
play in social categorization and cultural evaluation of its users? More
generally, to what extent are patterns of language use reflective of social
structure and of cultural values, of inequality and oppression? Can lan-
guage be in part constitutive of culture and society, of women and men
and their relationships?
These questions have undoubtedly inspired my work.

But how does cognition fit in the larger picture of grammatical gender and
culture? Sociolinguistic approaches to gender address the aspect of cognition
in indirect ways, by considering language reform practices as a tool for gener-
ating or facilitating social change, and by claiming that language constructs
and maintains gender inequality. When feminist linguists formulate specific
guidelines for the elimination of sexist language use, they seem to imply that
language plays a key role in speakers’ conceptualization of the world. For lan-
guage to be ‘culturally’ significant, it must be also ‘cognitively’ salient.
1.3 On Grammatical Gender and Cognition
The assumption that language plays a role in speakers’ understanding of the

world is supported by empirical sociolinguistic and psycholinguistic studies
that report a correlation between grammatical gender and the interpretation
of referent’s sex. For example, the generic use of the masculine gender is shown
to correlate with speakers’ strong tendency to interpret referents as male (e.g.
studies reported in Doleschal and Schmid 2001 and Nissen 2002). Moreover, a
number of psycholinguistic studies (e.g. Boroditsky, Schmidt, and Phillips 2003;
Sera et al. 2002) report indications of a grammatical gender ‘effect’ on speakers’
conceptualization of the inanimate world as ‘female’ or ‘male’. For example,
speakers of various languages are shown to attribute sex to inanimate objects
according to the grammatical gender of the nouns denoting these objects.
The relation of grammatical gender to speakers’ cognition is part of the
broader question about the role that language system plays in speakers’ cog-
nition. As Lucy (1996, 39) argues, the semiotic nature of language may have
interesting implications for the way in which speakers interpret experi-
ence and for the sociocultural world they live in. This view is known in the
introduction 7
literature as semiotic relativity (Lucy 1996). Is it possible that “the use of the
semiotic form we call language in and of itself fundamentally alters the vision
of the world held by humans in contrast to other species” (Lucy 1996, 38–39)?
In order to reflect on the symbolic component of language, Lucy (1996, 40)
draws on Peirce’s classification of icons, indexes and symbols. According to
Peirce (1940), icons and indexes are signs related to the objects that they denote
because of natural similarity and physical co-presence respectively. Symbols
are signs standing for their objects because of convention, or “by virtue of a
law [. . .] which operates to cause the Symbol to be interpreted as referring to
that Object” (Peirce 1940, 102). Language is a symbolic medium that maintains
aspects of icons and indexes. It socializes or objectifies individual activities,
including thought, and it is a flexible signaling mode that allows for diversity
to occur at all levels across languages (e.g. morphological, semantic but also
functional), among other things (Lucy 1996, 40).
Due to these semiotic properties of language, a certain kind of linguocen-
trism arises in cultural and cognitive phenomena. As Enfield (2000, 126) points
out, it is difficult “to isolate anything cognitive or cultural which is not already
imbued with language at a profound level”. This view is supported by the fact
that culture involves semiotic processes, which help humans share ideas and
beliefs, and that the basic semiotic tool that humans employ for this purpose
is language. In Enfield’s words,
while thoughts or private events independent of semiotic/linguistic

material are possible (. . .), for them to serve as cultural background, they
must be shared, and, further, be assumed to be shared. To achieve this
recursively cognizant shared-ness, some material with semiotic potential
must serve as a medium for individuals to use in aligning private repre-
sentations. (Enfield 2000, 131)
Semiotic relativity serves as a background assumption in my research, as a sort

of principle on the basis of which more specific questions can be addressed
with respect to how the relation between grammatical gender, culture and cog-
nition manifests itself in Greek.
Grammatical gender can be used as a case study for exploring how lan-
guage relates to cognition because of its semantic basis and its systematic and
obligatory use. The match between grammatical gender and referent’s sex in
reference to human beings foregrounds the relation of grammatical gender
to meaning, and, thus, to conceptualization. Following cognitive linguistics
(Croft and Cruse 2004), meaning here is understood as construal, that is, as
a process of re-structuring and interpreting experience in specific ways. The
8 chapter 1
relation between grammatical gender and conceptualization is further sup-

ported by the fact that gender is an inherent property in language system, a
morphological category that marks numerous linguistic items and that speak-
ers must use when they speak or write. The role of grammatical categories in
mediating speakers’ cognition has been examined by research on linguistic
relativity (Boas 1938; Sapir [1949] 1970; Whorf 1956). Linguistic relativity refers
to the correlation between cross-linguistic and cognitive differences, and pre-
supposes semiotic relativity. That is, it presupposes the mediation of language
in the interpretation of experience. Research on linguistic relativity shows that
grammar provides a set of options for schematicizing experience for verbal
expression and that grammatical categories guide speakers to unconscious
interpretations of experience, because they are systematic, obligatory and
automatic. The mediation of language to speakers’ cognition is expected to
take place at least when speakers think for speaking, as Slobin (1996; 2003)
shows. Given the approaches mentioned above, the following question arises:
do languages with systems of gender, such as Greek, orient speakers to spe-
cific ways of interpreting experience, that is, to the interpretation of person
as female/woman or male/man? This question will be examined in this book.
1.4 Summary of the Argument to be Presented in the Book
The book is divided into two main parts. Chapters 2, 3, 4 and 5 constitute the
theoretical part of my research, while chapters 6 and 7 constitute the empirical
part (data analysis and conclusions).
More specifically, in chapter 2, I describe the properties of gender, by focus-
ing on the features of the grammatical gender system in Modern Greek and
the ways in which grammatical gender is related to reference in general. The
key point made in this chapter is that grammatical gender in person reference
codifies the semantic distinction of sex.
In chapter 3, I study the relation between grammatical gender and culture
by drawing on sociolinguistic and feminist non-linguistic approaches.
Sociolinguistic studies on language and gender have shown that grammatical
gender is part of a wide range of linguistic resources available to speakers
for denoting and constructing gender in communication (Hellinger and
Bussmann 2001). Linguistic items marked by grammatical gender codify
biological difference between men and women, ascribe sex to referents
(Eckert and McConnell-Ginet 2003; Ochs 1992), and are used by speakers for
the construction of gender identities (Borba and Ostermann 2007; Hall and
O’Donovan 1996; Kulick 1998; Livia 1997). In addition, gender is constructed
introduction 9
on the basis of social hierarchy, as the generic use of the masculine shows
(Hellinger and Busssmann 2001; 2002; 2003). The relation between language
and gender is also explored by feminist non-linguistic approaches (e.g. Butler
1999; Irigaray 1985a; Wittig 1992). These approaches highlight the role of
language, and of grammatical gender in particular, in categorizing subjects as
women or men and constructing the dominant gender order.
In chapter 4, I explore the relation between grammatical gender and cogni-
tion through the lens of cognitive semantics and research on linguistic rela-
tivity. Lakoff (1987) challenges the idea that grammatical gender constitutes
an arbitrary category with no conceptual dimension, by arguing that gender
relates to conceptual categorization. More specifically, he makes the specula-
tion that genders in Dyirbal, an Australian Aboriginal language, are conceptual
categories with prototypical and less prototypical members. Following Lakoff’s
approach, I suggest that grammatical gender in Greek constitutes conceptual
categorization and that the masculine and feminine gender correspond to
conceptual categories in which men and women are prototypical members
respectively.
Moreover, Lakoff argues that conceptual categories are linked with met-
onymic cognitive models. As Köpcke and Zubin (2003) show, grammatical
gender in German may be related with metonymic cognitive models that
incorporate sociocultural stereotypes. Similar to their speculation, I suggest
that the generic use of the masculine gender is linked to a metonymic cog-
nitive model that incorporates the sociocultural stereotype of man with the
category of human/universal.
The grammaticization of the concept of sex in Greek is taken to have sig-
nificant consequences for speakers’ cognition according to linguistic relativity.
More specifically, as Slobin (1996; 2003) argues, language mediates speakers’
thinking before and while speaking. Language sets limits on what speakers
must say, depending on what is grammaticized, and, thus, guides speakers to
attend to specific aspects of experience when they speak. Because gender is a
grammatical category used in an obligatory and systematic way, it is expected
to guide speakers to attend to the sex distinction when they speak, and inter-
pret referents as female or male.
In chapter 5, I formulate the research question and explore the relation
between grammatical gender and person reference in interaction. In addition,
I present the method to be employed for the empirical investigation of my
research question. Based on the different approaches examined in chapters 3
and 4, grammatical gender is shown to have a complex nature. The cultural
and cognitive aspects of grammatical gender in person reference interweave in
interaction. More specifically, the role of grammatical gender in constructing
10 chapter 1
the hierarchical gender dichotomy presupposes the role of grammatical gen-

der in guiding speakers to the interpretation of referent as female or male. The
empirical investigation of the grammatical gender-culture-cognition inter-
play in interaction is especially interesting for two reasons. As Conversation
Analysis has shown, interaction is the locus of human sociality where partici-
pants jointly construct their sociocultural world (Schegloff 2006b). Moreover,
when speakers communicate with each other, they are engaged in verbal
conduct and, thus, in thinking for speaking (Slobin 1996). Therefore, interac-
tion can serve as a sort of ‘lab’ for exploring both culture and cognition with
respect to grammatical gender. In particular, my aim is to examine empirically
i) whether interaction provides indications of the role of grammatical gender
in guiding speakers to interpret referents as female or male, and ii) the implica-
tions of the use of grammatical gender for the sociocultural world that speak-
ers construct in interaction.
Person reference is an important notion in the study of grammatical gender
in interaction, because it is the actual ‘target’ for any cultural or cognitive
‘work’ that grammatical gender may be doing in interaction. As a number of
studies on person reference in interaction (Brown 2007; Enfield 2007; Hanks
2007; Haviland 2007) show, person reference involves a lot more than just
identifying specific referents. Referential forms usually express information
about social relations between speaker, addressee and third person, and their
social identities. Therefore, they maintain and reinforce these social relations
and identities. In a similar way, referential forms in Greek can be said to be
functionally complex, because they are inflected for grammatical gender and,
thus, mark referent’s sex in a compulsory manner. By drawing on Silverstein’s
(1976) approach to indexicality, Ochs (1992) argues that linguistic items
marking sex lexically or grammatically index referent’s gender. Following
Ochs, I suggest that linguistic items marked by grammatical gender in Greek
index referent’s gender and, thus, incorporate the information of gender in the
covert assumptions that participants share about context and in the routine
meanings produced in interaction. The link between grammatical gender
and covert assumptions and routine meanings foregrounds the notions of
‘ordinariness’, ‘taken-for-granted’ and ‘commonsense knowledge’ (Garfinkel
1967; Sacks 1984) as key in the study of grammatical gender in interaction.
In chapter 6, I examine empirically grammatical gender in interaction, by
analyzing naturally occurring conversations among friends and relatives. More
specifically, I examine the cultural and cognitive aspects of grammatical gen-
der in the following cases:
introduction 11
– in individual (εγώ [eɣó] ‘I’), collective (εμείς [emís] ‘we’) and indefinite or
generic second person (εσύ [esí] ‘you’) self-reference (i.e. reference to
speaker), in reference to recipient (εσύ/εσείς [esís] ‘you.sg/you.pl’), and in
third person reference (αυτός/αυτή [aftós/aftí] ‘he/she’, αυτοί/αυτές [aftí/
aftés] ‘they.masc/they.fem’);
– in cases in which participants orient to sociocultural gender, that is, they
recognize gender as a relevant category for interaction;
– in cases of non-match between grammatical gender and referents’ sex, that
is, in cases where the masculine grammatical gender is used for reference to
female persons only;
– and, in cases in which the use of grammatical gender creates problems in
the interpretation of referent’s sex that are resolved in interaction through
practices of repair.
Overall, the empirical analysis yields indications of the cognitive role of gram-
matical gender i) in cases in which the codification of referent’s gender via
grammatical gender becomes important for what participants do in interac-
tion, and ii) in cases in which the information of referent’s gender is simply
made available in interaction. In the first instance, indications are found in
cases of self-repair and exposed and embedded correction, that is, in repairs
initiated by the speaker and a person other than the speaker respectively, in
which the item marked by grammatical gender constitutes the repairable item.
These repairs are classified as direct indications of the cognitive role of gram-
matical gender, because they involve interventions made by speakers in the
interpretation of referent’s sex. In the second instance, indications are found
in next turns, in turns’ recipient-design, in the non-match between grammati-
cal gender and referents’ sex, and in the membership categorization device.
These indications are indirect or covert, because in the latter cases the infor-
mation of referent’s gender codified by grammatical gender passes unnoticed
by participants.
In addition, the analysis shows that specific meanings are produced in
interaction in routine and covert ways because of the compulsory use of gram-
matical gender in the composition of turns. Linguistic items marked by gram-
matical gender are gender membership categories that categorize speakers,
recipients and third persons as women or men, and invoke and sustain the
stereotypical association of man with the norm. When participants orient to
gender, they employ the information of gender that is made available by gram-
matical gender in interaction to construct their own or other people’s gender
identities. When gender is not interactionally relevant, the gendered catego-
rization of referents and the maintenance of gender hierarchy occur together
12 chapter 1
with various social actions, which are not related to participants’ orientation
to gender.
I end the book with a series of conclusions and reflections in chapter 7. The
main points to be discussed in this final chapter are the following: i) interpret-
ing indications of the cognitive role of grammatical gender as indications of
the role of language in mediating speakers’ thinking for speaking, ii) reflecting
on the relation of the generic use of the masculine with the covert reproduc-
tion of sexism in interaction, iii) and reconsidering the concept of gender per-
formativity through the lens of the present study.
CHAPTER 2
Grammatical Gender
2.1 Introduction
This chapter examines the grammatical gender system in Standard Modern

Greek, namely its properties and overall organization. Before describing
grammatical gender in Greek, some basic information on the language
profile would be helpful for the reader (see Holton, Mackridge, and Philipaki-
Warburton 1997, Joseph and Philipaki-Warburton 1987, and Mackridge 1985 for
detailed descriptions of the language, and Tsangalidis 2011 for an overview of the
language profile). Modern Greek descended from Ancient Greek and it belongs
to the Indo-European group of languages. It is spoken by about 13 million
speakers, with approximately 10 million of them living in Greece, and the rest
in Cyprus and parts of the Greek diaspora, such as Australia and the us. Greek
is a fusional, highly inflecting language, which means that several grammatical
categories are marked morphologically. For instance, nouns inflect for gender,
number and case, and verbs inflect for person, number, tense, aspect, voice,
and mood. Diglossia (Ferguson [1959] 1964) has been a key issue throughout
the history of the Greek language. Two varieties have developed since the
Hellenistic times, a written ‘high’ more prestigious variety, which remained
close to classical Greek, and a spoken vernacular ‘low’ variety. After the founding
of the Greek nation-state, the distinction between the ‘high’ and ‘low’ variety
became known as the Katharevousa vs. Dhimotiki (‘demotic’) distinction.
Katharevousa, the ‘puristic’ high variety of the language, was used in official
settings, and was usually favored by ‘conservative’ Greeks, while Dhimotiki, the
‘popular’ low variety of the language, was used by ordinary people for everyday
life. Diglossia ceased to exist in 1976, when Dhimotiki became the official
language. Standard Modern Greek refers to the present state of Dhimotiki that
still keeps certain more formal and archaic characteristics found in Katharevousa
(see Christidis 1999, 32–43, Horrocks 2010, 438–470, and Mackridge 1985, 11–14
for relevant discussions on the diglossic situation in Greece).
2.2 Noun Classes or Genders
Grammatical gender is a device of noun categorization, which forms part

of a broad continuum involving classifiers of various types. According to

14 CHAPTeR 2
Aikhenvald (2000, 1), classifiers refer to “grammatical means for the linguistic
categorization of nouns and nominals” and include a wide range of noun
categorization devices, such as noun classes or genders, noun classifiers,
numeral, possessed, relational, verbal or deictic classifiers.1 This section
examines noun classes or genders. Noun classes or genders are grammaticalized
agreement systems, which correlate with certain semantic characteristics,
such as sex, animacy or humanness, and they are realized outside the noun
itself (Aikhenvald 2000, 19). The number of noun classes across languages
varies. For example, Babungo (Grassfields Bantu, Benue-Congo) has fourteen
noun classes, while Spanish and French have only two. In many Indo-European
languages, including Greek, there are two or three noun classes, which always
include the feminine and the masculine. These small systems are often
referred as ‘genders’ or ‘grammatical genders’ (Hellinger and Bussmann 2001,
7). For instance, in Spanish, nouns are divided into two genders, masculine and
feminine (Nissen 2002), as shown in example 1, while in German, nouns are
divided into three genders, masculine and feminine, plus neuter (Bussmann
and Hellinger 2003), as shown in example 2.
(1) Masculine Feminine

le libro la casa
the.masc book.masc the.fem house.fem
(2) Masculine Feminine Neuter

der Platz die Entscheidung das Haus
the.masc place.masc the.fem decision.fem the.neuter house.neuter
Although the terms ‘noun class’ and ‘gender’ are used interchangeably in
literature,2 Aikhenvald (2000, 19) suggests that the term ‘noun class’ should
be used as the cover term for noun class and gender, and the term ‘gender’
should be reserved for small systems of two to three distinctions, which always
include the masculine and the feminine. This definition will be followed here.
1 For example, South and American Indian languages and inflecting Indic languages have
special morphemes, which correlate with animacy or shape, and are attached to numerals
(Aikhenvald 2000, 105). These are numeral classifiers. Also, Terêna (South Arawak) has special
morphemes attached to verbs, which characterize the S/O argument in terms of its shape,
size, form and animacy (Aikhenvald 2000, 152). These are verbal classifiers. See Aikhenvald
(2000) for a systematic approach to classifiers from a cross-linguistic perspective.
2 For example, Corbett (1991) uses the term gender as a cover term for all noun class systems.
GRAMMATICAL GENDER 15
Noun class systems have two definitional properties, according to

Aikhenvald’s typological approach (2000, 20–22):
a) Some constituent outside the noun itself, such as adjectives, numbers,

demonstratives, articles or adverbs, in the noun phrase and/or the clause,
must agree in noun class with a noun. Noun classes are defined syntacti-
cally and are realized with affixes or clitics.
b) Nouns are assigned to specific classes3 according to semantic, and some-
times according to morphological and phonological principles.
Noun class assignment has always a semantic basis, which usually relates to
animacy, humanness, and sex, and sometimes also shape and size (Aikhenvald
2006). In certain languages, noun class assignment is based exclusively on
semantic principles. For example, in Dravidian languages such as Kolami,
Parji and Ollari nouns denoting male humans belong to one class and nouns
denoting female humans and other entities belong to a different class (Corbett
1991, 10). The Australian language Diyari distinguishes nouns denoting female
human/animates from nouns denoting the rest, that is, male human/animates
and all inanimates (Austin 1981, 60 cited in Corbett 1991, 11).
In Dyirbal, a language spoken in North Queensland, Australia, nouns are
categorized in four classes: i) bayi for male referents and non-human animates,
ii) balan for female referents, fire, water and battle, iii) balam for non-protein
food, and iv) bala for the rest (Dixon 1972, 306–312). There are deviations from
these noun class assignment rules based on principles of mythological or
conceptual association or on the important-property principle. For example,
most birds are classified as balan rather than bayi, because birds are associated
mythologically with the spirits of dead women, and the moon is classified as
bayi, because in myths it represents the husband of the sun, which is classified
as balan. In addition, the fishing spear belongs to the bayi class rather than
the bala class, because it is linked conceptually with fish, which belongs to
the bayi class. The light and the stars belong to the balan class rather than the
bala class, because they are linked conceptually with fire, which belongs to
the balan class. While most fishes are classified as bayi, harmful fishes are clas-
sified as balan, because they display the important property of harmfulness,
which is linked conceptually with fire and battle (balan).
3 Each noun is assigned to one specific class. However, occasionally, a noun may belong
to more than one class. These nouns are often called nouns of ‘common gender’ (Corbett
1991, 181).
16 CHAPTeR 2
Manambu, a language from the Ndu family, spoken in the Sepik area of
New Guinea, distinguishes nouns into two classes, masculine and feminine
(Aikhenvald 2012). In general, noun class assignment is based on the size and
shape of referents. For example, adult humans are categorized as masculine
or feminine according to their sex, while children and babies are categorized
as masculine, if they are relatively big or old, or as feminine, if they are rela-
tively small or young. In addition, long, large or both inanimate objects are
categorized as masculine, while small and/or round objects are categorized as
feminine.
In many Indo-European languages such as Danish, Icelandic, Spanish,
Polish, German, Italian, Serbian/Croatian/Bosnian, Portuguese, Swedish,
Greek (see section 2.3), but also in Arab and Hebrew, gender assignment is
semantically motivated in person reference, because of the match between
grammatical gender and referent’s sex. As a number of studies included in the
three volumes edited by Hellinger and Bussmann (2001; 2002; 2003) show, in
languages with grammatical gender nouns denoting female humans are usu-
ally grammatically feminine, while nouns denoting male humans are usually
grammatically masculine. In examples 3 and 4, Spanish and German nouns
codify the sex of the person denoted both lexically and grammatically:

el hombre la mujer
‘man’ ‘woman’

der Mann die Frau
‘man’ ‘woman’
Contrary to person reference, gender assignment in inanimate reference is

generally considered to be semantically arbitrary. For example, there seems to
be no semantic principle for the categorization of sun as masculine in Greek
and French, feminine in German and Lithuanian, and neuter in Russian and
ancient Iranian (Hellinger and Bussmann 2001, 3). However, semantic motiva-
tion is also found in gender assignment in nouns denoting inanimate objects,
as Zubin and Köpcke (1981; 1984; 1986) and Köpcke, Panther and Zubin (2010)
show for German. For example, German nouns denoting physical entities (e.g.
das Phon ‘phon’, das Herz ‘heart’), superordinate categories of foodstuffs (e.g. das
Gemüse ‘vegetables’, das Obst ‘fruits’) and objects having relevance to human
needs (e.g. das Ding ‘thing’, das Gerät ‘tool’) are neuter. Nouns denoting fruit are
mostly feminine (e.g. Orange, Melone, Birne ‘pear’), wines are masculine (e.g.
Riesling, Elbling, Bordeaux) but soft drinks are feminine (e.g. Limonade, Cola,
Fanta), automobiles are masculine (e.g. Mercedes, Honda, Rover) but motor-
cycles are feminine (Harley Davidson, Kawasaki, Suzuki). Semantic motivation
is also found in the case of nouns ending in -mut. Masculine nouns ending in
-mut (e.g. Mut ‘courage’, Hochmut ‘arrogance’, Unmut ‘displeasure, resentment’)
are associated with the semantic characteristic of extroversion, that is, they
denote types of conduct directed towards controlling the outside world. In
contrast, feminine nouns ending in -mut (Wehmut ‘nostalgia’, Schwermut ‘mel-
ancholy’, Demut ‘humility’) are associated with the semantic characteristic of
introversion, that is, they denote types of conduct that place the self under
outside control.
Besides semantic principles, morphological and phonological principles
may also play an important role in gender assignment. In Russian, nouns
denoting humans are divided into feminine (e.g. mat ‘mother’) and masculine
(e.g. otec ‘father’) according to the semantic distinction of sex (Corbett 1991,
34). In all other cases, nouns are divided into masculine, feminine and neuter
according to morphological principles: nouns of declension 1 are masculine
(e.g. zakon ‘law’), nouns of declensions 2 and 3 are feminine (e.g. škola ‘school’,
kost΄ ‘bone’), and nouns of declension 4 are neuter (e.g. vino ‘wine’) (Corbett
1991, 36). In other languages, gender assignment in nouns denoting inanimate
referents may follow phonological principles. For example, in Katcha, a
Kordofanian language, nouns beginning with m- are feminine, unless they
have a male referent (Heine 1982, 200 cited in Aikhenvald 2000, 25).
Mixed principles of gender assignment can also be found in some
languages. For example, in Yimas, a language from the Sepik area in Papua
New Guinea, noun classes 1–4 are semantically motivated, while classes 5–11
are phonologically motivated (Foley 1986, 86ff.; 1991 cited in Aikhenvald 2000,
26). More specifically, nouns denoting male humans belong to class I, nouns
denoting female humans belong to class ii, nouns denoting animals belong to
class iii, and nouns denoting culturally important plants belong to class iv.
The assignment of the remaining classes is based on phonological principles.
Agreement is a definitional property of noun classes, as I mentioned above.
Noun class agreement can occur in noun phrases between modifiers and heads,
and/or in a clause between a predicate and its arguments. In noun phrases the
head determines noun class agreement with adjectives, participles, articles
or demonstratives. As the following example (found in Marcato and Thüne
2002, 194) illustrates, in Italian the adjective shows gender agreement with
the noun:
18 CHAPTeR 2
(5) la ragazza alta

the.fem.sg girl.fem.sg tall.fem.sg
‘the tall girl’
In clauses there is usually noun class agreement between the verb and the core
arguments, that is, subject and/or direct object. This is the case, for example, in
some prefixing Australian languages (Dixon 2002, 479–485 cited in Aikhenvald
2000, 34). Noun class agreement is also found with adverbs, though rarely (for
example in Lak, a North-east Caucasian language, Aikhenvald 2000, 34).
The assignment of agreement features is based on semantic or mixed seman-
tic and syntactic principles (Aikhenvald 2000, 29 and 37–39). For instance, in
German noun class agreement is syntactic, that is, based on the grammatical
gender of the noun, as shown in example 6 (found in Bussmann and Hellinger
2003, 146).
(6) der Platz an der Sonne

the.masc place.masc in the.dat.fem sun.dat.fem
‘the place in the sun’
However, in nouns denoting human beings semantic agreement may also

occur (Bussmann and Hellinger 2003, 146–147). For example, the noun das
Mädchen (‘the girl’) is grammatically neuter but is lexically specified as female.
For anaphora German speakers may select the neuter personal pronoun (es) or
the feminine pronoun (sie), that is, both syntactic and semantic agreement is
possible (more on grammatical gender and anaphora in section 2.4).
Noun classes can also be overt or covert, depending on whether they can be
inferred from the form of nouns or not. Overt noun class marking involves vari-
ous morphological processes, such as external affixation or apophony, and may
depend on case and number, among other factors (Aikhenvald 2000, 57–62).
For example, in Russian gender is marked via the different inflections and is
overt in nominative case.
In addition, noun classes tend to interact with other grammatical catego-
ries, such as number and person (Aikhenvald 2000, 242–270). For example,
in certain languages such as German, Russian and Manambu, gender is dis-
tinguished only in the singular and not in the plural. In other languages, the
number of genders differs for singular and plural. For example, in Polish three
genders are distinguished in singular, and two genders in plural. Gender dis-
tinctions are usually found in third person pronouns, for example, he, she, it in
English, and er, sie, es in German. However, in some languages gender is also
specified for second person (e.g. in most Afroasiatic).
2.3 Grammatical Gender in Modern Greek
In Modern Greek, the grammatical gender system is divided into three

inflectional paradigms or declensions, which correspond to masculine,
feminine and neuter (Triantafyllidis [1941] 2005), exemplified in 7.

o ουρανός η θάλασσα το βουνό
o uranόs i θálasa to vunό
‘the sky’ ‘the sea’ ‘the mountain’
All nouns, adjectives, articles and passive participles, and certain pronouns
and numerals are inflected for gender. Gender is inflected in singular and plu-
ral number and in all cases, that is, nominative, genitive, accusative and voca-
tive (see Tables 1, 2 and 3).
In Greek, nouns are assigned to specific genders according to semantic and
morphological principles. More specifically, in person reference gender assign-
ment is semantically motivated because of the match between grammatical
gender and referent’s sex. In general, nouns denoting male human beings are
grammatically masculine and nouns denoting female human beings are gram-
matically feminine. This semantic basis is illustrated with examples 8 and 9.
table 1 Declension I-Masculine
Singular Plural
nom o ουρανός οι ουρανοί

o uranόs i uraní
gen του ουρανού των ουρανών
tu uranú ton uranόn
acc τον ουρανό τους ουρανούς
ton uranό tus uranús
voc oυρανέ ουρανοί
urané uraní
20 CHAPTeR 2
table 2 Declension II-Feminine
Singular Plural
nom η θάλασσα οι θάλασσες

i θálasa i θálases
gen της θάλασσας των θαλασσών
tis θálasas ton θalasόn
acc τη θάλασσα τις θάλασσες
ti θálasa tis θálases
voc θάλασσα θάλασσες
θálasa θálases
table 3 Declension iii-Neuter
Singular Plural
nom το βουνό τα βουνά

to vunό ta vuná
gen του βουνού των βουνών
tu vunú ton vunόn
acc το βουνό τα βουνά
to vunό ta vuná
voc βουνό βουνά
vunό vuná
(8) o φοιτητής
o fititís
the.nom.masc student.nom.masc
‘the male student’
(9) η φοιτήτρια
i fitítria
the.nom.fem student.nom.fem
‘the female student’
In Greek, the match between grammatical gender and referent’s sex in person
reference is almost perfect, as was shown by a study conducted by Pavlidou,
Alvanoudi and Karafoti (2004) on the nouns included in the Dictionary of
Standard Modern Greek (1998). All nouns were codified according to their
grammatical gender and semantic content, i.e. whether they denote person,
animal, plant, abstract inanimate object, etc., and the sex of referent in per-
son reference, among other parameters. This study showed that the total num-
ber of nouns denoting person in Modern Greek included in the Dictionary
of Standard Modern Greek (1998) is 6225. 59.2% of these nouns are mascu-
line, 33.5% are feminine and 6.5% are neuter. 94.2% of the masculine nouns
denotes male humans, and 94.1% of the feminine nouns denotes female
humans (Pavlidou, Alvanoudi, and Karafoti 2004, 546). Thus, the correlation
between grammatical gender and referent’s sex in person reference is almost
perfect.
In addition, this study showed that 8 feminine nouns denote male humans,
half of which refer to male homosexuals with derogatory meaning, e.g.
λουμπίνα [lubína], πούστρα [pústra] and αδελφή [aðelfí], and only 3 masculine
nouns denote female humans, all carrying sexual connotations, e.g. πειρασμός
[pirazmόs] ‘a woman who causes sexual desire’ and κόμματος [kόmatos] ‘very
beautiful woman’. In addition, 49 neuter nouns denote male humans, e.g. το
αγόρι [to aɣόri] ‘boy’, and 30 neuter nouns denote female humans, e.g. το κορίτσι
[to korít͡si] ‘girl’. The assignment of neuter gender in the latter two nouns is due
to morphological principles, as we will see later on in this section.
In Greek, referent’s sex is marked not only grammatically but also lexically.
Certain nouns carry the semantic property of maleness/femaleness, that is, they
have lexical gender. For example, the nouns άντρας [ándras] ‘man’, θείος [θíos]
‘uncle’ and πατέρας [patéras] ‘father’ encode the semantic property of maleness
as part of their lexical meaning, while the nouns γυναίκα [ʝinéka] ‘woman’,
θεία [θía] ‘aunt’ and μητέρα [mitéra] ‘mother’ carry the semantic property of
femaleness. In these nouns there is correspondence between the masculine
and feminine grammatical gender and the lexical marking of the noun as male
or female. The analysis presented in this book focuses on linguistic items that
are grammatically marked as female or male. However, linguistic items that
are lexically marked as female or male will also be examined in a few points
throughout the text, when their use in interaction is considered to be relevant
with the analysis.
Sex-based gender assignment is also found in nouns denoting animals, e.g.
γάτος [ɣátos] ‘cat.masc’, γάτα [ɣáta] ‘cat.fem’. However, for many animals there
is a single form for both sexes, e.g. αλεπού [alepú] ‘fox.fem’, δελφίνι [ðelfíni]
22 CHAPTeR 2
‘dolphin.neuter’, or two forms of different grammatical genders which refer

to both males and females, e.g. τζίτζικας [d͡zíd͡zikas] ‘cicada.masc’, τζιτζίκι
[d͡zid͡zíci] ‘cicada.neuter’.
Gender assignment of inanimate referents is taken to be semantically arbi-
trary. For example, nouns denoting abstract entities may be masculine, e.g.
συλλογισμός [siloʝizmόs] ‘reasoning.masc’, feminine, e.g. ελευθερία [elefθería]
‘freedom.fem’, or neuter, e.g. πρόβλημα [prόvlima] ‘problem.neuter’. In a sim-
ilar way, nouns denoting physical phenomena may be masculine, e.g. κεραυνός
[ceravnόs] ‘thunder.masc’, feminine, e.g. βροχή [vroçí] ‘rain.fem’ or neuter,
e.g. δέντρο [ðéndro] ‘tree.neuter’. However, some sort of semantic motiva-
tion can also be found in gender assignment of inanimate referents. Nouns
denoting abstract entities, such as property or action, are often grammati-
cally feminine, e.g. αλλαγή [alaʝí] ‘change.fem’, ομορφιά [omorfçá] ‘beauty.fem’
(Anastasiadi-Symeonidi and Chila-Markopoulou 2003, 27). Nouns denoting
countries, islands, and cities are often feminine, e.g. Ελλάδα [eláða] ‘Greece.
fem’, Κρήτη [kríti] ‘Crete.fem’, Θεσσαλονίκη [θesaloníci] ‘Thessaloniki.fem’,
while nouns denoting rivers are often masculine, e.g. Έβρος [évros] ‘Evros.
masc’ (Anastasiadi-Symeonidi and Chila-Markopoulou 2003, 27). Also, nouns
denoting fruit-trees are usually feminine, e.g. μηλιά [miʎá] ‘apple tree.fem’,
while fruits are usually neuter, e.g. μήλο [mílo] ‘apple.neuter’ (Mackridge
1985, 49).
The semantic principles mentioned above interact with morphological
principles. In general, grammatical gender in Greek is associated with bound
morphemes, that is, specific suffixes in nouns. Overt gender marking is shown
in nominative case in singular number. More specifically, masculine nouns
end in -ας [as], e.g. τουρίστας [turìstas] ‘tourist’, in -ης [is], e.g. ναύτης [náftis]
‘sailor’, in -ες [es], e.g. καφές [kafés] ‘coffee’, in -ος [os], e.g. δρόμος [ðrόmos] ‘road’,
and in -ους [us] παππούς [papús] ‘grandfather’. Feminine nouns end in -α [a],
e.g. ελπίδα [elpíða] ‘hope’, in -η [i], e.g. ζάχαρη [záxari] ‘sugar’, in -ου [u], e.g.
αλεπού [alepú] ‘fox’, in -ος [os], e.g. ήπειρος [ípiros] ‘continent’, and in -ω [o],
e.g. ηχώ [ixό] ‘echo’. Finally, neuter nouns end in -ο [o], e.g. σίδερο [síðero] ‘iron’,
in -ι [i], e.g. κορίτσι [korít͡si] ‘girl’, in -ος [os], e.g. μέγεθος [méʝeθos] ‘size’, in -μα
[ma], e.g. όνομα [όnoma] ‘name’), in -ς, e.g. κρέας [kréas] ‘meat’, in -ν [n], e.g.
ενδιαφέρον [enðiaféron] ‘interest’, and in consonants other than -ν or -ς, e.g. ήπαρ
[ípar] ‘liver’ (see Triantafyllidis [1941] 2005, 225–253, and Holton, Mackridge
and Philippaki-Warburton 1997, 48–72 for a detailed presentation of the three
inflectional paradigms).
Covert gender marking is also found in Greek. For example, there is a num-
ber of nouns denoting persons, usually occupational terms, which have the
same form for both male and female referents. These are described as nouns of
‘common gender’ (Holton, Mackridge, and Philippaki-Warburton 1997, 61) and
their gender is distinguished by the article or other modifiers, as examples 10
and 11 show.
(10) o δικηγόρος η δικηγόρος

o ðiciɣόros i ðiciɣόros
the.masc lawyer.masc the.fem lawyer.fem
‘the male lawyer’ ‘the female lawyer’
(11) o συγγενής η συγγενής

o siɟenís i siɟenís
the.masc relative.masc the.fem relative.fem
‘the male relative’ ‘the female relative’
The degree of ‘overtness’ of gender depends on case and number. For example,
in singular number in nominative case certain masculine, feminine and neuter
nouns may all end in -os, as shown in example 12.

ουρανός πρόοδος λάθος
uranόs prόoðos láθos
‘sky.nom.sg’ ‘progress.nom.sg’ ‘mistake.nom.sg’
Also, in singular number in accusative case certain masculine, feminine and

neuter nouns may all end in -a (example 13).
(13) Μasculine Feminine Neuter

πατέρα καρδιά κύμα
patéra karðʝá címa
‘father.acc.sg’ ‘heart.acc.sg’ ‘wave.acc.sg’
Finally, in singular number in genitive case certain masculine, feminine and

neuter nouns may all end in -u (example 14).

δρόμου εισόδου σίδερου
ðrómu isóðu síðeru
‘road.gen.sg’ ‘entrance.gen.sg’ ‘iron.gen.sg’
24 CHAPTeR 2
In plural number in nominative case there are masculine and feminine nouns
which both end in -es or in -i (examples 15 and 16).

κανόνες χώρες
kanónes xóres
‘rule.nom.pl’ ‘country.nom.pl’

δρόμοι είσοδοι
ðrόmi ísoði
‘road.nom.pl’ ‘entry.nom.pl’
Also, in plural number in genitive case there are masculine, feminine and neu-
ter nouns which all end in -on (example 17).

δρόμων εφημερίδων φώτων
ðrómon efimeríðon fóton
‘road.gen.pl’ ‘newspaper.gen.pl’ ‘light.gen.pl’
The gender of the noun controls agreement with other elements in the noun
phrase or the predicate. In the noun phrases in examples 18 and 19, the gender
of the head noun controls agreement with the demonstrative, the article and
the adjective:
(18) αυτός ο καλός άντρας

aftόs o kalόs ándras
this.nom.masc the.nom.masc good.nom.masc man.nom.masc
‘this good man’
(19) αυτή η καλή γυναίκα

aftí i kalí ʝinéka
this.nom.fem the.nom.fem good.nom.fem woman.nom.fem
‘this good woman’
In examples 20 and 21, agreement is realized outside the noun phrase, with the
gender of the subject noun phrase controlling agreement with the adjective
and the participle in the predicate:
(20) O κήπος είναι όμορφος.

o cípos íne όmorfos
the.nom.masc garden.nom.masc is beautiful.nom.masc
‘The garden is beautiful.’
(21) H φοιτήτρια είναι πολύ θυμωμένη.

i fitítria íne polí θimoméni
the.nom.fem student.nom.fem is very angry.part.nom.fem
‘The female student is very angry.’
Gender agreement is based on mixed semantic and syntactic principles (see

Chila-Markopoulou 2003 for a systematic analysis of gender agreement in
Greek). Examples 18–21 illustrate cases of syntactic agreement, that is, agree-
ment based on the grammatical gender of the head noun (in examples 18, 19
and 21 agreement is also semantic). Semantic agreement may occur in nouns
with human referents in which there is a mismatch between grammatical gen-
der and natural gender (e.g. κορίτσι ‘girl.neuter’). In this case, syntactic agree-
ment is obligatory in the noun phrase (example 22).
(22) To όμορφο κορίτσι

to όmorfo korít͡si
the.nom.neuter beautiful.nom.neuter girl.nom.neuter
‘The beautiful girl’
However, in clauses (example 23 found in Chila-Markopoulou 2003, 149) or in

anaphora (example 24) semantic agreement may also be possible.
(23) Aυτό το κορίτσι, με

aftό to korít͡si me
this.nom.neuter the.nom.neuter girl.nom.neuter with
τόσες ικανότητες και
tóses ikanótites ce
so.many.acc.fem skills.acc.fem and
διακρίσεις, δεν έχει εμπιστοσύνη στον
ðiakrísis ðen éçi ebistosíni ston
excellence.acc.fem.pl neg has trust.acc.fem in.the.acc.masc
εαυτό του/της.
eaftó tu/tis
self.acc.masc its/her
‘In spite of all those skills and excellence, this girl does not trust itself/herself.’
26 CHAPTeR 2
(24) Tο κορίτσι ήρθε. Το/την

to korít͡si írθe to/tin
the.nom.neuter girl.nom.neuter come.3sg.past it.acc/she.acc
είδα στον κήπο.
íða ston cípo
see.1sg.past in.the.acc.masc garden.acc.masc
‘The girl came. I saw it/her in the garden.’
If a noun phrase consists of two nouns, which denote inanimate objects of the
same kind and are assigned to the same grammatical gender, then syntactic or
semantic agreement is possible in the predicate (example 25, found in Chila-
Markopoulou 2003, 153).
(25) Βρήκα την πόρτα και την

vríka tin pórta ce tin
find.1sg.past the.acc.fem door.acc.fem and the.acc.fem
ντουλάπα σπασμένες/ σπασμένα.
dulápa spazménes spazména
closet.acc.fem broken.part.acc.pl.fem/ neuter
‘I found the door and the closet broken.’
Gender distinctions are found in third person pronouns, e.g. the demonstra-
tives αυτός [aftós] ‘this.masc’, αυτή [aftí] ‘this.fem’, αυτό [aftó] ‘this.neuter’
and εκείνος [ecínos] ‘that.masc’, εκείνη [ecíni] ‘that.fem’, εκείνο [ecíno] ‘that.
neuter’.
Finally, it is worth mentioning that in some nouns grammatical gender is
associated with differences at the semantic and pragmatic level (Anastasiadi-
Symeonidi and Chila-Markopoulou 2003, 16–19). For example, some words
found in two genders may differ in meaning, e.g. μάντρα [mándra] ‘fence.fem’/
μαντρί [mandrí] ‘corral.neuter’, show stylistic variation, e.g αέρας [aéras]/
αγέρι [aʝéri] ‘wind.masc/wind.neuter’, or sociolinguistic variation, e.g. φόνος
[fónos]/φονικό [fonikó] ‘murder.masc/murder.neuter’. Semantic differences
can also be found in some nouns which are assigned to different genders in the
singular and the plural, e.g. δεσμός [ðezmós] ‘bond.masc.sg’, δεσμοί [ðezmí]
‘bonding.masc.pl’/δεσμά [ðezmá] ‘constraint.neuter.pl’.
2.4 Grammatical Gender and Reference: Preliminary Remarks
The role of grammatical gender in person reference is examined in detail

in chapter 5. However, some preliminary remarks need to be made in this
section with respect to the relation between grammatical gender and refer-
ence in general. Grammatical gender is considered to be one of the means
employed by languages for solving reference problems. Such problems may
involve a) achieving generic reference, b) solving coordination problems, and
c) maintaining reference.
Generic or indefinite reference is reference to female plus male referents
or to referents whose sex is unknown. Languages use different means to deal
with this problem. For example, Zande, a language of the Adamawa-Ubangian
branch of languages spoken in Zaire, Sudan and the Central African Republic,
has a special pronoun, ni, which is distinct from other personal pronouns and
is used if speakers refer to a non-specific or unknown person (Claudi 1985,
95–96 cited in Corbett 1991, 223). In Dyirbal there is no specific preference, as
either bayi or balam may be used for reference to female plus male referents
(Dixon 2010a, 238). In numerous Indo-European languages, including Greek,
the masculine grammatical gender is used for generic reference (Hellinger
and Bussmann 2001; 2002; 2003), while in Arawak (Aikhenvald 1999 cited in
Aikhenvald 2000, 54) and in a few Australian languages (Alpher 1987 cited in
Aikhenvald 2000, 54) the feminine gender is the preferred form for generic ref-
erence. The following example illustrates the generic use of the masculine in
Greek. The noun phrase οι καθηγητές that is marked by the masculine gender is
used for reference to all professors, either female or male.
(26) Οι καθηγητές του

i kaθiʝités tu
the.nom.masc.pl professors.nom.masc the.gen.neuter.sg
τμήματος έχουν συνάντηση.
tmímatos éxun sinádisi
department.gen.neuter have.3pl.pres meeting.acc.fem
‘The (male) professors of the department have a meeting.’
The use of the masculine or the feminine gender for generic reference is asso-
ciated with markedness relations in noun class systems. According to Dixon
(2010a, 237), the notion of functional markedness in grammar “relates to the
situation of use” of a specific term.4 The marked term is used in restricted
situations, while the unmarked term is employed in all other circumstances.5
As Aikhenvald (2000, 51) notes, a noun class is considered to be functionally
4 Markedness can also be formal (Dixon 2010a, 237). A term in a system with zero realization is
taken to be formally unmarked (e.g. singular number in English).
5 For example, singular number in English is functionally (and also formally) unmarked,
because contrary to plural number, which is used only for reference to two or more referents,
28 CHAPTeR 2
unmarked, if it is used as “a generic term and for indefinite reference”, among

other things.6 Functional markedness in noun class systems is also related with
the problem of coordination of nouns belonging to different noun classes or
what is referred as “noun class resolution” (Aikhenvald 2000, 52). Resolution
may follow semantic, syntactic principles or a combination of both. For exam-
ple, in the Dravidian language Tamil the problem of noun class resolution is
solved semantically (Asher 1985 cited in Corbett 1991, 269). In Tamil, nouns
are categorized into rational (humans, gods and demons) and non-rational or
neuter. If a noun phrase consists of nouns denoting female and male rationals,
the rational agreement form is used, as the following example shows (found in
Corbett 1991, 269):
(27) akkaa-vum aŋŋan-um neetu va-nt-aaŋka

elder.sister-and elder.brother-and yesterday come-past-3rd.pl.rational
‘Elder sister and elder brother came yesterday.’
In contrast, in Portuguese the problem of noun class resolution is solved syn-

tactically. If a noun phrase consists of two nouns belonging to different gen-
ders, the conflict is resolved in favor of the functionally unmarked gender, that
is, in favor of the masculine (example 28, found in Aikhenvald 2000, 53):
(28) um menino e uma menina bonitos

indef.masc boy and indef.fem girl beautiful.pl.masc
‘beautiful boy and girl’
Noun class resolution in Latin is semantic in nouns denoting persons, and syn-
tactic in all other nouns (Aikhenvald 2000, 53). Mixed syntactic and seman-
tic resolution is also found in Greek. If a noun phrase consists of two nouns
belonging to different genders and denoting male and female human beings,
the conflict is resolved in favor of the masculine, as the following example
shows.
singular is used for reference to one referent or “in a general sense when no number specifi-
cation is made” (Dixon 2010a, 237).
6 According to Aikhenvald (2000, 51), the other two criteria for defining a noun class as func-
tionally unmarked are the following: the noun class must be used when the noun class
distinction is neutralized or is of no relevance, and it may be used in default or neutral
agreement.
(29) Aυτός ο φοιτητής και

aftόs o fititís ce
this.nom.masc the.nom.masc student.nom.masc and
εκείνη η φοιτήτρια είναι
ecíni i fitítria íne
that.nom.fem the.nom.fem student.nom.fem are
πολύ χαρούμενοι.
polí xarúmeni
very happy.part.nom.masc.pl
‘This male student and that female student are very happy.’
Semantic resolution is found in nouns that belong to different genders but

denote human beings of the same sex (example 30).
(30) Αυτό το κορίτσι και

aftó to korít͡si ce
this.nom.neuter the.nom.neuter girl.nom.neuter and
η Mαρία είναι νικήτριες
i maría íne nicítries
the.nom.fem Maria.nom.fem are winners.nom.fem
στον διαγωνισμό.
ston ðiaɣonizmó
in.the.acc.masc competition.acc.masc
‘This girl and Maria are the winners in the competition.’
For nouns that belong to different genders and denote inanimate objects both
syntactic and semantic resolution is possible (example 31).
(31) Ο υπολογιστής και η

o ipoloʝistís ce i
the.nom.masc computer.nom.masc and the.nom.fem
καρέκλα είναι χαλασμένοι/ χαλασμένα.
karékla íne xalazméni xalazména
chair.nom.fem are broken.part.nom.pl.masc/ neuter
‘The computer and the chair are broken.’
However, if one of the nouns is neuter, semantic resolution is preferred, as

shown in example 32 (see Chila-Markopoulou 2003 for more examples in
Greek).
30 CHAPTeR 2
(32) O αναπτήρας και το

o anaptíras ce to
the.nom.masc lighter.nom.masc and the.nom.neuter
τσιγάρο είναι πεταμένα
t͡siɣáro íne petaména
cigarette.nom.neuter are lying.part.nom.neuter.pl
στο πάτωμα.
sto pátoma
in.the.acc.neuter floor.acc.neuter
‘The lighter and the cigarette are lying on the floor.’
Although the terms ‘functionally unmarked’ or ‘syntactic’ resolution may be

useful in describing gender systems, they limit our perspective on the variety
of the parameters involved in the use of gender. These parameters are social
and will be examined in detail in the next chapter.
The third problem that is resolved through the use of grammatical gender is
maintaining reference to the same entity. This phenomenon is known in con-
temporary linguistics as anaphora. Anaphora is defined by Huang (1994, 1) as a
relation between two linguistic elements, the anaphor and the antecedent, in
which the interpretation of the former is determined by the interpretation of
the latter. Gender systems are one of the various “reference-tracking systems”
provided by different languages “to keep track of the various entities referred
to in an ongoing discourse” (Huang 1994, 8).7 In this sense, noun phrases of
the same gender can be considered to be co-referential, while noun phrases of
different gender cannot. For example, in the following sentence in Greek the
feminine pronoun της is co-referential with the feminine noun phrase η Μαρία,
that is, the feminine pronoun (anaphor) selects as its referent the same person
depicted initially by the feminine noun (antecedent).
(33) H Μαρία πήγε στο

i maría píʝe sto
the.nom.fem Maria.nom.fem go.3sg.past to.the.acc.neuter
σπίτι της.
spíti tis
house.acc.neuter pron.poss.gen.fem
‘Maria went to her house.’
7 Reference-tracking systems also include switch-reference systems, switch-function systems,

and inference systems (Huang 1994).
2.5 Summary
Based on the discussion so far, we can conclude that the semantic distinction
of female/male sex is grammaticized in Greek. More specifically, the Greek
gender system is shown to have a semantic core in nouns denoting persons,
since the masculine and the feminine grammatical gender correlate with
referent’s male and female sex respectively. Grammatical gender constrains
speakers in specific ways. When speakers use their language, that is, when they
speak or write, they must denote systematically their own sex as well as other
referents’ sex. Given that grammatical gender is a feature inherent to the lan-
guage system that marks a vast number of linguistic items in Greek (cf. section
2.3), the codification of referents’ sex is systematic and compulsory. Speakers
simply do not have another choice; they are enforced by language system to
codify the semantic distinction of sex.
However trivial this may sound, it deserves our attention as linguists for
two reasons. First of all, it is not self-evident that all languages grammaticize
referent’s sex. As I mentioned earlier in this chapter, sex is just one of the various
semantic distinctions (e.g. shape, size) according to which nouns are catego-
rized in different languages. For example, in some Dravidian languages nouns
are divided into human and non-human, while in Siouan and Algonquian
languages nouns are divided into animate and inanimate (Aikhenvald 2000,
276). Moreover, there are languages with no gender systems, such as Finnish
or Turkish. Thus, the grammaticization of sex is an important and special fea-
ture of Greek which foregrounds the issue of language diversity and makes
Greek—as well as other Indo-European languages with grammatical gender—
distinctive. Second, and more importantly, gender is a feature of grammar that
displays a semantic basis. Given that grammatical gender is a morphological
category related to meaning and with a compulsory and systematic use in com-
munication, it should interweave with culture and cognition in specific ways.
In the next two chapters, I explore how this is done, by employing various
approaches within linguistics and beyond.
CHAPTER 3
Grammatical Gender and the Intersection of

Sociocultural Gender/Sex
3.1 Introduction
The relation between grammatical gender and culture has been investigated
primarily by sociolinguistic approaches since the mid 1970s. These approaches
explored the different lexico-grammatical tools that languages provide for
the construction of sociocultural gender. Grammatical gender constitutes
one of these tools. By codifying the semantic distinction of female/male sex
(cf. chapter 2), grammatical gender divides persons according to biological/
anatomical differences, and locates them on a bipolar order in which male/
masculine and female/feminine constitute the two poles respectively.
Moreover, this gender bipolarity is structured on the basis of social hierarchy,
which is associated with men’s dominance and women’s subordination. One
of the basic linguistic practices for maintaining gender hierarchy is the generic
use of the masculine. Aspects of the relation between grammatical gender
and culture have also been addressed in the work of feminist theorists work-
ing within the framework of psychoanalysis and philosophy. These theorists
underlined, among other things, that grammatical gender plays a role in con-
structing gendered subjects and sustaining gender inequality.
In this chapter, I examine the cultural aspects of grammatical gender, by
drawing on both linguistic and non-linguistic approaches. The reason for
selecting such an interdisciplinary perspective lies in the complexities that the
relation between grammatical gender and sociocultural gender foregrounds.
This relation has two sides, a ‘linguistic’ one and a ‘social’ one. These sides
are explored by linguistic and non-linguistic approaches in complementary
ways. In general, feminist theory has deeply influenced and shaped the ways in
which linguists theorize gender and its relation to language. For example, early
feminist linguistic research (e.g. Lakoff 1975; Spender 1980) focused on how the
linguistic representation of women and men is associated with issues of power
and equality, by drawing mainly on the agenda set by feminist empiricism
and feminist standpoint theory1 (Harding 1991). These linguistic approaches
1 Feminist empiricism aims at eliminating gender bias and discrimination against women at
research, teaching, implementation, policy-making and dissemination of data and informa-

grammatical gender & intersection of sociocultural gender/sex 33
aimed at uncovering the ways in which language system ‘reflects’ woman as a

negative inferior category and maintains gender inequality. They also argued
that language constitutes a tool for women’s emancipation. Yet, in the last
two decades a ‘discourse’ turn has occurred in feminist linguistic research,
which means that attention has shifted from language as system to language
as use (Eckert and McConnell-Ginet 2003, 4). As Eckert and McConnell-Ginet
(2003, 4) point out, this ‘discourse’ turn coincides with the ‘performance’ turn
in feminist theory, which is mostly associated with Butler’s (1999) work on
gender performativity (to be examined in more detail in the next section).
Butler showed that gender constitutes a sort of ‘doing’ and discursive effect. In
line with Butler’s theory, contemporary linguistic research theorizes gender as
a fluid and changing category, rather than static and solid, produced by mate-
rial and symbolic practices, and aims to uncover the ways in which language
use contributes to the construction of gender.
Butler’s theory has been deeply influential for linguists mainly because
Butler engages with theories situated within the broader epistemological
strand of feminist postmodernism (Harding 1991). Feminist postmodernism is
informed by poststructuralist theories addressing the role of language in con-
structing meaning, subjectivity and power (Andermahr, Lovell, and Wolkowitz
2000). Feminist postmodernism created space for an interdisciplinary dialogue
between feminist theory and linguistics, that generated a new strand of stud-
ies in language and gender research with challenging and insightful outcomes
(see for example the volumes edited by Bergvall, Bing and Freed in 1996, and
Livia and Hall in 1997).
The interdisciplinary spaces created between linguistics and feminist non-
linguistic approaches can help us gain a better understanding of the com-
plex relation between grammatical gender and culture. For this reason, both
approaches will inform the study of the relation between grammatical gender
and culture to be presented in this chapter.
3.2 The Intersection of Gender/Sex in Feminist Theory
Gender was introduced as an object of knowledge by the women’s movement

at the end of the 1960s and since then it has been established institutionally
tion, by adhering to the principle of scientific objectivity (Braidotti 2003, 199–200). Feminist
standpoint theory criticizes scientific objectivity, uncovers the role of power in knowledge
production, and privileges women’s experience in the formation of new paradigms of knowl-
edge (Braidotti 2003, 200–202).
34 chapter 3
as a topic of scholarly research in the field of Feminist, Women’s and Gender

Studies.2 Defining gender and conceptualizing the different ways in which
gender intersects with other sociocultural categories, such as sexuality, class or
ethnicity, has been central to feminist theory. Feminism, according to Braidotti
(1991, 147), is a critical theoretical movement through which women “situate
themselves in reality so to redesign their ‘feminine’ condition”.
One of the basic arguments employed by patriarchy for legitimizing and
naturalizing women’s oppression is biological determinism. According to this
view, gender inequality is causally related to biological differences between
women and men, and women’s socially inferior position is explained on the
basis of their body and brain features. This causal and deterministic relation
between biological sex and sociocultural gender has been deconstructed and
critically addressed by feminist theorists, starting from Simone de Beauvoir
([1949] 1972). In her classical book Second Sex, Beauvoir theorizes the relation
between women and men through the prism of dialectics and defines man
as the Subject and woman as the deviant subordinate Other. Beauvoir attacks
biological determinist arguments by distinguishing biological sex, that is, ana-
tomical differences among men and women, from sociocultural gender, that is,
social constructions such as behaviors, attitudes and norms, taken to be associ-
ated with biological differences.
In the years following Beauvoir’s work, feminists employed the analytic dis-
tinction between sex and gender in order to dissociate nature from culture
and challenge the tacit assumptions on which patriarchy is based. However,
as Lykke (2010, 25) points out, since the beginning of the 1990s the dichotomy
between sex and gender has undergone a process of deconstruction by a num-
ber of theorists, such as Haraway (1991), Butler (1999; 1993) and Braidotti (1994;
2002b). These theorists conceptualize gender/sex as a mixed cultural/natural
phenomenon, including both sociocultural aspects and bodily materiality.
They claim that gender should not be understood as a static homogeneous cat-
egory (WOMAN/MAN) or an attribute that people ‘have’, but rather as a fluid
heterogeneous category (WOMEN/MEN) and a process that is linked intersec-
tionally to other sociocultural categories and is open to transformation and
re-signification.
More specifically, Butler (1999) deconstructs the biological sex/socio-
cultural gender binary and argues that this is an artificial dichotomy. This
2 Women’s and Gender Studies is defined by Lykke (2004) as a post-disciplinary discipline

which employs interdisciplinarity to examine critically the intersections of gender, sexual-
ity, class, ethnicity and age, sociocultural hierarchies and power relations of inclusion and
exclusion.
dichotomy is based on the belief of a pre-social biological essence that pre-

exists and determines social relations. Butler describes this belief as the ‘meta-
physics of substance’. This sort of metaphysics legitimizes the economy of
compulsory heterosexuality. This economy consists of a set of regulatory prac-
tices that connect biological sex with sociocultural gender, sexual practice and
desire, and produce a matrix of intelligible, proper and non-intelligible, inap-
propriate or ‘abject’3 subjects (Butler 1999). By building on poststructuralist
philosophy of language, speech act theory and deconstruction, Butler (1999)
argues that gender does not exist before it is done, that is, before it is con-
structed through a set of reiterated performative practices. These practices are
both symbolic and material. For example, they include practices of ‘naming’
as well as norms on dressing, behaving and desiring. The reiteration of these
practices produces the materiality of the bodies as an effect and, thus, natural-
izes the dominant gender order. In Butler’s words,
Gender is the repeated stylization of the body, a set of repeated acts

within a highly rigid regulatory frame that congeal over time to produce
the appearance of substance, of a natural sort of being. (Butler 1999, 33)
In this sense, gender is performative, because it constitutes the very identity it

is purported to be.4 As Butler argues:
there is no gender identity behind the expressions of gender; that iden-

tity is performatively constituted by the very “expressions” that are said to
be its results (Butler 1999, 33)
Butler uses the concept of ‘queer’ in order to re-signify gender in non-

essentialist ways. She conceptualizes queer as a subject position, which sub-
verts the fixed static dichotomies of woman/man, feminine/masculine, hetero/
homo, and challenges the causal relation between body, sex, gender identity
and desire. The concept of ‘queer’ foregrounds the gender/sex intersection,
and according to Lykke (2010, 60), it should be understood as an image of an
alternative subjectivity or a feminist figuration, in Braidotti’s (2002b) terms.
3 Abject subjects are described by Butler (1993, 3) as the “ ‘unlivable’ and ‘uninhabitable’ zones
of social life which are nevertheless densely populated by those who do not enjoy the status
of the subject”.
4 Butler criticizes essentialist approaches to gender by drawing on Nietzche’s position that
“there is no ‘being’ behind doing, effecting, becoming; ‘the doer’ is merely a fiction added to
the deed-the deed is everything” (Butler 1999, 33).
36 chapter 3
Figurations are defined by Braidotti (2002b, 2–3) as alternative forms of

becoming, “materialistic mappings of situated, or embedded and embodied,
positions”, transformative accounts of the self, which “draw a cartographic
map of power-relations and thus can also help identify possible sites and strat-
egies of resistance”.5 In her own work, Braidotti (1994; 2002b) builds another
feminist figuration, the nomadic feminist. By drawing on Irigaray’s sexual dif-
ference theory and Deleuze’s theory on the subject’s embodied becoming,
Braidotti conceptualizes the difference of the female body in a positive and
affirmative way, and explores the role of sexual difference in the process of
becoming subjects. In a similar way, Haraway (1991) approaches the gender/
sex intersection by using the figuration of the ‘cyborg’, which “blurs categori-
cal distinctions (human/machine; nature/culture; male/female; oedipal/non-
oedipal)” (Braidotti 2003, 209). All these feminist figurations approach the
gender/sex intersection in non-essentialist ways and show that the subject
is an embodied and cultural entity or, a culture-nature network, according to
Lykke and Braidotti (1996).
In section 3.4, I return to feminist theorists in order to examine how their
work contributes to the study of the relation between grammatical gender
and the gender/sex intersection. Before that, I will first examine how the rela-
tion between grammatical gender and culture has been investigated within
linguistics.
3.3 Cultural Aspects of Grammatical Gender: The Perspective

of Linguistics
Speakers have access to a variety of resources, both material and symbolic, for
constructing gender identities. The basic symbolic resource available to speak-
ers is language system. As Ochs (1992) shows, language relates to gender in two
broad ways. On the one hand, there is a straightforward mapping of language
to gender (one-to-one correlation), in the sense that certain linguistic items
index gender referentially, directly and exclusively (more on this topic in chap-
ter 5). These are linguistic items in which gender constitutes “the actual con-
tent of a linguistic sign”, in Eckert and McConnell-Ginet’s words (2003, 60). A
5 Figurations should not be understood as metaphors but rather as lived maps. As Braidotti
(2002b, 3) argues, “being nomadic, homeless, an exile, a refugee, a Bosnian rape-in-war vic-
tim, an itinerant migrant, an illegal immigrant, is no metaphor. [. . .] These are highly specific
geo-political and historical locations—history tattooed on your body. One may be empow-
ered or beautified by it, but most people are not; some just die of it”.
small number of such items exist in every language community. For example,
in English the third person pronouns he and she, and the address terms Mr,
Mrs and Madam, are grammatically and lexically specified for sex respectively
(Ochs 1992, 338–339). In a similar way, linguistic items marked by grammatical
gender in Greek codify morphologically referent’s sex, and, thus, index gender
referentially, directly and exclusively.
On the other hand, language usually relates to gender in a non-straight-
forward manner, that is, non-referentially, indirectly and non-exclusively. For
example, the use of tag questions in English may index speaker’s feminine
gender, because tag questions are taken to express hesitancy or tentativeness,
which are stereotypically associated with feminine behavior (McElhinny 2003,
35, see also Eckert and McConnell-Ginet 2003, 167–173 for a discussion of stud-
ies on tag questions and gender). In this case, indexing gender is non-exclusive,
because women but also men use tag questions. In addition, indexing gender
via tag questions is constitutive and indirect or mediated, because tag ques-
tions correlate with other social information stereotypically associated with
women. This social information is based on speakers’ knowledge about norms
on gender roles and attributes. Although this book focuses on direct referen-
tial aspects of indexing gender, it is important to keep in mind that direct and
indirect indexing of gender coexist in language use. As Eckert and McConnell-
Ginet (2003) show, speakers may employ a large number of linguistic items or
structures from phonology, morphology, lexicon or syntax to construct gender.
Some of these items and structures may index gender directly, such as items
marked by grammatical gender, and some may index gender indirectly.
Given their relation to indexing gender, linguistic items marked by grammat-
ical gender play an important role in the construction of gender. As Hellinger
and Bussmann (2001, 19) point out, the basic function of grammatical gender
in person reference is “the communication of gendered messages of various
types”. As I show in the next sections, these messages involve the construction
of gender identities and the maintenance of gender hierarchy.
3.3.1 Grammatical Gender and the Construction of Gender Identities

The use of grammatical gender shapes the ways in which speakers design their
utterances. Speakers’ linguistic choices are highly constrained because of gram-
matical gender. Whenever they speak, they must codify female or male sex in a
compulsory way in a vast number of linguistic items, given gender agreement
(cf. chapter 2). The grammatical marking of sex has a specific social dimen-
sion attached to it. Linguistic items marked by grammatical gender “ascribe”
female or male sex to referent (Eckert and McConnell-Ginet 2003, 65), and,
thus, categorize speakers, hearers and third persons as women or men. In this
38 chapter 3
way, grammatical gender contributes to a “gendering” process, in McConnell-

Ginet’s (2003, 90) words, which involves the presumptions that interlocutors
carry about their own and other people’s gender. For example, in referring to a
third person via a grammatically feminine form, the speaker assumes that this
person is female/woman.
The use of grammatical gender contributes to the construction of gender
identities.6 As a number of studies show for Hindi (Hall and O’Donovan 1996),
French (Livia 1997; 2001) and Brazilian Portugese (Borba and Ostermann 2007;
Külick 1998), speakers employ grammatical gender to construct identities
that deviate from the norm of compulsory heterosexuality. One such case is
found in the community of hijras in India, who are born as boys but raised
as girls that dance at weddings and ceremonies (Hall and O’Donovan 1996).
The hijras have created a ‘third gender’ identity, which transgresses the female/
male bipolarity. During the process of their ‘becoming-hijra’ they select the
feminine gender instead of the masculine when they refer to themselves.
This linguistic practice is combined with other non-linguistic practices such
as dressing, facial gestures and body movement. After establishing a hijra-
identity, the hijras may switch between the feminine and the masculine gender
when they refer to others depending on the context, their relation to inter-
locutors and the social role performed. More specifically, the hijras employ
the feminine gender to construct relations of solidarity and social proximity
with other hijras, while they employ the masculine gender to construct rela-
tions of power and distance. For instance, they refer to younger hijras via the
feminine gender, but they employ the masculine gender to refer to hijras of
socially superior or inferior position, older age, in order to insult other hijras or
differentiate from them. In addition, they use the masculine gender to refer to
themselves at the period of their childhood before becoming hijras, to express
anger or put emphasis on what they say.
Similar to hijras, French homosexuals and transsexuals employ grammati-
cal gender to transgress static bipolar categories. Livia (1997; 2001, 160–192)
analyzes the use of grammatical gender in five autobiographies and two
biographies of transsexuals who changed their sex from male to female. The
autobiographies examined include Georgine Noël’s Appelez-moi Gina, Brigitte
Martel’s Né homme, comment je suis devenue femme, Inge Stephens’s Alain,
transexuelle, Sylviane Dullak’s Diane par Diane and Je serais elle, and the biog-
raphies include Catherine Rihoit’s Histoire de Jeanne transsexuelle and Marie
Mayrand’s Le Combat de la mére d’un transsexuel. The authors of these books
6 Identities should be understood here as discursive constructs emerging in interaction

through indexical processes, among others (Bucholtz and Hall 2005).
switch between feminine and masculine gender in self-reference to express

various social meanings. For example, the Belgian transsexual Georgine Noël
uses the feminine gender in the diary she keeps as a teenager to refer to herself
and, thus, differentiate from the masculine gender that is socially imposed on
her. When her parents discover her diary, she switches to masculine gender in
order to express her anger and disappointment. She also uses the masculine
gender when she refers to the events that took place before the sex reassign-
ment surgery in order to highlight her body difference before and after the
surgery.
French male homosexuals usually refer to themselves or their friends and
lovers via the feminine gender, and in this way they build an identity outside
the norm of compulsory heterosexuality (Pastre 1997 cited in Livia 2001, 186).
However, female homosexuals use the masculine gender less often to refer to
their lovers. As Livia (2001) points out, this may be related to the fact that the
masculine gender is functionally unmarked, and, thus, can be associated with
the category of human and not exclusively male. In contrast, the feminine gen-
der is “usually restricted to a plus interpretation and, if applied to a masculine
referent, implies femininity or a homosexual orientation” (Livia 2001, 189). In
addition, Livia (1997) shows that the use of grammatical gender in French is
related to social meanings linked with gender identities. In the film La Cage
aux folles by Marcello Danon a homosexual couple who own a cabaret for
transvestites employ the feminine gender to refer to other homosexuals. In this
way, they express their solidarity with these homosexuals or their sympathy for
them. However, when the gay couple wants to differentiate from other homo-
sexuals, they employ the masculine gender to refer to them.
Finally, the use of grammatical gender seems to be a common practice
among transvestites of north and south Brazil for establishing and negotiating
gender identities (Borba and Ostermann 2007; Kulick 1998). Transvestites in
North Brazil refer to others via the grammatically masculine word travesti
or the grammatically feminine word bicha (Kulick 1998, 214–218). The word
travesti is the preferred form in formal occasions, while the word bicha is the
preferred form in informal occasions. In addition, transvestites in south Brazil
usually refer to themselves or other transvestites via grammatically feminine
linguistic items (Borba and Ostermann 2007). Yet, they switch to masculine
gender in a number of occasions: a) in narratives about events that took place
when they changed their sex, b) when they report the speech produced by
third persons on transvestites, c) when they describe themselves in relation to
their families, and d) in order to differentiate from other transvestites.
The above studies show that speakers may use grammatical gender to con-
struct gender identities that deviate from the norm, and express various social
40 chapter 3
meanings that may be related to these identities, such as in-group solidarity,

sympathy, power or antagonism. These different sorts of subjectivities come
into social being through language. Although grammatical gender may act “as
a trap”, in Livia’s words (2001, 192), and limit people to a static bipolar gen-
der order, it can also be used as a device for expressing “gender fluidity” (Livia
2001, 192).
3.3.2 Grammatical Gender and the Maintenance of Gender Hierarchy

The dominant gender order is organized on the basis of social hierarchy
between men and women. According to this hierarchy, men constitute the
norm and women constitute the deviant and subordinate ‘other’. Structures of
discrimination against women are mainly sustained in society via institution-
alised practices in all kinds of social spheres (Connell 2002). These spheres are
both private and public, and include family, education, labor market, and state,
among others. Language is an additional medium for maintaining asymmet-
ric gender relations. As Lakoff (1975) points out, women experience inequality
through language in the way in which they are represented through language.
Two main patterns are observed across various languages with respect to lin-
guistic sexism: the semantic derogation of women and the generic use of the
masculine (cf. section 2.4). This section focuses on the second pattern, because
it explicitly involves the way in which speakers use grammatical gender.
However, a few things need to be mentioned about the semantic derogation of
women, before moving to the generic use of the masculine.
Reference to women across various languages is done in negative ways,
because feminine nouns tend to carry negative connotations. Schulz (1975)
describes this pattern as semantic derogation. For example, in English male-
specific and female-specific nouns are subjected into different social evalu-
ations by speakers, as the following pairs show: wizard/witch and bachelor/
spinster. The male-specific nouns wizard and bachelor carry positive con-
notations, while the female-specific nouns witch and spinster carry nega-
tive connotations. Similarly, in Finnish the feminine diminutive naikko-nen
‘woman-diminutive, trollop’ has negative content, while the masculine dimin-
utive miekko-nen ‘man-diminutive, little man’ does not (Engelberg 2002, 126).
In Spanish, the noun phrase mujer pública, which consists of a feminine head
noun and a feminine modifier, carries the meaning of ‘whore’, while the noun
phrase hombre public, which consists of a masculine head noun and a mascu-
line modifier, denotes a man ‘in the public eye’ (Nissen 2002, 260–261).
Greek displays a similar pattern of semantic derogation, according to
Pavlidou ([2002] 2006a, 44–45). For instance, contrary to the Greek masculine
diminutive noun αντράκι [andráci] ‘small man’, which has positive meaning,
the feminine diminutive noun γυναικάκι [ʝinekáci] ‘small woman’ has negative
meaning, i.e. denoting a woman of no importance.7 Moreover, feminine occu-
pational terms with explicit feminine demotic endings, such as -ινα [ina] and
-ισα [isa], e.g. βουλευτίνα [vuleftína] ‘female parliament member’, have under-
gone a process of semantic derogation (Pavlidou 1985). As Pavlidou (1985, see
also Pavlidou 2003 for an overview of the study in English) shows, the use of
these forms is not preferred in formal contexts. In Greek, women in occupations
are denoted by nouns of common gender stemming from katharevousa, e.g.
o/η φιλόλογος [o/i filόloɣοs] ‘philologist’, ο/η δημοσιογράφος [o/i ðimosioɣráfos]
‘journalist’ (cf. section 2.3), and by masculine nouns with the feminine article,
e.g. η ταμίας [i tamías] ‘cashier’, η βουλευτής [i vuleftís] ‘member of the parlia-
ment’. Triantafyllidis ([1953] 1963 cited in Pavlidou 2003, 186) suggested that
feminine occupational terms could be formed via explicit demotic feminine
suffixes. However, this solution was not adopted broadly. The semantic dero-
gation of these occupational terms seems to be related both to their explicit
feminine marking and the tension between the demotic and katharevousa
varieties (Pavlidou 1985).
The semantic derogation of women also involves the representation of
women as sexual objects or prostitutes. For example, Stanley (1977) shows that
in English the stereotypical woman is a prostitute. More specifically, 220 English
words denote women who are sexually available, e.g. slut, honey pot, but only
22 words denote men in a similar way. In addition, many English words origi-
nally denoting female humans, such as Nymph and Nymphet, acquired sexual
connotations and ended up denoting sexually available women or prostitutes
(Schulz 1975, 85). A similar pattern is also found in Finnish (Engelberg 2002).
In Spanish, many feminine words, such as mujer ‘woman’ and chica ‘girl’, carry
negative sexual connotations (Nissen 2002, 263). Yet, few masculine nouns
carry similar connotations. For instance, the masculine noun tío ‘uncle’ is usu-
ally employed by women to refer to a man as a sexual object. With respect
to Greek, Pavlidou, Alvanoudi and Karafoti (2004) found that 34 of the nouns
included in the Dictionary of Standard Modern Greek (1998) denote unethical
women, e.g. πατσαβούρα [pat͡savúra] ‘old bag’, πορνίδιo [porníðio] ‘little tart’, and
7 In addition, as Pavlidou (2006a, 45) points out, in Greek there are many nouns composed by
a lexical morpheme with negative meaning and a lexical morpheme encoding female sex,
such as βρομοθήλυκο [vromoθíliko] and βρομογύναικο [vromoʝíneko], both meaning ‘dirty slut’.
However, this pattern is not found in composite nouns including a lexical morpheme encod-
ing male sex. An exception to this are composite nouns which consist of a lexical morpheme
encoding both referent’s male sex and age, such as βρωμόγερος [vromόʝeros] ‘dirty old man’.
42 chapter 3
only 15 denote unethical men, e.g. σάτυρος [sátiros] ‘satyr, lustful man’, καριόλης
[karʝόlis] ‘bastard’.
Women can also be represented negatively in less direct ways when speakers
refer to the inanimate world. For example, as Mathiot (1979) shows, American
English speakers often refer to typhoons and broken cars via the feminine pro-
noun she, because emotional instability and weakness are stereotypically asso-
ciated with women. However, they usually refer to cars working properly via
the masculine pronoun he, because power is stereotypically associated with
men. Moreover, male speakers often personify inanimate objects as female,
as the following utterance shows: “She gets me where I want to go” (a trucker
refers to his van). According to Mathiot (1979), this use reflects men’s view of
women as objects. A similar pattern exists in the use of the masculine and
the feminine pronoun in Tasmanian English (Pawley 2002, see also Wierzbicka
2002 for a relevant discussion). Speakers tend to refer to the animate and inan-
imate world via the masculine and the feminine pronoun respectively. The
feminine gender is usually employed for reference to vehicles, e.g. “But when
’e first bought the bulldozer ’e told me he took ’er out in the bush”, storms, e.g.
“She made a mess of that crop”, houses, e.g. She’s certainly got a view, this one”
or abstract concepts as profession, e.g. “I’ve given up the morning job I have
’er away”. These patterns reflect the patriarchal worldview, which identifies
the human with the male and objectifies women (Wierzbicka 2002).
The other medium for marking female difference as a negative category
and establishing men as the norm is the generic use of the masculine. No
matter how ‘innocent’ this practice may look on the surface, it is actually a
highly regulatory practice, which reflects gender hierarchy at the social level
and is based on specific assumptions about how the sociocultural world is
organized (Hellinger and Bussmann 2001, 14–15). These assumptions involve
what Hellinger (2001, 108) describes as “the ideology of MAN (male as norm)”.
According to this ideology, the male/masculine is considered to be the higher
and more prestigious category, while the female/feminine is considered to be
secondary and subordinate.
Several researchers exploring the relation between gender and language
point out that the generic use of the masculine is a political choice, which
establishes a man-made perspective of the world. Following up the discussion
in section 2.4, more examples are given with respect to this rather common
pattern across Indo-European languages. In Italian and German, masculine
nouns may refer to both men and women, e.g. cittadini ‘male and female
citizens’ (Marcato and Thüne 2002, 201), and Jeder Wähler sollte von seinem
Stimmrecht Gebrauch machen ‘Every (masculine) voter (masculine) should
exercise his right to vote’ (Bussmann and Hellinger 2003, 158). In addition, in
Icelandic many masculine nouns denoting occupations can refer to either men
or women, e.g. leikari ‘actor’ and prestur ‘priest’ (Grönberg 2002, 173). This is
also found in French, e.g. Elle est l’auteur de deux romans ‘She is the author
(masculine) of two novels’ (Schafroth 2003, 100–101), and in Polish, e.g. Ona
jest dobrym lekarzem ‘This is a good doctor (masculine)’ (Koniuszaniec and
Blaszkowska 2003, 268). The generic use of items grammatically or lexically
specified as male has also been reported for English8 (e.g. Eakins and Eakins
1978; Graham 1975; Μartyna 1983; Miller and Swift 1988; Nilsen 1977). For exam-
ple, Graham (1975) found that masculine pronouns are used more frequently
than feminine pronouns in English textbooks, and that most of these mascu-
line pronouns refer to male humans.
The generic use of the masculine is also found in Greek, as I mentioned in
section 2.4. A few additional points need to be made here. As Pavlidou (2006a)
points out, the feminine gender may sound unusual or incorrect in certain
usages, because of the generic use of the masculine. This is shown in the fol-
lowing example (found in Pavlidou 2006a, 42).
(34) Μπορεί κανείς να καλέσει πρώτα

borí kanís na kalési próta
can.3sg.pres anyone.nom.masc subj call.3sg.pfv first
το νούμερο
to número
the.acc.neuter number.acc.neuter
‘One (male) can call the number first’
*Μπορεί καμία να καλέσει πρώτα

borí kamía na kalési próta
can.3sg.pres anyone.nom.fem subj call.3sg.pfv first
το νούμερο
to número
the.acc.neuter number.acc.neuter
‘One (female) can call the number first’
8 In English, third person pronouns he, she and it are grammatically marked as male, female
and inanimate. However, given that no gender agreement is observed between head nouns
and modifiers, English cannot be said to possess a noun class system (Aikhenvald 2000, 21).
See Curzan (2003) for the history of gender in English.
44 chapter 3
Moreover, a link between the use of the masculine noun άνθρωπος [ánθropos]
‘human being’ and false generics is reported in a study conducted by Makri-
Tsilipakou (1989). The study shows that Greek speakers do not use the noun
άνθρωπος to refer to female humans only.
Interestingly in Greek, male dominance is reflected not only in the way in
which language is used but also in the way in which the language system is
organized. The study of the nouns in the Dictionary of Standard Modern Greek
(Pavlidou, Alvanoudi, and Karafoti 2004) showed that overall—in reference to
human/animates and inanimates—the feminine gender is dominant morpho-
logically, since there are twice as many feminine nouns (45%) as masculine
nouns (23.8%). Yet, the masculine nouns denoting humans (3574) are twice
as many as the corresponding feminine (1949), that is, most nouns denoting
humans are grammatically masculine.
Overall, the generic use of the masculine is considered by feminist linguists
as an exclusionary practice, which makes women less visible, and guides speak-
ers to understand human as male or male as human (Engelberg 2002, 114).
Discourse seems to be ‘colonized’ by the masculine gender, forcing women to
undergo a process of negative ‘othering’ and ‘silencing’ (Spender 1980). In mak-
ing such claims, sociolinguistic studies seem to presuppose or implicate that
grammatical gender has some sort of cognitive depth and, therefore, ‘affects’
or guides the way in which speakers understand the world. This assumption
about the cognitive role of language also seems to underlie or motivate femi-
nist linguistic debates on language reform.
The non-sexist use of language has been treated as a practice for eliminat-
ing linguistic sexism, and, therefore, affecting or facilitating social change. In
Cameron’s words (1985, 1), language constitutes “an essential part of the strug-
gle for liberation”. Feminist initiatives on language reform, or what Pauwels
(2003) defined as ‘feminist linguistic activism’, target linguistic sexism and
explore alternative ways in which women will ‘speak’ about their different
locations and experience and female difference will be represented through
language in positive ways. Reform practices involve various sorts of interven-
tions in language use, which concern the semantic derogation of women and
the generic use of the masculine. Guidelines for non-sexist language use can be
found in Nilsen (1977), Miller and Swift (1988), Kramarae and Treichler (1985),
Frank and Treichler (1989), and Tsokalidou (1996) for the Greek language. With
respect to the generic use of the masculine, two reform strategies have been
proposed. These are summarized by Pavlidou (2006a, 51) as following: a) add-
ing the feminine gender whenever the masculine gender is used for generic
reference, and b) avoiding denoting male or female sex. These strategies are
illustrated with the following examples for Greek (found in Pavlidou 2006a, 51).
Instead of employing the masculine gender only (example 35), generic refer-
ence to Greek people can be achieved either by employing the feminine noun
together with the masculine noun (example 36) or by using a noun which does
not codify the male or female sex of referents (example 37).
(35) Oι Έλληνες λατρεύουν τη

i élines latrévun ti
the.nom.masc.pl Greeks.nom.masc love.3pl.pres the.acc.fem
θάλασσα.
θálasa
sea.acc.fem
‘(Male) Greeks love the sea.’
(36) Oι Έλληνες και οι

i élines ce i
the.nom.masc.pl Greeks.nom.masc and the.nom.fem.pl
Ελληνίδες λατρεύουν τη θάλασσα.
eliníðes latrévun ti θálasa
Greeks.nom.fem love.3pl.pres the.acc.fem sea.acc.fem
‘Male and female Greeks love the sea.’
(37) O ελληνικός λαός λατρεύει

o elinikόs laόs latrévi
the.nom.masc Greek.nom.masc population.nom.masc love.3sg.pres
τη θάλασσα.
ti θálasa
the.acc.fem sea.acc.fem
‘The Greek people love the sea.’
Although these strategies seem to be useful in theory, they prove to be inef-

fective for Greek when they are put into practice (Pavlidou 2006a). For exam-
ple, repeating the feminine form together with the masculine form is a time
consuming practice, which entails extra cognitive work that speakers need to
undertake when they speak. This work is generally not required when speakers
use the masculine gender only (more on this topic in chapter 5). This is shown
with examples 38 and 39.
46 chapter 3
(38) Όποιος/α εργάτης/τρια έχει

όpços/a erɣátis/tria éçi
whoever.nom.masc/fem worker.nom.masc/fem has
πρόβλημα με το
prόvlima me to
problem.acc.neuter with the.acc.neuter
ωράριό του/της πρέπει να . . .
oráriό tu/tis prépi na
schedule.acc.neuter his/her must subj
‘Every male and female worker who has a problem with his/her schedule
must . . . ’
(39) Oι μαθητές και

i maθités ce
the.nom.fem.pl students.nom.masc and
οι μαθήτριες αισθάνονται
i maθítries esθánode
the.nom.fem.pl students.nom.fem feel.3pl.pres
κουρασμένοι και κουρασμένες.
kurazméni ce kurazménes
tired.part.nom.masc.pl and tired.part.nom.fem.pl
‘The male and female students feel tired.’
As Pavlidou (2006a, 53) highlights, avoiding denoting the female/male sex dis-
tinction in person reference may be a successful strategy for German, which
has a plural participle that makes no distinction between the masculine and
the feminine, e.g. die Lehrenden ‘teachers’. Yet, this strategy is ineffective for
Greek in which most linguistic items inflect for gender (cf. chapter 2). For
intance, in Greek the corresponding participle for die Lehrenden is grammati-
cally marked as male or female in a compulsory way, e.g. οι διδάσκοντες και οι
διδάσκουσες [i ðiðáskodes ce i ðiðáskuses] ‘the male teachers and the female
teachers’ (Pavlidou 2006a, 53).
Besides linguistics, the relation between grammatical gender and culture
has been addressed in the context of feminist non-linguistic theories, more
specifically in the work of Irigaray (1985a; 1985b; 1985c; 1993), Wittig (1992)
and Butler (1993; 1997; 1999). These theories offer a systematic philosophical
and psychoanalytic approach on sociocultural gender and associate the use of
grammatical gender with the construction of gendered subjectivity (Alvanoudi
2008). In the following section, I examine how these theories contribute to our
understanding of the relation between grammatical gender and the gender/
sex intersection.
3.4 Cultural Aspects of Grammatical Gender: The Perspective of

Feminist Non-Linguistic Theories
The feminist theories examined in this section are situated within the broader
epistemological framework of feminist postmodernism9 (briefly discussed
in section 3.1), and more specifically within the framework of French post-
structuralism, which considers postmodernism as a critical mode of thinking.
In poststructuralist theory, language is taken to be a medium of constructing
meaning and social reality. Subjectivity is understood as a discursively pro-
duced process, which encompasses conscious and unconscious dimensions,
and is embodied (Braidotti 2003). Poststructuralist thinking criticizes the sub-
ject of the western philosophical tradition as a ‘disembodied’ abstract rational
individual and moves beyond in theorizing the self as a locus of contradictions,
diversity and change. The category WOMAN is challenged as a unified category
and emphasis is given on the differences within the category itself, taking class,
age, race, ethnicity and sexual orientation into account. These differences are
asserted in positive ways and are linked with newly formed empowered sub-
jects (Braidotti 2003).
In this section, I examine three feminist theorists working in the realm of
this epistemological paradigm. More specifically, I focus on the insights they
provide on the cultural aspects of grammatical gender, by taking the general
context of their work into account.
Luce Irigaray is considered to be the basic philosopher of sexual difference
theory. She argues that the patriarchal symbolic economy10 is based on the
9 Feminist postmodernism is an epistemological strand emerging from the critical

engagement of feminism with postmodernism. Postmodernism has been diagnosed
by Lyotard (1984) as a condition of the post-industrial society, characterized by the col-
lapse of the grand- or meta-narratives, and by a crisis of the premises of Enlightment
(e.g. humanism, marxism, teleologism). Feminist postmodernism challenges the static
and homogeneous categories WOMAN/MAN and criticizes the epistemological rational
subject of Enlightment’s thought as phallogocentric and masculine. Phallogocentrism is
a term related to phallocentrism, which is a term derived from Lacanian psychoanaly-
sis. Phallocentrism refers to “the patriarchal symbolic order in which the phallus is posi-
tioned as the primary signifier, privileging masculinity at the expense of femininity”
(Andermahr, Lovell, and Wolkowitz 2000, 163). Phallogocentrism refers to “any discourse
which attempts the phallic mastery of the world through the sign” (Andermahr, Lovell,
and Wolkowitz 2000, 163).
10 The concept of symbolic in Irigaray’s work derives from Lacanian psychoanalysis. The
Symbolic is one of the three orders distinguished by Lacan (the Imaginary and the Real
are the other two). The Symbolic order is “a product of language which as a system
imposes rules upon the human infant” (Andermahr, Lovell, and Wolkowitz 2000, 219).
48 chapter 3
production of subordinate others versus the male as the One and the Same
universal subject (Irigaray 1985a). This economy is reproduced via the domi-
nant (Lacanian) psychoanalytic paradigm, which tends to ignore women.
By establishing hierarchical dichotomies such as phallic/non-phallic, penis/
vagina-clitoris, more/less, psychoanalysis theorizes female sexual difference as
a pejorative other. For Irigaray, the issue of representing female difference and
subverting the idea of woman as a negative category is directly linked with
language. Language constructs sexed subjects by categorizing them as female
or male, and locates them in asymmetric positions: man is the universal sub-
ject and woman is the negative subordinate other or “what is left of a mirror
invested by the (masculine) ‘subject’ to reflect himself, to copy himself”, in
Irigaray’s (1985b, 30) words.
For Irigaray (1993) grammatical gender is one of the basic tools that sus-
tain the patriarchal symbolic economy. For example, Irigaray argues that the
generic use of the masculine and the negative evaluation of grammatically
feminine words in French affect the ways in which women experience them-
selves as negative others. For instance, she points to the semantic derogation
of the feminine noun doctoresse and the asymmetries in the semantic pairs un
ordinateur (‘computer’)/la machine à écrire (‘typewriter’) or un avion (‘plane’)/
une voiture (‘car’), in which masculine nouns denote objects of higher value
compared with the objects denoted by feminine nouns. Irigaray treats lan-
guage as a key notion for women’s emancipation. Women need to speak ‘as
women’ and use language in creative ways in order to represent their differ-
ence in positive ways, and, thus, make themselves visible.11
The role of language in the construction of gender is also central in Monique
Wittig’s (1992) work. Wittig criticizes biological determinism for naturalizing
and legitimizing social inequality between men and women, and deconstructs
the category of sex. According to this theorist, sex should not be understood
as a given and natural category, but rather as a political category. In her words,
“there is but sex that is oppressed and sex that oppresses” (Wittig 1992, 2). Sex
divides subjects according to biological difference. At the same time it cate-
gorizes subjects as legitimate or illegitimate, and locates them to hierarchal
11 Irigaray (1985b) suggests that women should employ the strategy of mimesis, that is, re-
appropriate old meanings and start producing new radical ones. She describes mimesis
as a woman’s effort “to recover the place of her exploitation by discourse, without allow-
ing herself to be simply reduced to it” (Irigaray 1985b, 76). Mimesis involves the use of
innovative practices for the expression of fluid pluralistic identities. For instance, Irigaray
(1985b) employs the metaphor the two lips in order to conceptualize a new feminine
imaginary.
positions in the symbolic order. For Wittig, the process of women becoming-
subjects constitutes a cognitive process of rethinking and re-evaluating the
social world. This process is linked with language.
By drawing on a Marxist framework of analysis, Wittig argues that language
belongs to the superstructure, which reproduces the dominant ideology. She
suggests that language shapes speakers’ understanding of the world, by enforc-
ing a specific view on what is ‘real’ and legitimate, as is shown in the following
extract:
Language casts sheaves of reality upon the social body, stamping it and
violently shaping it . . . for there is a plasticity of the real to language: lan-
guage has a plastic action upon the real. (Wittig 1992, 78)
Grammatical gender plays an important role in this process, because it denotes

female vs. male sex difference, and enforces the division of subjects according
to their sex. For this reason, Wittig (1992, 77) describes grammatical gender as a
“linguistic index of the political opposition between the sexes and of the domi-
nation of women”. She associates the generic use of the masculine with the
ideology of man as the universal subject, and she adopts the extreme position
that grammatical gender should be deleted from languages, because it divides
subjects on arbitrary ontological grounds. In her own work, Wittig attempts to
restore an undivided ‘I’ and universalize women as subjects, by employing the
impersonal pronoun on, which does not mark subjects as female or male, or by
using the feminine gender for generic reference.
Finally, language is a key notion in Butler’s work on gender performativity.
Butler draws on various theoretical traditions in order to conceptualize gen-
der performativity and the role of language—or discourse—in this process
(cf. section 3.2). One line of thinking that Butler follows in her work derives
from Foucault’s work on power and discourse, and more specifically from his
concept of bio-power. Bio-power refers to educational, medical or other proce-
dures designated to subjugate the population, control the individual body and
define what counts as a ‘proper’ body or an intelligible subject. By drawing on the
notion of bio-power, Butler (1993) theorizes the ways in which bodies are ‘mate-
rialized as sexed’ through discourse and are divided into intelligible and non-
intelligible, liveable and unliveable or abject. Discursively constructed bodies
are constituted by the very acts that name these bodies, that is, by language.
Butler draws on Austin’s (1962) speech act theory, Derrida’s ([1972] 1988)
deconstructive approach to Austin’s work, and Althusser’s (1984) notion of
interpellation in order to examine the role of language in gender performa-
tivity. Austin (1962) claims that certain utterances are performative (i.e. they
50 chapter 3
perform what is being said), and he associates their performative force with
certain felicity conditions, which are related to specific contexts and conven-
tions. For example, the utterance I now pronounce you husband and wife is
infelicitous, if it is not uttered by a person designated to do so in an appropri-
ate context, e.g. by a priest at a church. Following Austin, Butler suggests that
words do not simply describe subjects. Words categorize subjects and locate
them to specific social positions. In this way, language interpellates subjects
into social existence in the context of a ritual, material and pre-existing struc-
ture, in Althusser’s (1984) terms. For example, when a baby is born and the
doctor says It’s a girl! a process of ‘girling’ is inaugarated through which
the baby is categorized as female and located in the dominant gender order. The
use of a linguistic item lexically marked for female sex produces the illusion of
an essentialist identity, that is, the effect of female sex. I suggest that this func-
tion can also be extended to all linguistic items marked by grammatical gender.
These items categorize speakers and others according to their sex and sustain
the bipolar gender order. In this sense, the use of grammatical gender consti-
tutes a performative practice, which produces the materiality of sexed bodies.
It is worth mentioning that Butler does not accept a deterministic view on
the role of language in the construction of the social world. By building on
Derrida’s idea that linguistic signs are always subject to re-appropriation, reit-
eration and re-signification, she argues that words can be used in unforeseen
ways and acquire new meanings. This process of re-signification can enable
subjects to form new kinds of political agency.12
3.5 Summary
If we take both linguistic and non-linguistic approaches into account, the cul-
tural aspects of grammatical gender can be defined as following. Grammatical
gender ascribes sex to referents and locates them in asymmetrical positions in
the bipolar gender order. It naturalizes biological difference, which supports
social hierarchy and inequality between women and men. It constructs gender
12 A typical case of how re-signification can become a tool for political action is the way in
which the word queer has been re-appropriated and re-signified by the lgbt movement.
Originally, the word queer denoted the strange, unusual, sexually deviant, that is, it had
negative meanings. However, it acquired positive meaning, when it was re-appropriated
by lgbt activists to denote a critical stance towards the norm of compulsory heterosexu-
ality. Moreover, the concept queer has been used by Butler as a tool for conceptualizating
gender in non-essentialist ways (cf. section 3.2).
identities and reproduces sexism. The latter is mainly achieved through the
generic use of the masculine. In a few words, grammatical gender constructs
sociocultural gender on the basis of social hierarchy.
This social dimension of grammatical gender seems to be interconnected
with a cognitive dimension. The latter dimension is often implied or presup-
posed by sociolinguistic approaches, which associate the generic use of the
masculine with women’s exclusion from speakers’ thinking of the world. The
idea that grammatical gender plays an important role in cognition is also
reflected in numerous initiatives taken by feminist linguists and other theo-
rists with respect to language reform, which aim at increasing women’s vis-
ibility in the social/symbolic sphere. But is there actually a relation between
grammatical gender and speakers’ cognition? This question is examined in the
next chapter.
CHAPTER 4
Grammatical Gender and Speakers’ Cognition
4.1 Introduction
The question whether grammatical gender plays a role in speakers’ thinking

of the world is part of the more general question on the relation between lan-
guage and cognition. This question has received different answers within lin-
guistics and has been an object of investigation across various disciplines (see
Gleitman and Papafragou 2005, Kati 2013, Kati, Kondyli and Nikiforidou 1999
for an overview and critical account of different approaches). In very broad
terms, language has been treated as a phenomenon either mirroring or medi-
ating and restructuring cognition.1 In formalist theories of language and the
representational theory of mind (e.g. Chomsky 1965, Fodor 1975, 1983 cited in
Kati 2013, 48 and 73), which have been predominant in the study of language
and cognition, language is generally defined as an autonomous subsystem of
the mind and a computational code, which transfers pre-existing meaning. On
the contrary, various other (non-formalist) approaches within linguistics, such
as ethnography of communication, systemic functional grammar, linguistic
anthropology, cognitive linguistics and research on linguistic relativity con-
sider language as an important tool for cognition, which is embodied, socially
embedded and interrelated with meaning (Kati 2013).
In this chapter, I examine the relation between grammatical gender and
speakers’ cognition by drawing on two approaches from this second strand of
research: Lakoff’s (1987) approach to the relation between noun classes and
conceptual categorization, and Slobin’s (1996; 2003) theory on the role of lan-
guage in mediating ‘thinking for speaking’.
4.2 Cognitive Aspects of Grammatical Gender
4.2.1 Grammatical Gender and Conceptual Categorization

The relation between noun class systems and conceptual categorization is
examined by cognitive linguistics and more specifically by cognitive seman-
tics. For cognitive semantics, language is a “methodological tool for uncov-
ering conceptual organization and structure” (Εvans and Green 2006, 153).
1 This classification does not cover the rich complexity that is observed within each strand.

grammatical gender and speakers ’ cognition 53
More specifically, language is assumed to ‘prompt for’ meaning construction

and, thus, guide ‘construal’ (Croft and Cruse 2004) or conceptualization. This
means that language plays an active role in speakers’ cognition, by mediat-
ing the way in which speakers construct and interpret their experience of the
physical and sociocultural world. For cognitive semantics, meanings are con-
ventionally related with words and other linguistic items, such as morphemes,
and they constitute concepts. In this sense, grammatical gender can be said
to have some sort of conceptual content, given its semantic basis in person
reference. The information of referent’s sex that is codified morphologically
via grammatical gender is linked with the concept of sex. Therefore, grammati-
cal gender guides the construal of referents as female or male. Lakoff’s (1987)
approach to conceptual categorization will be our starting point for exploring
how grammatical gender is related to speakers’ thinking of person as female
or male.
One of the problems addressed by research on cognitive semantics is cat-
egorization. Categorization refers to the ability that humans have to identify
entities or objects as members of a particular group (Evans and Green 2006,
168). In his book Women, Fire and Other Dangerous Things, Lakoff (1987)
argues that there is a link between categorization and noun classes. In order
to approach categorization, Lakoff draws mainly on the idea formulated by
Rosch and others (e.g. Rosch 1978 cited in Lakoff 1987, 44) that conceptual cat-
egories are organized on the basis of prototypes or cognitive reference points.
According to this view, categories have fuzzy boundaries and consist of more
representative or prototypical members and less representative or peripheral
members. For example, chair is a more representative member of the category
furniture compared to carpet. In other words, chair constitutes a ‘better
example’ of the category.2 Lakoff argues that categories are linked with struc-
tures of knowledge, which he describes as idealized cognitive models. These
models are based on particular schematizations of experience, they constitute
theories about how the world is organized and they include speakers’ socio-
cultural knowledge. For example, the category bachelor is understood accord-
ing to a specific idealized cognitive model about the institution of marriage
and the age in which men usually get married (Lakoff 1987, 70). Pope, how-
ever, does not belong to the category bachelor, because Popes are linked with a
different idealized cognitive model, that of the Catholic Church.
2 Contrary to prototype theory, the classical approach (Taylor 2003) views categories as sets of
common features with clear-cut boundaries. For instance, man is defined as ‘male’, ‘human’,
‘adult’, while woman is defined as ‘female’, ‘human’ and ‘adult’.
54 chapter 4
Idealized cognitive models are metonymic, if they include concepts A and

B, which are associated metonymically, that is, B, which is either part of A
or is closely associated with it, stands for A (Lakoff 1987, 77–78). Metonymy
constitutes a sort of conceptual mapping or projection (Barcelona 2000), or
a reference-point phenomenon (Langacker 1993). In Radden and Kövesces’
(1999, 21) words, it is “a cognitive process in which one conceptual entity,
the vehicle, provides mental access to another conceptual entity, the target”
within the same idealized cognitive model. For example, I may refer to the us
army attacking Iraq, by saying Bush attacked Iraq. In this sentence, the name
Bush is the vehicle that provides mental access to the target, i.e. the us army.
According to Lakoff (1987, 84), a metonymic model consists of the following:
i) a “target” concept A to be understood for some purpose in some

context;
ii) a conceptual structure containing both A and another concept B;
iii) concept B which is either part of A or closely associated with it in that
conceptual structure, and
iv) which is easier to understand, remember, recognize or more immediately
useful for the given purpose in the given context, compared to A.
Social stereotypes are typical cases of metonymy in which one member of the
category is used as a vehicle for understanding the category as a whole. For
example, in patriarchal cultures housewife mothers tend to be taken as the
best examples of what mothers should be like. This stereotype defines which
members are considered to be the most representative and less representative
ones. Housewife mothers are prototypical members, while working mothers
are less central members (Lakoff 1987, 79–82).3
Lakoff (1987) investigates the relation between noun classes and conceptual
categorization by using the noun class system in Dyirbal as one of his exam-
ples. As we saw in section 2.2, Dyirbal has four noun classes organized on the
basis of semantic principles: i) bayi for male referents and non-human ani-
mates, ii) balan for female referents, fire, water and battle, iii) balam for non-
protein food (except for meat), and iv) bala for the rest. Lakoff (1987, 91–104)
suggests that these noun classes constitute conceptual categories with more
representative/prototypical and less representative/peripheral members. For
example, male and female humans are prototypical members in the categories
3 For this reason the category mother is said to have a radial structure (Lakoff 1987, 83–84).
bayi and balan respectively, while the moon and the stars are peripheral mem-
bers in the corresponding categories.
Although Lakoff examines a noun class system in which nouns are assigned
to specific classes according to semantic principles (cf. section 2.2), I suggest
that his approach may apply to grammatical gender systems as well. Although
in these systems gender assignment is mostly based on morphological or
phonological principles, there is a semantic basis when it comes to gen-
der assignment in nouns with human referents. As we saw in section 2.2, in
many languages with grammatical gender the categorization of nouns denot-
ing inanimate objects is semantically arbitrary to a large extent. However, in
person reference it is semantically motivated because of the match between
grammatical gender and referent’s sex. In addition, as section 4.3 shows, the
gender/sex match seems to be so important for speakers’ cognition that its
‘effect’ is not limited to speakers’ thinking of the human/animate world as
female or male but can also be extended to speakers’ thinking of the inanimate
world as ‘female’ or ‘male’. Following Lakoff’s approach, I suggest that in Greek
the concept of sex is grammaticized. Similar to bayi or balan in Dyirbal, the
masculine and feminine grammatical gender categories in Greek constitute
conceptual categories with prototypical and peripheral members. Male and
female humans are prototypical members in the masculine and feminine gen-
der categories respectively. In contrast, sun (masc) or lighter (masc), and rain
(fem) or door (fem), are peripheral members in the corresponding categories.
Additionally, Lakoff’s approach is useful in our understanding of the cogni-
tive role of grammatical gender in the following way. Köpcke and Zubin (2003)
suggest that there is a relation between grammatical gender and metonymy in
language use, by drawing on Lakoff’s idea about cognitive metonymic models.
These scholars argue that grammatically neuter linguistic items referring to
female persons are associated with a metonymic model, which embodies
the sociocultural stereotype of sexual innocence, social naiveté and depen-
dent social status. This mechanism is activated by the mismatch between the
neuter grammatical gender and referent’s female sex. When grammatically
feminine and neuter linguistic items are used in the same context to refer to
a female person, feminine gender is associated with a different metonymic
model that embodies the stereotype of sexual experience, social maturity and
sophistication.
Köpcke and Zubin (2003) examine the use of grammatical gender in refer-
ence to female humans in Karl Weggerl’s story Legende von den drei Pfändern
der Liebe. In this story, a young village pot-maker travels away from home to sell
his wares. His wife who is a young girl gives him three pledges of love, a hair
56 chapter 4
ribbon, a ring and a knife. During his travel a mysterious woman visits him at
night and sleeps with him. Every morning he gives her one of the love pledges.
On his way back home he finds the ribbon, then the ring and finally the girl
with the knife in her breast, and he realizes that his wife was the mysterious
woman that visited him at night. Köpcke and Zubin point to two archetypes
on which this story is based: the archetype of the sexually innocent village
girl and the archetype of the experienced independent and amoral woman of
the city. The former is evoked by the use of grammatically neuter nouns and
pronouns in anaphora, e.g. Mädchen and es. The latter is evoked by the use of
grammatically feminine nouns and pronouns, e.g. Frau and sie. For instance,
in the first part of the story neuter items are used to refer to the girl, and in the
second part of the story feminine items are used to refer to the woman. Yet,
in the third part of the story in which the hero realizes that the girl and the
woman are the same person, reference to the same female person is achieved
via feminine and neuter items. This is shown in the following extract: “when
he entered the main room, there lay his girl (sein Mädchen) on the bier. And
he knew that she (sie) was the one (es) whom (die) he had loved three times,
and three times betrayed and now there was his knife thrust in the middle of
her (ihrer) white breast” (Köpcke and Zubin 2003, 149). Neuter linguistic items
evoke metonymically the archetype of the innocent village girl character, while
feminine linguistic items evoke metonymically the archetype of the mysteri-
ous amoral woman character.4 The referent is the same but each time she is
associated with properties of different stereotypes.
Following Köpcke and Zubin’s argument, I suggest that metonymy may be
linked with other uses of grammatical gender as well. More specifically, meton-
ymy can help us understand how the generic use of the masculine operates. As
I mentioned in section 3.3.2, the generic use of the masculine depends on the
stereotypical assumption that man is the universal category or the norm. This
sociocultural stereotype relates to the cognitive dimension of the masculine
grammatical gender in two ways. First, this stereotype structures the category
human, by defining which members are taken to be prototypical in this cat-
egory and which members are taken to be less representative. Given speakers’
knowledge about gender hierarchy, it is reasonable to assume that male
humans are prototypes or cognitive-reference points for the category human,
while female humans are peripheral members. Second, this stereotype can be
said to form a metonymic cognitive model on which the generic use of the
masculine is based. This model consists of the following elements:
4 The pronoun die refers to the female person non-metonymically (Köpcke and Zubin 2003).
i) a concept A (‘target’) that needs to be understood for some purpose in

some context, i.e. the concept of human/universal;
ii) a conceptual structure containing concepts A and B, i.e. the cultural asso-
ciation of human/universal with male sex (man is the ‘prototypical
human’), and
iii) a concept B (‘vehicle’) which is either part of A or closely associated with
it in that conceptual structure, i.e. the concept of male sex which is part
of the category human/universal.
In this metonymic model, male sex operates as the vehicle through which
the human/universal (target) is accessed and understood. This metonymic
model is associated with the use of the masculine gender in certain contexts,
as shown in example 40. Suppose that an employee at a museum produces the
following utterance:
(40) Σήμερα επισκέφτηκαν το

símera episkéftikan to
today visit.3pl.past the.acc.neuter
μουσείο οι μαθητές
musío i maθités
museum.acc.neuter the.nom.masc.pl students.nom.masc
του έκτου Δημοτικού
tu éktu ðimotikú
the.gen.neuter sixth.gen.neuter primary.gen.neuter
Σχολείου μαζί με
sxolíu mazí me
school.gen.neuter together with
τους δασκάλους.
tus ðaskálus
the.acc.masc.pl teachers.acc.masc
‘Today the (male) students of the sixth Primary School visited the museum
together with the (male) teachers.’
The speaker refers to all students and teachers of the primary school, that is,
males plus females, who took part in the school trip. In order to do that, she
employs grammatically masculine nouns, which control agreement with the
modifiers, i.e. articles. In this case, generic reference is achieved metonymi-
cally. Male sex, which is morphologically codified, activates metonymically
the category of human/universal and guides speakers to interpret referents as
not exclusively male. Therefore, the category of human/universal is a sort of
58 chapter 4
inference that is produced because speakers presuppose the cultural associa-

tion of male sex/man with human/universal (more on this in chapter 5).
The example above illustrates generic reference to third persons. The
inference of human/universal is also produced when speakers refer to self or
hearer, as shown in example 41. Here the female speaker uses the first person
plural verb to refer to a collectivity which includes herself and her male friend
Dimitris, and she uses the masculine participle to describe their emotional
condition.
(41) Εγώ και ο Δημήτρης

eɣό ce o ðimítris
I and the.nom.masc Dimitris.nom.masc
είμαστε ενθουσιασμένοι με την
ímaste enθusiazméni me tin
cop.1pl.pres excited.part.masc.pl with the.acc.fem
ταινία.
tenía
movie.acc.fem
‘Me and Dimitris are excited about the movie.’
However, there may be cases in language use in which the masculine gender is
used for indefinite reference but the metonymic model of male sex as human/
universal is not necessarily activated. This is illustrated with example 42.
(42) Η Υπουργός Μεταφορών

i ipurɣόs metaforόn
the.nom.fem minister.nom.fem transport.gen.fem.pl
ζήτησε από τους οδηγούς
zítise apό tus oðiɣús
demand.3sg.past from the.acc.masc.pl drivers.acc.masc
των ταξί να σταματήσουν
ton taksí na stamatísun
the.gen.neuter.pl taxi subj stop.3pl.pfv
την απεργία.
tin aperʝía
the.acc.fem strike.acc.fem
‘The (female) Minister of Transport asked the (male) taxi drivers to stop the strike.’
The noun phrase marked by the masculine gender refers to all taxi drivers. Yet,
these referents are more likely to be interpreted as male only, rather than as
male plus female, given speakers’ knowledge about the job of driving a taxi as
stereotypically masculine.
Overall, based on the approaches mentioned in this section, the following
hypotheses can be made. Sex is a concept that gets grammaticized in Greek
and the use of the masculine grammatical gender is linked with metonymy.
The grammaticization of the concept of sex is expected to have significant con-
sequences for speakers’ cognition according to research on linguistic relativity.
The latter research shows that grammatical categories play a significant role in
mediating speakers’ thinking. This topic is examined in the next section.
4.2.2 Grammatical Gender and the Linguistic Mediation of Thought

The idea that language mediates speakers’ cognition is associated with the
principle of linguistic relativity or the Sapir-Whorf hypothesis, as it is widely
known. Linguistic relativity refers to the correlation between linguistic and
cognitive differences, and is based on a more general sort of relativity, which
has been formulated by Lucy (1996) as semiotic relativity. According to semiotic
relativity, language constitutes a semiotic system that mediates speakers’
experience of the world. Linguistic relativity originated in American linguistic
anthropology in the beginning of the 20th century and is basically associated
with the work of Boas, Sapir and Whorf.5 The idea of linguistic relativity can
be formulated briefly as following. Grammatical categories guide speakers
to specific unconscious interpretations of experience, because they are
obligatory, automatic and systematic. Therefore, differences across languages
are expected to correlate with differences in respect to what is conceptually
salient in a speech community.
Boas ([1911] 1991; 1938) examines the different ways in which speakers of dif-
ferent languages describe the same situation given the different grammatical
categories that languages provide. In his words,
Grammar [. . .] determines those aspects of each experience that must be

expressed. [. . .] These obligatory aspects are expressed by means of gram-
matical devices. The aspects chosen in different groups of languages vary
fundamentally. (Boas 1938, 132–133)
5 The roots of linguistic relativity can also be traced in German romanticism and Humboldt’s
philosophy, as well as structuralist paradigms, more specifically Saussure’s theory of language
as a system of signs and Durkheim’s sociology (Gumperz and Levinson 1996b, 4–5).
60 chapter 4
For example, when English speakers produce the utterance Τhe man is sick,
they codify number, tense and definiteness. However, in Siouan speakers must
codify whether the man is standing or moving, in Kwakiutl they must codify
whether the man is visible or not by the speaker and whether he is close to
speaker, hearer or third person. Thus, languages set limits on what speakers
must say and on which particular aspects of experience must be verbalized.
According to Boas, the concepts that get grammaticized in a language tend
to be conceptually salient for its speakers. Moreover, the interpretations of
experience that are associated with specific grammatical categories are uncon-
scious because grammatical categories are automatic, obligatory and system-
atic. In a similar vein, Sapir (1970) argues that language constitutes a ‘guide’ to
social reality which channels the way in which speakers think about the world.
In Sapir’s words:
Language is a guide to ‘social reality’. [. . .] Human beings do not live in

the objective world alone [. . .] but are very much at the mercy of the par-
ticular language which has become the medium of expression for their
society. [. . .] The fact of the matter is that the ‘real world’ is to a large
extent unconsciously built up on the language habits of the group. [. . .]
We see and hear and otherwise experience very largely as we do because
the language habits of our community predispose certain choices of
interpretation. (Sapir 1970, 162)
Building on the ideas of Boas and Sapir, Whorf (1956) describes the role of lan-
guage in guiding speakers’ thinking of the world in the following way:
We dissect nature along lines laid down by our native languages. The cat-
egories and types that we isolate from the world of phenomena we do not
find there because they stare each observer in the face; on the contrary,
the world is presented in a kaleidoscopic flux of impression which has
to be organized by our minds—and this means by the linguistic systems
of our minds. We cut nature up, organize it into concepts, and ascribe
significances as we do, largely because we are parties to an agreement
to organize it in this way—an agreement that holds throughout our
speech community and is codified in the patterns of our language. The
agreement is, of course, an implicit and unstated one, but its terms are
absolutely obligatory; we cannot talk at all except by subscribing to the
organization and classification of data which the agreement decrees.
(Whorf 1956, 213–214)
If language guides speakers’ thinking, then the differences among languages

are expected to correlate with differences in the way speakers of these lan-
guages think. This idea constitutes the “linguistic relativity principle” (Whorf
1956, 221). In Whorf’s words:
Users of markedly different grammars are pointed by their grammars

towards different types of observations and different evaluations of
externally similar acts of observation, and hence are not equivalent as
observers but must arrive at somewhat different views of the world.
(Whorf 1956, 221)
Whorf examines the relation between ‘fashions of speaking’ established by

grammatical categories and speakers’ habitual thought. Habitual thought
refers to “the microcosm that each man carries about within himself, by which
he measures and understands what he can of the macrocosm” (Whorf 1956,
147), or in Lucy’s (1992, 44) reading, to concepts that speakers use in their every-
day life in order to interpret their experience. For example, Hopi and English
codify number in different ways. As Whorf argues, these differences correlate
with differences in the way Hopi and English speakers understand the concept
of time. According to Whorf’s speculations, Hopi speakers treat days or years
as recurrent events, while English speakers treat them as objects that can be
counted and measured. Whorf associates habitual thought with both classifi-
cations that are overtly marked in a systematic way, such as the plural number
in English, and more covert classifications, such as the feminine and masculine
gender in English, which is grammatical only in third person pronouns he and
she. He defines the first sort of classifications as ‘phenotypes’ and the second
sort as ‘cryptotypes’, and he suggests that cryptotypes ‘affect’ speakers’ inter-
pretations in more significant ways, because they lack morphological marking
and, thus, cannot be detected easily by speakers.
Speakers’ cognitive differences may correlate with differences in the struc-
ture of languages or in the use of language. Lucy (1996) defines the first kind of
relativity as structural and the second as discursive (or functional). Research on
structural relativity addresses the structural diversity of natural languages, and
explores “whether and to what extent the characteristics of specific languages
have an impact on the thought or behavior of those who speak them” (Lucy
1996, 41). Research on discursive relativity addresses the functional diversity
found within the same language, and explores “whether patterns of use have
an impact on thought either directly or by virtue of amplifying or channeling
any effects due to linguistic structure” (Lucy 1996, 52; see Hymes 1966 for the
62 chapter 4
formulation of this idea). These patterns of use may involve differences in the
ways in which language is used by subgroups in the same language community
or in contexts of speaking (e.g. Gumperz 1996). Although the perspective taken
by discursive relativity is an interesting one—especially for scholars working
on sociolinguistics—it is not relevant to the question examined in this chap-
ter. The question about the cognitive role of grammatical gender foregrounds
the idea of structural relativity, because it specifically involves the relation of a
feature of grammar to speakers’ cognition.
Issues of structural relativity are examined by a number of “psychologically
informed” (Sidnell and Enfield 2012, 302), or Neo-Whorfian, studies, situated
mainly within the context of psycholinguistics. Some of these studies can be
found in the volumes edited by Gumperz and Levinson (1996a), Niemeier and
Dirven (2000), Pütz and Verspoor (2000), Bowerman and Levinson (2001), and
Gentner and Goldin-Meadow (2003). These studies treat linguistic relativity as
a hypothesis that can be empirically investigated via experimental methods.6
Within the framework of structural relativity Slobin (1996; 2003) formu-
lated the hypothesis that language mediates speakers’ cognition at least at the
time of speaking. Slobin replaced the terms language and thought with the
terms speaking and thinking respectively, and he argued that these phenom-
ena should be conceptualized as mental processes that occur when speakers
formulate utterances, rather than as abstract entities or static phenomena.
When speakers design and produce utterances, they choose to codify particu-
lar aspects of experience depending on what is grammaticized in the language
they speak. Thus, the mental representations constructed at the time of speak-
ing differ across languages. Slobin defines the sort of thinking that emerges
while speaking as ‘thinking for speaking’. In his words:
The activity of thinking takes on a particular quality when it is employed

in the activity of speaking. In the evanescent time frame of constructing
utterances in discourse, one fits one’s thoughts into available linguistic
forms. A particular utterance is never a direct reflection of “objective” or
perceived reality or of an inevitable and universal mental representa-
tion of a situation. This is evident within any given language, because
the same situation can be described in different ways; and it is evident
6 This hypothetico-deductive appproach to linguistic relativity (see Hill and Mannheim 1992
for the term) has been criticized by Hill and Mannheim (1992), Lee (1996) and Enfield (2002)
as deviating from the original Whorfian position. According to this position, linguistic rela-
tivity is a principle or an ‘axiom’, in Foley’s (1997, 192) terms, that is, a sort of background
assumption for addressing questions and conducting research.
across languages, because each language provides a limited set of options

for the grammatical encoding of characteristics of objects and events.
“Thinking for speaking” involves picking those characteristics that (a) fit
some conceptualization of the event, and (b) are readily encodable in the
language. (Slobin 1987, 435 cited in Slobin 2003, 158)
Similar to Boas, Slobin argues that concepts that get grammaticized in lan-
guages are conceptually salient for the speakers of these languages. For exam-
ple, these concepts are used frequently in speech and are acquired by speakers
at an early stage. In addition, speakers are shown to have easy lexical access
to these concepts. According to ‘thinking for speaking’ hypothesis, speakers’
thinking before and at the time of speaking will be attuned to the concepts
that get grammaticized in the language that they speak. The experiments
conducted by Slobin and Berman with English, German, Spanish and Hebrew
speakers (Berman and Slobin 1994 cited in Slobin 1996, 93) provide indica-
tions of the mediation of language structure to speakers’ thinking. Slobin and
Berman examined time and space expressions in speech. Adults and 3–5 years
old children speaking the abovementioned languages were asked to describe
the events presented in the picture book Frog, Where Are You? (Mayer 1969
cited in Slobin 1996, 72). These scholars found that speakers of different lan-
guages describe the same events in different ways depending on how each lan-
guage codifies the concept of time and space. For example, English codifies
the durative/non durative distinction in the verb and, thus, guides speakers to
attend to the duration of an event or to the lack of it in a compulsory way. In
contrast, Hebrew does not grammaticize this distinction and, thus, it does not
guide speakers to attend to this specific feature of experience when they think
for speaking.7
The structural diversity observed across languages provides analysts
with a rich source for the possible different ways in which languages can
mediate speakers’ thinking for speaking. For example, Turkish is a language
that grammaticalizes evidentiality (Aikhenvald 2004). This means that when
Turkish speakers refer to some event that took place in the past, they must
codify whether they witnessed the event they refer to or whether they heard
it from someone else. Therefore, language is expected to guide them to attend
to this particular feature of experience, when they speak. On the contrary,
speaking Greek does not involve the same mental activity, given that this
7 In his later work, Slobin (2003) argues that language mediates speakers’ thinking not only
at the time of speaking but also at the time of hearing, reading and looking, and it involves
mental processes such as imagination and memory.
64 chapter 4
language does not grammaticalize evidentiality. Also, demonstratives across

languages provide an interesting example on how structural differences among
languages may require speakers to codify different conceptual distinctions. As
comparative typological research has shown, languages differ with respect to
the meanings of demonstratives. For example, English has a two-term system
of nominal and adverbial demonstratives (see Dixon 2010b, 225 for the terms)
that indicate proximity to (this/that) or distance from (here/there) speaker. In
contrast, Quechua has a three-term system in which nominal and adverbial
demonstratives indicate proximity to speaker (kay ‘this one here’), proximity
to addressee (chay ‘that one there’) and distance from both speaker and
addressee (taqay ‘that one over there’) (Dixon 2010b, 239; Weber 1989, 38). This
means that English speakers must use a bipartite code and, thus, attend to the
proximity to/distance from speaker distinction, while Quechua speakers must
use a tripartite code and, thus, attend to the distinction between proximity to
speaker, proximity to addressee and distance from both speaker and addressee.
Even if one disagrees with a strong version of linguistic relativity, according
to which structural differences across languages correlate with differences in
speakers’ non-verbal thinking (e.g. Levinson 2003),8 there seems to be some
sort of consensus on a weaker version of linguistic relativity, according to
which language mediates at least speakers’ thinking at the time of speaking.9
As Slobin (1996, 76) highlights, no matter how subtle or trivial the impact of
language on thinking for speaking may sound, it is not obvious and is certainly
worth scholars’ attention.
In line with anthropological linguistic research (e.g. Bickel 2000; Hanks 1990),
Slobin argues that ‘thinking for speaking’ is not an autonomous phenomenon,
independent of the cultural context in which language is used. On the contrary,
8 According to Levinson (2003), differences among languages may be related with differences
in speakers’ experiencing for speaking. Given that experience will be verbalized at some
point, it is possible that speakers codify experience non-linguistically in ways that are appro-
priate for its verbal expression. Therefore, speakers of different languages are expected to
differ not only in their thinking for speaking, but also in their experiencing for speaking.
9 For example, a weak version of linguistic relativity seems to be supported by Levelt (1989),
whose work informed Slobin’s approach to thinking for speaking. According to Levelt (1989),
speech production is divided into four systems: i) the production of preverbal messages
which consist of the conceptual information that needs to be expressed for realizing the
speakers’ intention (Conceptualizer), ii) the grammatical and phonological codification of
preverbal messages (Formulator), iii) the production of external speech (Articulator), and
iv) the understanding and control of speech (Speech-Comprehension System). Differences
among languages can be found in Formulators, which differ across languages and require
speakers to codify different conceptual distinctions.
‘thinking for speaking’ is related to speakers’ sociocultural practices, that Slobin

(2003, 180) describes as “online processes that include thinking and speaking”.
This point raised by Slobin foregrounds the social embeddeness of mental
phenomena. Cognition is grounded in sociocultural practices that speakers
employ to conduct their everyday activities and construct their common world
(more on this point in chapter 5).10 In Gumperz and Levinson’s (1996b, 230)
words, “we cannot think of a ‘world-view’ as inherent in a language, somehow
detached from all the practices established for its use”.
Ιf we take research on linguistic relativity into account, we arrive at a series
of theoretical hypotheses about the cognitive role of grammatical gender.
Because sex is a grammaticized concept in Greek (cf. section 4.2.1), it is con-
ceptually salient for Greek speakers. Moreover, given that grammatical gen-
der is a compulsory morphological category that is employed systematically
in communication, it is expected to guide speakers to interpret experience in
automatic and unconscious ways. More specifically, grammatical gender is
expected to orient speakers to the interpretation of person as female or male.
According to Slobin’s thinking for speaking hypothesis, grammatical gender
mediates speakers’ thinking at least at the time of speaking. When Greek
speakers design and produce utterances, they must codify referent’s sex in a
compulsory way. Thus, in communication speakers’ thinking is attuned to the
conceptual distinction of sex, because the Greek language guides speakers
to attend to referent’s sex when they think for speaking. In this sense, gram-
matical gender in Greek seems to presuppose a particular kind of thinking for
speaking which is absent from languages with no grammatical gender.
4.3 Indications of the Cognitive Role of Grammatical Gender:

Empirical Studies
The theoretical hypotheses that I formulated in the previous section are sup-
ported by a number of empirical studies from sociolinguistics and psycholin-
guistics to be discussed in this section. Although these studies have different
aims, are based on different theoretical backgrounds and employ different
methodological tools, they all show a correlation between grammatical gender
and speakers’ cognition.
10 As Bickel (2000, 185) insightfully points out, “sociocultural practices, whether verbal or
nonverbal, [. . .] sustain the cognitive style and bias of awareness that is required by a
particular grammar”.
66 chapter 4
The first group of studies has a sociolinguistic orientation (Braun 2001;

Doleschal and Schmid 2001; Engelberg 2002; McKay and Fulkerson 1979; Nissen
2002). These studies aim to explore the relation between language and sexism
in various languages in which sex is grammatically or lexically codified.
In a study conducted by Doleschal (1993 and 1997 cited in Doleschal and
Schmid 2001, 266), Russian speakers were given a questionnaire with sen-
tences. Each sentence included a masculine noun. Speakers were asked to
give a proper name to the person denoted by each noun. The study showed
that speakers interpreted the majority of referents as male, e.g. vrag ‘enemy’,
aptekar΄ ‘pharmacist’ and avtor ‘author’. Yet, this interpretation was shown to
depend on speakers’ knowledge about stereotypically feminine and masculine
occupations. For example, most speakers interpreted the person denoted by
the masculine noun vrač ‘doctor’ as female, problably because the majority of
doctors in the former ussr were women. In another study (Schmid 1998 cited
in Doleschal and Schmid 2001, 266), speakers were given the initial sentence
of a text and were asked to compose short stories. Each sentence consisted of
a masculine noun, e.g. specialist, with no clues that would indicate the sex of
the referent. Speakers interpreted the persons denoted by masculine nouns
mostly as male.
In a similar study conducted by Nissen (1997 cited in Nissen 2002, 259),
Spanish speakers were given sentences with a masculine noun (e.g. los niños
‘the children’), an epicene noun which does not distinguish between female
or male sex (e.g. la población infantil ‘the population of children’), and a noun
phrase with masculine and feminine nouns (e.g. los niños y niñas ‘the boys and
girls’). They were asked to give proper names to referents. The study found that
when the masculine or epicene nouns were used, speakers tended to intrepret
referents as male. When both masculine and feminine nouns were used, speak-
ers interpreted referents as male and female.
Similar findings are reported for the generic use of the masculine gender
in German (Braun et al. 1998; Ιrmen and Köhncke 1996; Klein 1988 cited in
Bussmann and Hellinger 2003, 161). These studies report that speakers do not
always interpret referents as male plus female or gender-indefinite, when the
masculine gender is used for generic reference. They also indicate that the
interpretation of referent’s sex depends on the context of use.
Indications of the correlation of grammatical gender with the interpreta-
tion of sex are also provided for Greek. In a study conducted by Pavlidou (1985),
Greek speakers were given a text consisting of lexical morphemes in which
grammatical gender was not specified, e.g. πρόεδρ- [prόeðr] for ‘president’,
and συνήγορ- [siníɣοr] for ‘advocate’. Speakers were asked to add articles and
endings to these morphemes. Pavlidou found that the majority of speakers
ascribed male sex to nouns. The only exception to this pattern was the noun
κοινωνικός λειτουργός [cinonikόs liturɣόs] ‘social worker’, which was interpreted
by most speakers as female. According to Pavlidou (1985), speakers’ interpre-
tations were associated with the generic use of the masculine, which guides
speakers to the interpretation of referent as male, and with speakers’ knowl-
edge about stereotypically feminine and masculine occupations. The generic
use of the masculine noun άνθρωπος ‘human being’ was examined by Makri-
Tsilipakou (1989) (cf. section 3.3.2). In this study, speakers were given a ques-
tionnaire with sentences consisting of the noun άνθρωπος with no contextual
clues about referent’s sex. They were asked to give a name to the person
denoted by the noun. Makri-Tsilipakou found that speakers tended to interpret
referents as male. A more recent study conducted by Pavlidou and Alvanoudi
(2013) also indicates that the feminine and the masculine grammatical gen-
der in Greek correlate with the interpretation of person as female and male
respectively. In a sex attribution task, Greek speakers were asked to attribute
male or female proper names to human beings depicted in pictures. Speakers
assigned male or female sex to human beings according to the grammatical
gender of the noun denoting these items. Therefore, an almost perfect match
between grammatical gender and referent’s sex was found for Greek. The find-
ings of this study are presented in detail later in this section.
The studies discussed above indicate that there is a correlation between
grammatical gender and the interpretation of referent’s sex. They show that
the generic use of the masculine is linked with speakers’ strong tendency to
interpret referent(s) as male. They also uncover the role that context plays
in these interpretations. For example, if the noun denotes a stereotypically
masculine activity, it is likely that speakers will interpret referent(s) as male.
Additionally, these studies show that the use of feminine gender ensures that
reference to women is achieved.
Interestingly, a male bias in the interpretation of referent’s sex is also reported
for languages with no gender system at all, such as Finnish (Engelberg 2002)
and Turkish (Braun 2001), or in languages with limited gender distinctions,
such as English (MacKay and Fulkerson 1979). For example, a study conducted
by Braun (2001) showed that Turkish speakers interpret the sex of referents
denoted by gender-indefinite nouns, based on their knowledge about stereo-
typical masculine and feminine occupations. For instance, the nouns sekreter
‘secretary’ and kuyumcu ‘goldseller’ denote stereotypically feminine and mas-
culine occupations respectively. Speakers tend to interpret sekreter as referring
to a female person and kuyumcu as referring to a male person. Therefore, even
in a language with no gender system, gender-related associations still exist and
“remain hidden on a deeper semantic level” (Hellinger and Bussmann 2001, 11).
68 chapter 4
This is an interesting finding, if we consider Whorf’s idea that covert classifica-

tions can affect speakers’ cognition (cf. section 4.2.2). Some reflections on this
topic are presented in chapter 7.
The correlation between grammatical gender and the interpretation
of sex in person reference seems to be so strong that sometimes it can also
be extended to the conceptualization of the inanimate world as ‘female’ or
‘male’. Empirical data from language use and experiments indicate that this is
the case.
Sometimes speakers of languages with grammatical gender represent
the inanimate world as ‘male’ or ‘female’ in poetry, songs or in the media.
For example, the city of Thessaloniki is grammatically feminine in Greek
(Θεσσαλονίκη [θesaloníci] ‘Thessaloniki.FEM’). Very often Thessaloniki
is represented as a woman: in songs (Θεσσαλονίκη μου μεγάλη φτωχομάνα
[θesaloníci mu meɣáli ftoxomána] ‘My Thessaloniki, the great mother of the
poor’), in descriptions (η νύμφη του Θερμαϊκού [i nímfi tu θermaikú] ‘the bride of
Thermaikos sea’) or in magazines. For example, on the frontpage of a newspaper
magazine Thessaloniki is represented as a young woman (Pavlidou 2006b). The
picture of the woman is accompanied by a caption describing the city with
grammatically feminine adjectives, such as ερωτική [eroticí] ‘erotic.FEM’. A
similar example is also found in a monthly free-press magazine presenting
architects’ work on redesigning the city of Athens. The noun denoting Athens
is grammatically feminine (Αθήνα [aθína] ‘Αthens.FEM’). Ιn the picture
accompanying the article, a naked female doll reclines knees bent on Omonoia
square, staring up at the Acropolis, and allows the masses to pass through it
and inside her. The picture is entitled Ομόνοια: το ‘αιδοίο’ της Αθήνας [omόnia to
eðío tis aθínas] ‘Omonoia: the “vulva” of Athens’. Here Athens is represented
as a woman because of the match between the feminine gender and referent’s
female sex. Wierzbicka (2002) mentions similar cases for Russian and German.
In Russian poetry, moon (luna) and death (smert) are represented as women,
because Russian nouns denoting moon and death are feminine. However, in
German poetry these items are represented as men, because they are denoted
by masculine nouns (der Mond, der Tod).
The role of grammatical gender in speakers’ cognition is also examined by
a number of psycholinguistic empirical studies, which examine issues of
structural relativity (e.g. Boroditsky, Schmidt, and Phillips 2003; Flaherty
2001; Imai et al. 2013; Mills 1986; Pavlidou and Alvanoudi 2013; Saalbach,
Imai, and Schalk 2012; Sera, Berge, and del Castillio Pintando 1994; Sera et al.
2002; Topsakal 1995).11 These studies show that grammatical gender mediates
11 Another group of studies (Clarke et al. 1981; Ervin 1962; Flaherty 2001; Konishi 1993; Μills
1986) examines whether masculine and feminine nouns carry connotations of masculin-
speakers’ thinking of the inanimate world as ‘female’ or ‘male’. However, the

findings reported by these studies differ with respect to which languages can
produce such an effect. For example, Sera et al. (2002) argue that the role of gram-
matical gender in speakers’ cognition is limited to languages with a two-gender
system, while other studies (e.g. Boroditsky, Schmidt, and Phillips 2003) indi-
cate that grammatical gender effect can also be found in languages with a three-
gender system.
A common question examined by the majority of these studies is whether
grammatical gender correlates with sex attribution to inanimate objects.12 In
sex attribution tasks, speakers were presented with objects denoted by words
or depicted by pictures. They were asked to give objects a name, for example a
proper name that would attribute female or male sex to them. The sex attribu-
tion task conducted by Palvidou and Alvanoudi (2013) for the Greek language
is of special interest here. The aim of this task was to examine whether gram-
matical gender has an effect on Greek speakers’ cognition, and whether this
effect is restricted to certain semantic categories. Speakers were presented with
a set of 40 colored pictures or drawings depicting persons, animals, and natural
and artificial inanimate objects (e.g. ανανάς ‘pineapple.masc’, βαλίτσα ‘suitcase.
fem’, αυγό ‘egg.neuter’, γλάρος ‘seagull.masc’, καμήλα ‘camel.fem’) and a list
of 20 proper names for women and men in equal numbers (e.g. Μαρία ‘Maria.
fem’, Νίκος ‘Nikos.masc’). None of the objects was stereotypically related with
women or men. Speakers were asked to attribute sex to depicted items by
choosing a proper name from the list. The hypothesis tested was that subjects
would assign male or female sex to objects/animals/persons according to the
grammatical gender of the noun denoting those items.
The sex attribution task yielded a high correlation between grammatical
gender and the sex attributed to objects/animals/persons: 70.3% of the mas-
culine nouns correlate with male sex and 61.8% of the feminine nouns cor-
relate with female sex. More specifically, the following results were found with
ity and femininity respectively, by employing semantic differential tests. In these tests,
speakers are asked to rate the meanings of nouns on bipolar adjective scales that include,
for example, factors of potency (e.g. strong/weak) or evaluation (good/bad), and are asso-
ciated with ‘masculinity’ and ‘femininity’ (see Pavlidou and Alvanoudi 2013, 110 for a sum-
mary of the criticism that has been articulated with respect to these tests).
12 Sex attribution to inanimate objects is one of the different questions investigated by these
studies. For example, Boroditsky, Schmidt and Phillips (2003) examined whether gram-
matical gender ‘affects’ speakers’ memory of word-name pairs, assessment of similarity
between persons and objects, and description of objects. Contrary to the previous stud-
ies that were conducted in speakers’ native language, the tasks employed by Boroditsky,
Schmidt and Phillips were conducted in the same non-native genderless language, i.e.
English, to ensure that speakers understand stimuli and instructions in the same way.
70 chapter 4
respect to the different semantic categories examined. For persons there is

an almost perfect match between masculine nouns and male sex (91.5%) and
feminine nouns and female sex (93.8%). A similar correlation between gram-
matical gender and sex is found in the semantic category of animals: 71.3% of
the masculine nouns were given male sex, while 67.8% of the feminine nouns
were given female sex. A less pronounced correlation between grammatical
gender and sex is found in the case of inanimate objects: 60.9% of the mascu-
line nouns were assigned male sex, while 52.8% of the feminine nouns were
assigned female sex.
These findings indicate a grammatical gender effect on speakers’ thinking
of humans/animates and inanimate objects in Greek. The correlation between
grammatical gender and sex found in the semantic category of person seems
to be so strong that it can be extended also to the semantic category of inani-
mate objects. This finding provides further support for the hypothesis that I
formulated in the previous section on the cognitive role of grammatical gender
in person reference in Greek.
In general, the findings reported by psycholinguistic studies uncover
aspects of the relation between grammatical gender and cognition. However,
these studies show certain ‘weaknesses’, some of which can be summarized as
following. First, these studies do not provide a systematic pattern of results,
given that some studies indicate a grammatical gender effect in languages with
a two-gender system, while others indicate such an effect only for languages
with a three-gender system. Therefore, it is not easy to arrive at any definite
conclusions about the cognitive role of grammatical gender across languages
in general. Grammatical gender seems to leave some sort of ‘cognitive trace’
on speakers’ minds. Yet, more research needs to be done before we make any
sorts of generalizations. Second, it is not clear what ‘thinking of the inanimate
world as female or male’ actually means for speakers’ everyday mental life. In
sex attribution tasks, speakers are asked to attribute sex to various items. While
this task makes sense when the items depicted are humans or animals (e.g.
people usually name their pets), it makes little or no sense when the items are
inanimate objects. This experimental condition is artificial and is not directly
related to speakers’ everyday life experience (cf. Imai et al. 2013). In spite of
their problems, I believe that these studies show that grammatical gender
interweaves somehow with conceptualization.
The sociolinguistic and psycholinguistic studies discussed in this section
test the role of grammatical gender in speaker’s thinking in experimental con-
ditions. Yet, their findings could also be indicative of the role of grammatical
gender in speakers’ thinking in communication.
4.4 Summary
The theoretical approaches and empirical studies presented in this chapter

shed light on different aspects of the relation between grammatical gender
and speakers’ cognition and, thus, contribute to our understanding of the
cognitive role of grammatical gender in person reference in complementary
ways. The argument presented in this chapter can be summarized as follow-
ing. Based on Lakoff’s (1987) approach to the relation between noun classes
in Dyirbal and conceptual categorization (cf. section 4.2.1), I assumed that the
categories of the masculine and feminine gender in Greek constitute concep-
tual categories in which female and male humans are prototypical members.
Moreover, following Lakoff’s (1987) theory on idealized metonymic cognitive
models and Köpcke and Zubin’s (2003) study on the relation of these models
to grammatical gender in German, I suggested that the generic use of the
masculine depends on an idealized metonymic cognitive model that incorpo-
rates the sociocultural stereotype of man as the category of human/universal.
Βy drawing on research about linguistic relativity (e.g. Slobin 1996, cf. section
4.2.2), I formulated the hypothesis that grammatical gender guides Greek
speakers to interpret person as female or male in automatic and unconscious
ways. Finally, I showed that these theoretical hypotheses about the cognitive
role of grammatical gender seem to be supported by various empirical studies
that indicate a relation between grammatical gender and speakers’ cognition
(cf. section 4.3).
Given the above, the cognitive dimension of grammatical gender in per-
son reference can be described as such: in language use grammatical gender
guides/orients speakers to interpret person(s) as female or male.
CHAPTER 5
Grammatical Gender, Culture and

Cognition in Interaction
5.1 The Research Question
The cultural and cognitive aspects of grammatical gender examined in the

previous two chapters are not separate phenomena. On the contrary, they
interrelate with each other when speakers use language in interaction, and they
form a sort of ‘hybrid’. This hybrid can be briefly described as follows. Grammatical
gender ascribes sex to referents and constructs sociocultural gender on the
basis of gender hierarchy (cf. chapter 3). This social dimension of grammatical
gender presupposes a cognitive one, that is, it presupposes that grammatical
gender guides speakers to interpret referents as female or male in automatic
and unconscious ways (cf. chapter 4). In the case of the generic use of the
masculine this interpretation is linked with the inference of human/universal.
This inference is based on the stereotypical association of male sex/man with
the norm and the knowledge that speakers share about gender hierarchy as a
feature of the sociocultural world they live in. Following Slobin (1996; 2003),
the interpretation of referent’s sex and the inference of human/universal
can be said to be part of the mental representations constructed on-line in
communication, that is, at the time of speaking, when speakers produce and
understand utterances in interaction. Based on the discussion so far, I suggest
that the use of grammatical gender in Greek pressuposses a specific kind of
‘thinking for speaking’ that maintains and reproduces an asymmetrical bipolar
gender order.
My aim in this book is to examine empirically the interrelation between
the social and cognitive dimensions of grammatical gender in interaction. As
Hellinger and Bussmann (2001, 21) suggest, research on grammatical gender
should address the ways in which “gendered messages are interpreted” and
more generally, the ways in which “the perception and construction of the
universe is influenced by linguistic, social and cultural parameters”. The study
presented in this book contributes to our understanding of the complex inter-
section of grammatical gender with culture and cognition.
The empirical investigation of grammatical gender in interaction in particu-
lar is interesting for two reasons. First, interaction seems to be an appropri-
ate or ‘promising’ locus for the study of the relation between language and

grammatical gender, culture and cognition in interaction 73
cognition, given Slobin’s (1996; 2003) ‘thinking for speaking’ hypothesis. As I

mentioned in section 4.2.2, even if one disagrees with the idea that language
affects non-verbal thinking, there seems to be some sort of consensus on the
idea that language mediates at least speakers’ thinking at the time of speak-
ing. If grammatical gender is shown to guide speakers to the interpretation of
referents as female or male in experimental conditions, what happens in inter-
action when speakers design and produce utterances, that is, when their think-
ing for speaking is necessarily attuned to the categories of the language they
speak? Does interaction operate as a sort of “natural laboratory”, in Heritage
and Atkinson’s words (1984, 3), in which the cognitive role of grammatical gen-
der can be investigated empirically?
But why choose naturally occurring interaction as a sort of ‘lab’ for testing
the gender-culture-cognition complex instead of employing experimental
methods, one could reasonably ask. There are various reasons for this choice.
In section 4.3, I pointed to some of the problems detected in the experimental
tasks employed for the study of the relation between grammatical gender and
cognition. Here I add another point of criticism. In general, experiments are
useful in providing indications of the relation between language and cognition
(and personally I have also ‘experimented’ with this method, cf. Pavlidou and
Alvanoudi 2013). However, one should always keep in mind that these indi-
cations involve forms of thinking that are not necessarily related to speakers’
everyday thinking. Experiments examine what “people CAN think or expe-
rience” and not “what people usually DO think and experience”, to borrow
Hanks’ (1996, 234) words. In addition, experiments treat language and cogni-
tion as separate and autonomous phenomena, which constitute independent
(language) and dependent variables (cognition), and are detached from his-
torical and cultural context. I do not agree with this view (cf. Introduction).
As was highlighted in section 4.2.2, ‘thinking for speaking’ phenomena are
related to (or materialized through) speakers’ sociocultural practices. Οpposite
to experiments, in interaction one can examine how language and cognition
interweave with sociocultural practices. In this way, interaction offers a great
methodological advantage that experiments do not have. It ‘contextualizes’
mental phenomena into speakers’ everyday practices, and provides analysts
with access to the complex interplay among language, culture and cognition.
The second reason for investigating the relation between grammatical gen-
der and culture in interaction concerns the role that interaction plays in main-
taining culture. As Conversation Analysis (e.g. Schegloff 2006b, more on ca in
section 5.3) shows, interaction is the locus in which speakers construct various
aspects of daily social life. In Schegloff’s words:
74 chapter 5
the fundamental or primordial scene of social life is that of direct inter-

action between members of a social species, typically ones who are
physically co-present. For humans, talking in interaction appears to be
a distinctive form of this primary constituent of social life, and ordinary
conversation is very likely the basic form of organization for talk-in-
interaction. Conversational interaction may then be thought of as a form
of social organization through which the work of the constitutive institu-
tions of societies gets done—institutions such as the economy, the polity,
the family, socialization, etc. (Schegloff 1996a, 4)
According to ca, sociocultural categories, such as gender or power, are pro-

duced by interlocutors’ practices as sorts of “doing” at the micro-level of inter-
action, as outcomes of joint coordinated actions that interlocutors accomplish
in interaction. Therefore, interaction can be an appropriate locus for the inves-
tigation of the relation between grammatical gender and the construction of
gender as a social category.
Given the above, the research question I explore empirically in the rest of
this book is the following: how does the interrelation between the cultural and
cognitive aspects of grammatical gender in person reference manifest itself in
interaction? More specifically:
i) does interaction provide indications of the role of grammatical gender in

guiding speakers to interpret referent(s) as female or male, and
ii) how does the use of grammatical gender contribute to the construction of
sociocultural gender?
These questions bring the relation of grammatical gender to person reference

in interaction into the spotlight of this study. Person reference in interaction is
examined in the next section.
5.2 Grammatical Gender and Person Reference in Interaction
How does person reference relate to the use of grammatical gender? Do lin-
guistic items marked by grammatical gender merely identify individuals or do
they display more complex functions, because they codify referent’s sex in a
compulsory and systematic way? In this section, I examine these questions
and I show that the interrelation between the social and cognitive dimensions
of grammatical gender manifests itself in interaction through speakers’ covert
assumptions about referent’s sex as an aspect of social context.
When speakers refer to person, time, place or object, they use linguistic forms
that will enable hearers to identify the specific person, time, place or object
that speakers intend to be identified by hearers. According to Enfield (2012,
433), “when we say that a speaker makes reference to something in interaction,
we mean that the speaker establishes or maintains a communicative focus on
some entity, usually in order to say something about it”. Speakers have access to
rich lexical resources for making reference. For example, in Greek speakers can
accomplish person reference by employing various lexical means. They can
use proper names, e.g. Αγγελική [anɟelikí] ‘Angeliki.fem’, definite noun phrases,
e.g. ο γιατρός [o ʝatrós] ‘the.masc doctor.masc’, indefinite noun phrases, e.g.
μία γυναίκα [mía ʝinéka] ‘a.fem woman.fem’, demonstratives, e.g. εκείνη [ecíni]
‘that.fem’, vocatives, e.g. κυρία [ciría] ‘lady/Madam.fem’, or first, second and
third person pronouns and verbs, which point to the speaker, to the addressee
or to a person who is neither speaker or addressee in the speech event, e.g.
εγώ [eɣó] ‘I’, εσύ [esí] ‘you’, αυτός/αυτή [aftós/aftí] ‘he/she’, and είμαι [íme] ‘cop.
pres.1sg’, είσαι [íse] ‘cop.pres.2sg’, είναι [íne] ‘cop.pres.3sg’.
The selection of a referential formulation among various alternative ones
in interaction is a contextually bound and cognitively rich phenomenon. As
Hanks (2007) argues, person reference involves a practice of individuating a
referent under a perspective, which presupposes and activates background
knowledge, and should be theorized as a sort of construal. In his words:
In formulating a person with a descriptor, a name or a participant deic-

tic, the speaker construes the person in a certain way. [. . .] Construal
unavoidably engages both the local context of utterance and the back-
ground of social common sense. (Hanks 2007, 149)
Person reference is related to indexicality, that is, the broader phenomenon of

“contextual dependency” of utterances (Levinson 2004, 98). The role of con-
text is made clear in deictic reference. For instance, in the following utterance
(example 43) the speaker refers to the addressee, by employing the second per-
son singular pronoun and verb. These linguistic items mark person deixis, that
is, they encode participants’ roles in the speech event (Huang 2007, 136).
(43) Εσύ έλα εδώ.

esí éla eðό
you.sg come.2sg.imperat here
‘You, come here.’
76 chapter 5
Here person reference presupposes knowledge of the ‘here-and-now’ of the

speech event, e.g. who the addressee is, and some monitoring of the physical
gestures the speaker might use, e.g. finger pointing towards the referent. In
other words, reference cannot be achieved without contextual support.
Knowledge about the context of utterance is also important for non-deictic
reference. Imagine that a waiter in a restaurant produces the following
utterance, addressing another waiter (example 44).
(44) Το έξι ζήτησε λογαριασμό.

to éksi zítise loɣarʝazmό
the.nom.neuter six ask.for.3sg.past bill.acc.masc
‘(Table number) six asked for the bill’.
The other waiter is able to identify the referent introduced by the definite noun
phrase το έξι, because in restaurants waiters usually refer to customers via the
number of the table where the latter sit.
As Enfield (2012, 434) notes, the selection of a referential formulation is
based on “factors specific to the speech event, including who the speaker is,
who s/he takes the addressee to be, what the relationship is between the two,
and what the speaker’s communicative purpose is—that is, the social action s/
he wants to produce”. For instance, imagine that I ask a woman sitting next to
me on a bus to help me with the time (example 45):
(45) Κυρία, μπορείτε σας παρακαλώ

ciría boríte sas parakalό
madam.fem can.2pl.pres you.pl.acc please.1sg.pres
να μου πείτε τι
na mu píte ti
subj i.gen tell.2pl.pfv what
ώρα είναι?
όra íne
time.nom.fem is
‘Madam, can you please tell me what time it is?’
With this utterance I request some bit of information. Because I do not know
the woman, I use the address form κυρία, which encodes social distance
between me—the speaker—and the addressee. Plus, because the request is a
face-threatening act (Brown and Levinson 1987), I employ this specific address
term as a politeness strategy, together with indirecteness and second person
plural verbs, which also index social distance between speaker and addressee.
Thus, my selection of the specific referential formulation is shaped by the
social relationship I hold with the third person to be addressed and the com-
municative purpose I have.
This address term does not simply individuate and identify a referent. It also
encodes information about the referent. More specifically, it indexes the social
relation between speaker and referent, and because it is a grammatically femi-
nine noun, it indexes referent’s feminine gender (cf. chapter 3). This aspect of
person reference in interaction will be the focus of this section.
The complex phenomenon of person reference in interaction is examined
by a number of studies on person reference in interaction included in the vol-
ume edited by Enfield and Stivers in 2007. More specifically, the studies con-
ducted by Brown (2007), Enfield (2007), Hanks (2007) and Haviland (2007)
show that reference to third person in interaction is functionally complex,
because it combines, in Hanks’ words (2007, 154), “the referential effect of pick-
ing up the individual” with “the indexical effect” of displaying certain social
information about the individual. In choosing some referential formulation
speakers do more things than just identifying a referent.
For example, Lao speakers employ pragmatically unmarked, or default, and
pragmatically marked formulations for initial third person reference in interac-
tion (Enfield 2007). Default formulations consist of the person’s first name plus
a prefix that appropriately denotes the referent’s social position relative to the
speaker. Marked formulations consist of the name without the prefix or with a
prefix that does not conform to social relations. Default formulations involve
minimum cognitive effort from the part of the speaker and the addressee(s),
and are designed to be interpreted with no special attention given to the man-
ner of formulation selected. In contrast, marked formulations are departures
from the default, which show speaker’s special effort to accomplish some-
thing more than merely identifying the referent. These formulations invite the
addressee to interpret the ‘special’ thing that the speaker tries to accomplish.
However, as Enfield (2007) shows, even default formulations in Lao interaction
always do more than just referring to person. They encode a person’s hierarchi-
cal position relative to others in Lao society, and, thus, “make publicly overt
and thereby instantiate and stabilize cultural values about persons and their
social relations” (Enfield 2007, 119). In a conversation, two male Lao speak-
ers use the proper name tia together with the prefix bak2 to refer to a man
(Enfield 2007, 105). The prefix in this case is a social deictic element indexing
familiarity and lack of respect. The information of familiarity is not relevant
for the interactional business in which participants are engaged. However, this
78 chapter 5
information is made available in Lao interaction, whenever it is morphologically

marked, and constitutes part of the routine meanings produced in interaction
that usually remain unnoticed by participants.
Similar cases are observed in third person reference in Tzeltal (Brown 2007)
and Yucatec Maya (Hanks 2007), both languages spoken in Mexico. Default for-
mulations for third person reference in Tzeltal include kinship terms or names,
among others, which depend on the social identity of the referent and her/
his relationship to speaker, addressee or other co-participants. For example,
the default formulation for referring to a person not kin related to speaker or
addressee and socially distant may be jme’tik/jtatik plus name, which indexes
social distance and respect, or me’el ‘old woman’/mammal ‘old man’ plus name,
which indexes neutrality or lack of respect (Brown 2007). This social informa-
tion is part of the meanings that, in Brown’s (2007, 199) words, are “routine,
unremarkable in the context”. In a similar way, in Yucatec Maya kinship terms,
such as in-tiyo ‘my uncle’ and in-fáamilya ‘my wife’, are default formulations for
initial third person reference (Hanks 2007). Besides selecting and identifying a
referent, these terms also encode referent’s kin relation to speaker, and presup-
pose and activate participants’ knowledge on kin relations. Haviland (2007, 232)
describes the multiple indexical projections associated with expressions refer-
ring to person via the term ‘referring dupliciter’ (or ‘multipliciter’). Contrary to
Schegloff’s (1996b; 2007c) idea that certain formulations in interaction serve as
‘referring simpliciter’, that is, doing referring and nothing else, Haviland (2007,
232) claims that “a socially embedded act of person reference is never simple”.
When participants refer to person, they “always ‘do’ more than refer” (Haviland
2007, 249); they perform some sort of indexical work. For instance, speakers of
Zinacantec Tzotzil have different nicknames by which they are known in dif-
ferent social settings (Haviland 2007). When a speaker refers to a third person
via a nickname, s/he uses a nickname that is recognized successfully by the
addressee. In this case, nicknames index speaker’s and addressee’s relation to
third person, for example they indicate whether third person and addressee
belong in the same social group or not.
Referring expressions in Greek may index various social meanings (e.g. see
Makri-Tsilipakou 1984 for a study of address terms in Greek) and they mark
or index referent’s gender, because they are inflected for grammatical gender.
For instance, in example 45 the feminine noun selects a specific third person
and at the same time it construes this person as female/woman. In this way,
the information of referent’s gender is made available in interaction. Person
reference in Greek is functionally complex, because referring expressions do
not simply select and identify a third person. They always index referent’s gen-
der, that is, they display a specific indexical effect, in Hanks’ (2007) terms, and,
thus, operate as referring dupliciter, in Haviland’s (2007) terms. As I suggested

in chapter 3, following Ochs’ (1992) work, linguistic items marked by grammat-
ical gender index gender referentially, directly and exclusively. In the following
section, I examine the relation between grammatical gender and indexicality
in more detail.
5.2.1 Indexing Gender: The Role of Grammatical Gender

In arguing that certain linguistic items index gender referentially, Ochs draws
on Silverstein’s (1976) work on referential and non-referential indexicality. As
Silverstein argues, language does not only represent reality but also indexes,
and, thus, presupposes and creates aspects of social context, such as identities
and social relations. Linguistic items may index aspects of context referentially
or non-referentially. Referential indexes index an aspect of context via their ref-
erential content, that is, via the description of events or states of affairs, and
contribute to referential speech event. For example, referential indexes may
include elements of space and person deixis, such as the English demonstra-
tives here and there and the personal pronouns he and she. In contrast, non-
referential indexes “signal a particular value of one or more contextual variable”
and do not contribute to accomplishing reference (Silverstein 1976, 29). Non-
referential indexes usually include social deictic elements, which index social
identities and relations. For example, such elements are the verb particle -s in
Koasati which indexes speaker’s female sex, the politeness particles khráb and
khá in Thai which are used by men and women only respectively, second per-
son singular and plural pronouns (also known as the T/V distinction), such as
εσύ [esí] and εσείς [esís] in Greek which index social proximity and distance
between speaker and addressee respectively, and the mother-in-law language
in Dyirbal, that is, an ‘avoidance’ language used in the presence of a mother-
in-law with whom close social contact is taboo (examples found in Dixon 1972;
1980; 2002; Levinson 1983; Silverstein 1976). The difference between referential
and non-referential indexes is illustrated with the following example. In replac-
ing the second person singular pronoun εσύ with the second person plural
pronoun εσείς (i.e. non-referential indexes), the information on social distance
between speaker and addressee indexed by the second person singular and plu-
ral pronoun changes. Yet, the referent remains the same. In contrast, in replac-
ing the proximal demonstrative εδώ [eðό] ‘here’ with the distal demonstrative
εκεί [ecí] ‘there’ (i.e. referential indexes), the place indicated by the demonstra-
tive changes. In other words, referential indexes are truth conditional.
Indexes are further classified on the basis of indexical presupposition
and indexical creativity (Silverstein 1976). Referential indexes are usually
associated with indexical presupposition, that is, they presuppose the
80 chapter 5
(physical or cognitive) existence of particular aspects of context. For example,

the interpretation of the deictic expression αυτή η καρέκλα [aftí i karékla] ‘this
chair’ presupposes a real object in the physical context to which the speaker
refers, and which can be identified as such by the addressee. Non-referential
indexes are usually associated with indexical creativity, that is, they construct
aspects of social context or in Silverstein’s (1976, 34) words, they “make explicit
and overt the parameters of structure of the ongoing events”. For example,
the second person plural pronoun creates the social relation of distance or
hierarchy between speaker and addressee as an aspect of context.1
Indexical presupposition is a relevant aspect for linguistic items marked by
grammatical gender in Greek, given the relation of these items with referen-
tial indexicality. Items marked by grammatical gender, which either refer to or
describe person, constitute referential indexes of gender, that is, they index
referent’s feminine or masculine gender and presuppose it as an aspect of con-
text, i.e. they presuppose the existence of a female or male person (a woman
or a man).2
This indexical effect is produced every time grammatical gender is used
in interaction. In example 46, the grammatically feminine participle indexes
speaker’s feminine gender and presupposes that the referent, i.e. the speaker,
is a woman.
(46) Eίμαι ενθουσιασμένη.

íme enθusiazméni
cop.1sg.pres excited.part.nom.fem
‘I am excited.’
1 Referential and non-referential indexes may also be related to indexical creativity and
presupposition respectively (Silverstein 1976). For example, personal pronouns I and you
(i.e. referential indexes) presuppose a speaker and hearer at context and they also create
these roles as aspects of context, while the mother-in-law vocabulary in Dyirbal (i.e. non-
referential indexes) presupposes the existence of relatives in context.
2 Linguistic items marked by grammatical gender appear to be similar to social deictic items,
due to the fact that in both cases a bit of ‘social’ information about the context is grammati-
cized. However, these items are not identical. According to Levinson (1983, 127–129), social
deixis is related with conventional implicature, i.e. non-truth-conditional inference that is
conventionally associated with certain lexical items or linguistic constructions. In contrast,
referential indexes are truth-conditional. For example, if I replace second person singular
with second person plural in the following sentence Είσαι ψηλός/Είστε ψηλός [íse psilós/íste
psilós] ‘You are.sg tall.masc/You are.pl tall.masc’, truth conditions remain intact, but the
inference about the social relation between speakers and addressee changes. In contrast,
replacing the masculine adjective ψηλός with the feminine adjective ψηλή affects truth condi-
tions, because the referent changes (a person can be either female or male).
In example 47, the noun phrase with the masculine head noun and the mas-
culine modifier identifies a third person and at the same time presupposes
masculine gender as part of referent’s identity, that is, it combines referential
effect with indexical effect.
(47) Ο μαθητής ήρθε.

o maθitís írθe
the.nom.masc student.nom.masc come.3sg.past
‘The male student came.’
In both cases, referent’s gender is codified as part of the assumptions that

interlocutors share about context, or as part of the meanings produced in
interaction in covert, automatic and compulsory ways that usually remain
unremarkable (cf. section 5.2). The relation of grammatical gender to refer-
ential indexicality foregrounds the role of context in the complex interplay
between grammatical gender, culture and cognition. The cultural and cogni-
tive aspects of grammatical gender are not abstract concepts, separated from
language use. They are grounded in the production of utterances in everyday
interaction, when speakers think for speaking, and they manifest themselves
in interaction through covert assumptions about referents’ gender as an aspect
of social context.
One of the covert assumptions associated with the use of grammatical
gender in interaction is the stereotype of man as the norm and the asymmet-
ric relation between women and men. In the next section, I show how this
assumption is linked with strategies that speakers use in interaction to inter-
pret what is being said and done.
5.2.2 The Generic Use of the Masculine Gender and Default Inferences
Person reference in everyday interaction in English and also in other languages
is organized on the basis of two principles or preferences: recognition and
minimization (Sacks and Schegloff 1979; Stivers, Enfield, and Levinson 2007).
According to the principle of recognition, speakers use a recognitional
reference form that the recipient will know and can use to identify a person.
According to the principle of minimization, speakers use a single form, whether
a name, a description, etc. to refer to person. Levinson (2007) argues that the
principle of minimization is part of the more general principle of economy,
according to which speakers minimize the expressive means in referring to
person. For example, speakers use a single referring expression (minimization)
or a name rather than a description. The generic use of the masculine seems
to align with the principle of economy in person reference in interaction. In
using the masculine forms alone for achieving reference to males plus females,
82 chapter 5
speakers minimize their expressive means and invite their recipients to make
the necessary inference of human/universal. Therefore, the masculine gender
seems to be related with the interpretative strategies that participants employ
in interaction.
As Enfield (2007) shows, there is a relation between interpretative strate-
gies and person reference in Lao interaction. For example, default referential
formulations are designed to pass unnoticed by participants. They also allow
participants to recognize cases in which speakers deviate from the norm and
use pragmatically marked formulations, trying to express additional meanings
besides referring to a specific person. Inferences or conversational implica-
tures play an important role in understanding speakers’ intentions in commu-
nication, according to Grice (1989).
Levinson (2000) explores the role of inferences in communication by focus-
ing on generalized conversational implicatures, that is, implicatures whose
interpretation does not depend on context. More specifically, Levinson argues
that generalized conversational implicatures are associated with default infer-
ences or preferred interpretations in communication. These preferred inter-
pretations include various phenomena in communication, such as preference
structure in interaction, presuppositions, and illocutionary force. According to
Levinson, generalized conversational implicatures are produced on the basis
of three neo-Gricean principles: Quantity (do not say less than is required),
Informativeness (do not say more than is required), and Manner (do not use
a marked expression without reason). The principle of Informativeness is of
special interest here. This principle consists of the maxim of minimization
(speaker’s maxim) and the rule of enrichment (recipient’s corollary) (Levinson
2000, 114). The maxim of minimization enjoins the speaker to say as little as
necessary, that is, produce the minimal linguistic information sufficient to
achieve her/his communicational ends. The rule of enrichment dictates the
recipient to infer that what is generally said is stereotypically and specifically
exemplified, that is, to amplify the informational content of the speaker’s utter-
ance and interpret it according to stereotypical assumptions. For example, the
utterance ‘John said “Hello” to the secretary and then he smiled’ produces
the generalized conversational implicature ‘John said “Hello” to the female sec-
retary and then John smiled’, based on the knowledge that interlocutors share
about the occupation of secretary as stereotypically feminine (Levinson 2000,
117). In other words, the semantically general expression secretary implicates
the semantically specific interpretation ‘female secretary’.
Levinson’s approach sheds light on the way in which the generic use of the
masculine operates in communication. The generic use of the masculine seems
to be related with the systematic pragmatic inferences that interlocutors make
in communication. Interlocutors share some general expectations about how

language is used. For example, they expect that not all meaning is explicitly
articulated and much is inferred according to their common knowledge about
the sociocultural world. The use of the masculine gender for generic reference
is based on such an expectation. When speakers employ masculine items to
identify a group of male and female or gender-indefinite referents, they pro-
duce the minimal linguistic information sufficient to achieve their commu-
nicational ends. At the same time, they invite recipients to infer reference to
male and female or gender-indefinite referents based on the stereotypic (met-
onymic) association of male sex with the category of human/universal (cf.
section 4.2.1). Thus, the inference of human/universal is a default inference or
preferred interpretation, in Levinson’s (2000) terms. This inference is produced
in communication in systematic and automatic ways, based on the sociocul-
tural stereotype of man as the norm, and it aligns with the general principle of
economy in person reference in interaction.
However, the inference of human/universal via male sex does not form a
generalized conventional implicature for two reasons. First, if the inference
of human/universal was a generalized conventional implicature, it would be
‘isomorphic’ with the utterance. This means that each term in the implicature
would have a denotation that would be “a subset of the denotations of the cor-
responding expressions” in the utterance (Levinson 2000, 115). For instance,
the implicature ‘female secretary’ has a denotation that is a subset of the deno-
tations of the expression secretary. This isomorphy is not found in the case of
the generic use of the masculine: the masculine expression codifies male sex
but the inference produced consists of male plus female—or non-exclusively
male—referents. In other words, the generic use of the masculine does not
involve a semantically specific interpretation emerging from a semantically
general expression. Second, contrary to generalized conversational implica-
tures that do not require contextual conditions in order to arise, the inference
of human/universal via male sex is context-dependent (cf. section 4.2.1). For
example, when reference is done to stereotypically masculine occupations,
referents may be interpreted as exclusively male, and not as male plus female.
So far, I have showed that the use of grammatical gender in interaction is
linked with covert assumptions that interlocutors share about referent’s gen-
der as an aspect of social context and man as the norm. These assumptions
are the ‘vehicle’ through which the interrelation between the social and cogni-
tive dimensions of grammatical gender manifests itself in interaction. In the
next section, I turn to the methodological tools that will allow us to examine
empirically grammatical gender in interaction.
84 chapter 5
5.3 Approaching Grammatical Gender in Interaction through the Lens

of Conversation Analysis
For the empirical investigation of grammatical gender in interaction I choose

Conversation Analysis. The reasons for this choice are discussed in the
following.
ca is a method for analyzing language use and social interaction. The dis-
tinctive features of ca are summarized by Stivers and Sidnell (2012, 2) as fol-
lows. ca assumes that social interaction “is orderly at a minute level of detail”,
and “this orderliness is conceived of as the product of shared methods of rea-
soning and action to which all competent social interactants attend”. ca exam-
ines the practices, actions, activities and the overall structure of interactions,
by analyzing records of spontaneous, naturally occuring interaction, which are
transcribed in full detail. It is an inductive qualitative method that is based on
a case-by-case analysis, and seeks to discover patterns using “a combination of
distributional regularities, commonalities in contexts of use, participant ori-
entations and deviant case analysis”. Usually conversation analysts examine
a number of cases that allows them to make generalizations about the orga-
nization of a particular phenomenon, and the practices through which this
phenomenon is produced. Yet, ca can also be based on the analysis of a single
episode of interaction, as Schegloff (1987) shows. According to Schegloff (1987,
102), this mode of analysis is justified by the assumption that “social action
done through talk is organized and orderly not, and not only, as a matter of
rule or as a statistical regularity, but on a case by case, action by action, basis”.
ca has its roots in the sociological work of Goffman (1983) and Garfinkel
(1967). As Stivers and Sidnell (2012, 3) point out, these roots are visible in ca’s
founding idea that an “institutionalized ‘interaction order’ (Goffman 1983),
comprising shared methods of reasoning and action (Garfinkel 1967), forms
the foundation of ordinary action in the social world”. According to Garfinkel
(1967), these methods or practices constitute commonsense knowledge3 that
participants share, and serve as the background of everyday social life that is
taken for granted, or, in other words, what is ‘seen’ but remains ‘unnoticed’.
3 Commonsense knowledge serves as an ‘umbrella’ category which includes norms for under-
standing behavior, language use and knowledge about the sociocultural world (e.g. social
institutions and relations) that is taken for granted (Garfinkel 1967). This concept resembles
the concepts of common ground proposed by Clark (1996), and social cognition proposed by
van Dijk (2009).
Participants employ these methods to construct and sustain their socio-

cultural world. In line with this view, ca theorizes interaction as a locus of
social life (Schegloff 2006b, cf. section 5.1).
ca is not interested in what interactants “hypothetically or imaginably
understood or the procedures or constraints to which they could conceiv-
ably have been oriented” (Heritage and Atkinson 1984, 1). It rather examines
the actual conduct of participants in interaction and the procedures through
which participants produce their own behavior and understand and deal
with the behavior of others. In this sense, ca employs an ‘emic’ perspective
on interaction. Instead of applying pre-existing analytic interpretations on
the data examined, ca aims at describing aspects of interaction in the way
in which they have been recognized by and reacted upon by participants. In
Schegloff’s (1997, 168) words, ca aims at describing a conversational episode
“in its endogenous constitution, what it was for the parties involved in it, in
its course, as embodied and displayed in the very details of its realization”. For
this reason, interaction serves as a sort of natural laboratory, in which analysts
can examine the ordinary actions through which participants accomplish dif-
ferent aspects of their sociocultural world and ‘do’ social life (Sacks 1984). But
how are participants’ understandings of social actions uncovered and publicly
displayed in interaction?
Activities in interaction are accomplished through sequences. Sequences
are courses of action implemented through talk, which consist of adjacency
pairs composed of two turns. These turns are produced by different speakers,
and are adjacently placed and relatively ordered as first pair parts and second
pair parts (Schegloff 2007a, 13–21). Turns are composed by turn constructional
units (TCUs), that is, clauses, phrases and lexical items which constitute at least
one recognizable action in context. First pair parts initiate some exchange, e.g.
question, request, offer, invitation, and second pair parts are responsive to the
action of a prior turn, e.g. answer, reject, accept, decline. The notions of ‘first-
ness’ and ‘secondness’ refer to the design features of these turns, rather than to
their order. First pair parts project the relevance of specific second pair parts,
that is, they set powerful constraints on what the recipient should do and
on how the action accomplished by the recipient should be understood. For
example, if a question (first pair part) is followed by an answer (second pair
part), the answer is heard as responsive to the turn that preceded. If an answer
is not produced, its non-occurence can be heard as ‘non-answering’ and can be
interpreted as “relative absence” (Schegloff 2007a, 20).
In this sense, participants’ understandings depend on the relationship
of adjacency or nextness that characterizes the organization of talk-in-
86 chapter 5
interaction. As Schegloff (2007a, 14) notes, any element in talk-in-interaction

comes next after a prior one. For example, in a tcu a word comes next after a
word, in a multi-unit turn a tcu comes next after a tcu, in a sequence a turn
comes next after a turn. This moving from some element to a hearably-next-
one with nothing intervening is described by Scegloff (2007a, 15) as “progres-
sivity”. In his words:
Should something intervene between some element and what is hear-

able as a/the next one due—should something violate or interfere with
their contiguity, whether next sound, next word, or next turn—it will be
heard as qualifying the progressivity of the talk, and will be examined
for its import, for what understanding should be accorded it. Each next
element of such a progression can be inspected by participants to find
how it reaffirms the understanding-so-far of what has preceded, or favors
one or more of the several such understandings that are being entertained,
or how it requires reconfiguration of that understanding. (Schegloff
2007a, 15)
Therefore, next turns are understood by co-participants as displaying their

speaker’s understanding of the just-prior turn (next-turn proof procedure,
Sacks, Schegloff, and Jefferson 1974) and embodying an action responsive
to the just-prior turn so understood (Schegloff 2007a, 15). For example, the
answer Ναι. [né] ‘Yes.’ to the question Ήρθε o Nίκος; [írθe o níkos] ‘Did Nikos
come?’ displays speaker’s understanding of prior turn as a question. Thus, the
organization of talk-in-interaction allows participants to build joint or inter-
subjective understandings, which can be maintained, monitored and repaired,
if necessary (Sidnell 2010, 12).
To sum up, interaction provides analysts with access to participants’ inter-
subjective understandings. As Heritage (1984, 259) claims, “a context of pub-
licly displayed and continuously up-dated intersubjective understandings is
systematically sustained” through the organization of interaction on a turn-
by-turn basis. The fact that this context of understandings is made available for
analysis constitutes a major methodological advantage when it comes to the
study of the relation between language and cognition in interaction. As Sidnell
and Enfield (2012, 303–304) argue, the methods employed by participants in
interaction for building and sustaining their intersubjective understandings
can be employed by analysts as a metholodological tool which resolves “one
of the most difficult methodological problems of all relativity arguments (. . .)
how to show that a grammatical or lexical peculiarity has nonlinguistic (i.e.
cognitive, cultural, action relevant) consequences.”4 In a similar vein, Levinson

(2006, 86) argues that ca is “more cognitive” compared to other theories of
discourse, because it takes participants’ orientations and interpretations into
account and, thus, it is closer to their “mental world”. As Hougaard (2008, 180)
emphasizes, ca focuses not on what participants “may mean, think, believe or
intend” but on what they actually do in interaction, that is, what social actions
they accomplish and what categories they orient to. Based on the ideas of the
researchers mentioned above, I suggest that ca’s ‘emic’ perspective constitutes
a potential ‘cognitive’ perspective, which provides analysts with access to par-
ticipants’ intersubjective understandings.
Conversation analytic work has addressed various aspects of cognition
through the notions of commonsense knowledge, shared understanding
and socially shared cognition (Schegloff 1991), among others, as Potter and
Molder (2005, 23–28) highlight. More specifically, Schegloff (1991) points to
certain areas of inquiry that one can undertake in order to explore the rela-
tion between language and cognition. These areas include person reference
and its relation to participants’ commonsense knowledge, and the practices
employed by participants for repairing trouble in interaction and maintaining
their intersubjective understandings.5 In the next chapter, I show that both
these areas are involved in the empirical study of the relation between gram-
matical gender and cognition in interaction.
Overall, the preceding discussion points to two methodological advantages
that ca displays with respect to the research question examined in this book.
On the one hand, ca provides analysts with access to speakers’ cognition in
interaction. As was shown in chapter 3, speakers’ cognition at the time of
speaking is mediated by language (Slobin 1996; 2003). Therefore, part of the
mental representations that are constructed when participants interact with
each other is expected to be language-specific and associated with speakers’
thinking for speaking. Based on this hypothesis, I assume that indications of
the cognitive role of grammatical gender can be searched and possibly found
in the data of interaction, that is, in the context of participants’ publicly
4 Sidnell and Enfield (2012) formulate a new version of linguistic relativity, according to which
lexico-grammatical differences among languages correlate with differences in the kinds of
social actions that participants can achieve in social interaction.
5 Schegloff (2006a) criticizes the dominant approach within cognitive science as focusing on
individual mind and separating cognitive processes from interaction. He argues that the
study of language and cognition should be based on empirical observations about interac-
tion, which constitutes the locus of human sociality.
88 chapter 5
displayed intersubjective understandings. On the other hand, as ca has shown,

interaction is the locus in which speakers construct various aspects of daily
social life. Gender is one of these aspects. By analyzing data of naturally occur-
ring interaction, I aim to show how the use of grammatical gender contributes
to the construction of sociocultural gender as an aspect of the social world that
participants jointly construct in interaction.
5.3.1 Gender Membership Categories and Grammatical Gender

Linguistic items marked by grammatical gender are linked with the practices
that participants employ for categorizing selves and others as women or men
in interaction. In particular, because grammatically masculine or feminine
items ascribe female or male sex to referents and index their feminine or mas-
culine gender, they constitute gender membership categories. The notion of
membership category was introduced by Sacks (1972a; 1972b) to describe the
practices that participants employ for categorizing persons as members of
specific categories. Membership categories together with the rules of their use
form what Sacks (1972b, 332) calls ‘membership categorization device’.
More specifically, membership categories are classifications of various kinds,
such as γυναίκα [ʝinéka] ‘woman.fem’/άντρας [ándras] ‘man.masc’, νέος [néos]
‘young.masc’/νέα [néa] ‘young.fem’, γέρος [ʝéros] ‘old.masc’/γριά [ɣriá] ‘old.
fem’, χριστιανός [xristçanós] ‘christian.masc’/χριστιανή [xristçaní] ‘christian.
fem’, μουσουλμάνος [musulmános] ‘muslim.masc’/μουσουλμάνα [musulmána]
‘muslim.fem’, καθηγητής [kaθiʝitís] ‘professor.masc’/καθηγήτρια [kaθiʝítria]
‘professor.fem’, αριστερός [aristerós] ‘leftist.masc’/αριστερή [aristerí] ‘leftist.
fem’, δεξιός [ðeksiós] ‘rightist.masc’/δεξιά [ðeksiá] ‘rightist.fem’, Έλληνας [éli-
nas] ‘Greek.masc’/Ελληνίδα [eliníða] ‘Greek.fem’, etc. They are organized into
collections of categories, such as sex, age, religion, political views, etc. Some
collections, like sex or age, are appropriate for categorizing any member of any
undefined population. Yet, other collections apply only to members of specific
populations, for example the category τερματοφύλακας [termatofílakas] ‘goalie.
masc’ is used to describe only members of a football team. The same person
can be categorized on the basis of different categories, e.g. as woman (sex),
young (age), leftist (political views), etc.
Membership categories are important for what participants do in interaction
because of the special features they have (Sacks 1972a; 1972b; 1995; Schegloff
2007b). Following Schegloff (2007b), these features can be summarized as fol-
lowing. Membership categories are inference-rich, that is, they organize—as
a sort of “store house” (Schegloff 2007b, 469)—commonsense knowledge
that speakers share about identities, roles and behaviors of the members of a
community, and provide access to this knowledge. If a person is taken to be a
member of a specific category, s/he is considered to be a presumptive repre-

sentative of this category, that is, all inferences produced about the members
of the category will apply to that specific referent as well. Membership cat-
egories are also protected against induction, that is, if a member of a category
deviates from what is considered to be typical behavior for the members of
the category, knowledge about the category is not revised and this member is
treated as an exception.
Finally, membership categories are associated with “category-bound activi-
ties” (Sacks 1972b, 335). These activities or forms of conduct are usually asso-
ciated with the members of a category, based on speakers’ knowledge about
typical or specially characteristic features that members of a category have.
A well-known example given by Sacks (1972b, 330) to illustrate the notion of
category-bound activity is the phrase Τhe baby cried. The mommy picked it up.
Referring to crying can activate the category ‘baby’ as a stage of life, because
crying is specially characteristic of babies. As Schegloff (2007b, 470) notes in his
tutorial article on membership categorization, “the doing of a category-bound
action can introduce into a scene or an occasion the relevance of the category
to which that action is bound”. This is related with the hearer’s and viewer’s
maxim formulated by Sacks (1972b). According to these maxims, participants
are expected to understand the referring to or doing of a category-bound activ-
ity as such, that is, as typical of members of this category.
When it comes to Greek conversation, speakers have a variety of resources
or practices for categorizing themselves and others as women or men. Gender
membership categories in Greek include nouns in which referent’s sex is
lexically marked, e.g. γυναίκα [ʝinéka] ‘woman’/άντρας [ándras] ‘man’, κορίτσι
[korít͡si] ‘girl’/αγόρι [aɣόri] ‘boy’. They also include all linguistic items in which
referent’s sex is grammaticized, such as nouns and articles (ο φοιτητής [o fititís]
‘the.masc student.masc’/η φοιτήτρια [i fitítria] ‘the.fem student.fem), pro-
nouns (αυτός [aftόs] ‘he’/αυτή [aftí] ‘she’), adjectives (χαρούμενος [xarúmenos]
‘happy.masc’/χαρούμενη [xarúmeni] ‘happy.fem’), and participles (δεμένος
[ðeménos] ‘bound.masc’/δεμένη [ðeméni] ‘bound.fem’). In the case of the lat-
ter membership categories, the gendered categorization of referents may be
combined with additional sorts of categorization which are related to the lexi-
cal morpheme, such as age (νέος [néos] ‘young.masc’/νέα [néa] ‘young.fem’)
or religion ( χριστιανός [xristçanόs] ‘christian.masc’/χριστιανή [xristçaní] ‘chris-
tian.fem’).
More specifically, linguistic items marked by grammatical gender are suit-
able for the categorization of any member of an undefined population, they
categorize referents as typical/most representative members of the category
of male or female sex, organize participants’ knowledge about the dominant
90 chapter 5
gender order and are inference rich. I suggest that these inferences include
speakers’ covert assumptions about referent’s gender, given that items marked
by grammatical gender operate as referential indexes (cf. section 5.2.1).
Inferences also include speakers’ knowledge about the stereotypic association
of man with the norm, and gender hierarchy (cf. section 5.2.2). Defining items
marked by grammatical gender as gender membership categories foregrounds
the contextually grounded nature of these items in interaction. It associates
grammatical gender with practices that participants employ in interaction in
order to accomplish a particular aspect of social life, gender.
As a number of conversation analytic studies (e.g. Kitzinger 2007; Stockill
and Kitzinger 2007)6 show, the deployment of gender membership categories
in English conversation (e.g. woman, man, male, girl) is not necessarily related
to participants’ orientation to gender. Stockill and Kitzinger (2007, 231) argue
that although gender membership categories are gendered linguistically, they
are not always “gendered interactionally”, that is, they are not always used “pri-
marily to foreground gender”. Yet, they make gender available in talk. However,
as Kitzinger (2000; 2005) argues, various aspects of social life are often repro-
duced in interaction in tacit ways, without participants explicitly orienting to
them. For example, the norm of heterosexuality is reproduced as an unques-
tioned assumption in English conversation when participants accomplish
social actions, which are not related to gender (Kitzinger 2005). The gender
membership categories husband/wife denote specific roles in the heterosexual
kin system, locate referents in the heterosexual order and, thus, make the infer-
ence of heterosexuality available in interaction. Moreover, as Kitzinger (2005)
points out, participants’ lack of orientation to the norm of heterosexuality is a
sort of ‘required’ condition ensuring that this norm is reproduced smoothly as
part of the taken-for-granted world. In this way, heterosexuality becomes part
of the tacit presuppositions of everyday life that participants ‘see’ but do not
notice, in Garfinkel’s (1967) terms.
In general, ca favors a mode of analysis that does not impose a-priori
social categories on the data and is based on aspects of interaction to which
participants explicitly orient. According to Schegloff (1997), the analysis of
conversational episodes should examine only those categories that are shown
to be relevant for participants, that is, procedurally consequential for what
participants do in interaction. Yet, this analytic choice, that Enfield (2007, 113)
describes as the ‘Members-Only Filter’, limits the scope of analysis. Consider,
6 Kitzinger (2000) refers to conversation analytic studies on gender via the term ‘feminist con-
versation analysis’. Speer and Stokoe (2011) provide a review of the issues examined by these
studies.
for example, person reference in interaction (cf. section 5.2). As Enfield (2007)
shows, default formulations encode information about the social relation
between speaker and third person. Regardless of whether participants
recognize these relations as relevant for what they do in interaction or not,
these relations are made available in interaction and constitute part of the
routine meanings produced in interaction. Following up these arguments,
one should expect that a lot of interesting things may be happening when
“nothing special is happening” in interaction, in Kitzinger’s (2005, 259) words.
Given that power relations are usually reproduced in interaction in tacit and
hidden ways, Kitzinger (2005) suggests that analysts should treat participants’
everyday world as problematic and explore the assumptions on which this
world is based. In her words, we as analysts might ask what is happening:
when the second is in a preferred relation to the first; when the yes—no
question is followed by a yes—no answer; when the recognitional refer-
ent is recognized or the nonrecognitional referent is treated as adequate;
[. . .] when presumed ordinary experiences are treated as ordinary—what
is happening then, how is that done, and what kind of world must we be
living in that these things run off smoothly? This kind of analysis may
enable us better to understand (as activists and as conversation analysts)
the routine achievement of a taken-for-granted world (Schegloff 1986).
(Kitzinger 2005, 259)
Taking the studies discussed in this section into account, in the next chapter,
I examine empirically grammatical gender in interaction both when gender is
relevant interactionally and when it is not.
In this chapter, I discussed the theoretical and methodological tools that
will help me analyze empirically grammatical gender in interaction: indexical-
ity in relation to gender membership category. To put it briefly, linguistic items
marked by grammatical gender are gender membership categories, which are
linked with covert assumptions that participants share about referent’s gender.
In the next chapter, I apply this notion to the analysis of naturally occurring
data in Greek conversation.
CHAPTER 6
The Empirical Investigation of Grammatical

Gender in Interaction
6.1 Introduction
In this chapter, I examine empirically grammatical gender in Greek conversa-

tion. The data I analyze are generated from 39 fully transcribed audio-recorded
naturally occurring informal face-to-face conversations among friends and rel-
atives from the Corpus of Spoken Greek of the Institute of Modern Greek Studies
(see Pavlidou 2012b for a description of the features of the corpus, e.g. data col-
lection, size, discourse types included etc.). The total duration of the conver-
sations examined is 22 hours and 43 minutes, and the total number of words
is 402350 (each of the conversations examined had a minimum duration of
30 minutes).
The present chapter is divided into two parts. In the first part (section 6.2),
I explore the implications of the compulsory use of grammatical gender in the
composition of turns for the construction of sociocultural gender in interac-
tion. I examine grammatical gender in individual, collective and indefinite
second person self-reference, in reference to recipient(s) and third person(s).
I analyze conversational episodes in which participants orient to their gen-
der identity as well as conversational episodes in which participants do not
foreground gender interactionally. Special attention is given to cases in which
there is no match between grammatical gender and referents’ sex, that is,
when participants employ the masculine grammatical gender in reference to
female persons only.
In the second part (section 6.3), I examine whether interaction provides
indications of the cognitive role of grammatical gender. More specifically,
I analyze cases in which the use of grammatical gender in interaction creates
problems in the interpretation of referent’s sex, which are resolved by partici-
pants through practices of repair. I show that these cases provide indications of
the cognitive role of grammatical gender. In addition, I analyze cases in which
the use of grammatical gender does not create any sort of trouble for partici-
pants, and I show that covert indications of the relation between grammatical
gender and speakers’ cognition can also be found in such cases.
Before moving on with the analysis, a few remarks need to be made about
ca’s methodological tools that will help the reader follow up the analysis

Empirical Investigation of Grammatical Gender in Interaction 93
resented in the next sections. According to ca, conversations involve people

p
taking turns at speaking. Αccording to Sacks, Schegloff and Jefferson (1974,
700–701), the turn-taking system in conversation is locally managed and party-
administered, and consists of the following features:
(1) Speaker-change recurs, or at least occurs.

(2) Overwhelmingly, one party talks at a time.
(3) Occurrances of more than one speaker at a time are common, but brief.
(4) Transitions (from one turn to a next) with no gap and no overlap are com-
mon. Together with transitions characterized by slight gap or slight over-
lap, they make up the vast majority of transitions.
(5) Turn order is not fixed, but varies.
(6) Turn size is not fixed, but varies.
(7) Length of conversation is not specified in advance.
(8) What parties say is not specified in advance.
(9) Relative distribution of turns is not specified in advance.
(10) Number of parties can vary.
(11) Talk can be continuous or discontinuous.
(12) Turn-allocation techniques are obviously used. A current speaker may
select a next speaker (as when s/he addresses a question to another
party); or parties may self-select in starting to talk.
(13) Various ‘turn-constructional units’ are employed; e.g. turns can be pro-
jectedly one word long, or they can be sentential in length.
(14) Repair mechanisms exist for dealing with turn-taking errors and viola-
tions; e.g. if two parties find themselves talking at the same time, one of
them will stop prematurely, thus repairing the trouble.
As was mentioned in section 5.3, turns consist of at least one tcu, which con-
stitutes at least one recognizable action, such as question, answer, disagree-
ment, offer, request, invitation, announcement, promise or narration. The
action done by some talk is “grounded in its position, not just its composition”
(Schegloff 2007a, 20). A speaker beginning to talk in a turn has the right and
obligation to produce one tcu, and as a speaker approaches the possible com-
pletion of a first tcu in a turn, transition to next speaker can become relevant
(Schegloff 2007a, 4). Current speaker selects next speaker by addressing her/
him “with a turn whose action requires a responsive action next—for example,
with a question that makes an answer relevant next” (Schegloff 2007a, 4). Or,
next speaker self-selects to take the next turn. First pair parts make specific
second pair parts relevant. Second pair parts are divived into preferred and
dispreferred, depending on whether they align with or distance from the first
94 chapter 6
pair parts (Schegloff 2007a, 58–63). For example, agreement is a preferred

second pair part to an assessment, while disagreement is a dispreferred first
pair part (Pomerantz 1984). Preference and dispreference refer to a structural
relationship of sequence parts and constitute “a social/interactional feature
of sequences and of orientations to them” (Schegloff 2007a, 61), rather than a
psychological notion.
One of the practices that participants employ for accomplishing social
actions in interaction is the selection of elements composing turns (Schegloff
2006b).1 As expected, the analysis of grammatical gender in interaction pre-
sented in this chapter focuses on the composition of turns. Word-selection
is key in terms of how a turn will be interpreted by recipients. As Schegloff
(2006b, 81) points out, “the elements of the talk are selected and deployed to
accomplish actions and to do so recognizably; and recipients attend to talk
to find what the speaker is doing by saying it in those words, in that way”. This
is the reason why turns in interaction are always recipient-designed (Sacks,
Schegloff, and Jefferson 1974), that is, they are designed by speakers in such
a way as to be appropriate and relevant for recipients. Wilkinson (2010, 112)
exemplifies recipient design as following: “Speakers say different things, in
different words, implementing different actions, depending on who they take
their recipient(s) to be”. For instance, in referring to a third person known to
my recipient I may select a proper name, that is, a recognitional reference form
(Schegloff 1996b, 459), following the preference of recognition (cf. section
5.2.2) and taking my recipient into account, i.e. assuming that my recipient
knows the referent and can use this formulation to identify him or her suc-
cessfully. That is, I design my turn in a way which diplays an orientation to the
specific person I address.
1 Word selection is one of the basic problems of talk in interaction that get to be examined via
ca. According to Schegloff (2006b, 71, 73, 77 and 82), the other problems of talk in interaction
are the following: i) the “turn-taking” problem, which concerns who and when should talk or
act next, and how this affects the construction and understanding of the turns or acts them-
selves, ii) the “sequence-organizational” problem, which concerns how successive turns or
actions are formed up to be ‘coherent’ with the prior one and constitute a “course of action”,
and the nature of that coherence, iii) the “trouble” problem, which concerns how partici-
pants deal with trouble in speaking, hearing, or understanding the talk or other conduct such
that the interaction does not freeze in place, intersubjectivity is maintained or restored, and
that the turn, sequence and activity can progress to possible completion, and iv) the “overall
structural-organizational” problem, which concerns how an occasion of interaction is struc-
tured, and how placement in the overall structure informs the construction and understand-
ing of the talk and other conduct as turns, as sequences of actions, etc.
The conversational episodes analyzed in the next sections2 are presented

together with an English translation. Although grammatical gender may occur
in various turns in each extract, special focus is given only to certain usages,
depending on the particular phenomenon examined each time. The turns
in which these usages occur are followed by glossing explicating the specific
lexico-grammatical features of the elements composing the turns.
6.2 The Routine Achievement of Sociocultural Gender in Interaction:

The Role of Grammatical Gender
The first question to examine empirically is how the use of grammatical gender
contributes to the construction of sociocultural gender in interaction. What
are the implications of the compulsory use of items marked by grammatical
gender in the composition of turns for the sociocultural world that partici-
pants jointly construct in interaction? In the next section, I start exploring this
question by analyzing grammatical gender in self-reference in interaction.
6.2.1 Self-reference and Grammatical Gender

Self-reference in interaction is divided into individual and collective (Lerner
and Kitzinger 2007). In individual self-reference a speaker refers only to herself
or himself via first person singular, while in collective self-reference a speaker
refers to self collectively, that is, refers to self as a member of a collectivity, via
first person plural.3 In the first two extracts, I analyze grammatical gender in
individual self-reference.
Three friends, Vaso, Yana and Katerina, participate in the conversational epi-
sode in Extract 1. At lines 2, 5 and 7, Vaso invites her friends Yana and Katerina
to the movies.
Extract 1 (Ι.14.Α.19.1): Lord of the Rings

1 (1.3)
2 Vaso Παιδιά:,=
Hey guys,
2 Extracts 1, 8, 13, 14 and 27 were analyzed for the first time in a talk that I gave in the Symposium
Gender and the Greek Language at Freie Universität Berlin, February 2, 2012. The written ver-
sion of the analysis can be found in the special issue on gender and the Greek language in the
journal Gender and Language (Alvanoudi forthcoming).
3 For a systematic analysis of pronoun systems across languages and their semantics see
Mühlhäusler and Harré (1990).
96 chapter 6
3 Yana =Μπορεί να πάω την^ Μπέμπτη.

I may go on Thursday.
4 (0.7)
5 Vaso δε σηκω[νόμαστε όπως] είμαστε, .hh >να πάμε να
((in a laughing tone. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
why don’t we stand up, .hh and go to
6 Yana? [°(. . . . . . . . . . .)]
7 Vaso δούμε< τον Άρχοντα των Δαχτυλι[διώ:ν¿ ]
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .))
see the Lord of the Rings¿
8 Yana [Θα πρέπει]
You should
9 να ’χεις κλείσει πολύ νωρίτερα εισιτήρι[ο.]
have booked a ticket much earlier.
10 Vaso [Ε] τώρα,
Uh now,
11 έχει παλιώ:σει μωρέ. είναι
it became old you fool. it’s been
12 [ δυο βδομάδε:ς. ]
two weeks.
13 Κaterina → [Εγώ νιώθω πολύ κουρα]σμένη παιδιά από >χθε:ς.
I feel.1sg.pres very tired.part.nom.fem guys PREP yesterday
Hey guys I feel very tired.fem since yesterday.
14 δηλαδή, >δε μπορώ να πάρω τα πόδια μου.°<=
that.is neg can.1sg.pres subj move.1sg.pfv the.acc.
neuter.pl legs.acc.neuter my
that is, I cannot move my legs.
15 Vaso =Χθες έκατσες μέχρι τις τέσσερι:ς?
Yesterday did you stay until four?
Katerina rejects Vaso’s invitation at lines 13 and 14. She refers to herself via
the personal pronoun εγώ ([eɣó] ‘I’) and the first person singular verbs νιώθω
([ɲóθo] ‘feel’) and δε μπορώ να πάρω ([ðe boró na páro] ‘cannot move’). The
speaker uses the grammatically feminine participle κουρασμένη ([kurazméni]
‘tired’) to describe herself and give a non-fault ability account (Heritage 1988)
for her rejection. The feminine participle is a gender membership category
that categorizes the speaker as woman. However, the speaker is not using this
category “for its gendered properties or to invoke gendered attributes”, to quote
Kitzinger (2007, 44). Instead, the speaker employs the feminine participle as a
tool for rejecting Vaso’s invitation. In particular, Katerina’s turn is a dispreferred
second pair part that distances from the first pair part that her co-participant
produced; it is a rejection delivered after an invitation. As Schegloff (2007a,

66) points out, dispreferred responses are often treated as “accountable” and
are designed in such a way as to avoid too overt a disalignment with the first
pair part. Katerina treats her rejection as accountable, by offering an excuse
for not wanting to go to the movies. She describes herself as tired because of
what she did yesterday. She employs the feminine participle, which encodes
the meaning of tiredness, together with other elements composing the turn,
to accomplish the action of rejection. However, more things are happening in
interaction than that. Although sociocultural gender is not a relevant category
for what the speaker is doing at that particular moment in interaction, it is
indexed via grammatical gender and, thus, is covertly assumed by participants
as an aspect of the speaker’s identity. This covert assumption is part of the
routine and unremarkable meanings that are produced in this conversational
episode when grammatical gender is used in self-reference.
A similar case is examined in Extract 2. Dimitris tells a story about an inci-
dent that he experienced during his exam period at the university.
Extract 2 [Ι.20.Α.28.4]: Exams

1 Dimitris [Απ’ τις πεντέμιση] ώρα
From half past five
2 → ξύπνιος έχω αρχίσει να:: (0.7) χάνω ντον^ γκόσμο
((noise. . . . . . . . . . .)) ((noise. . . . .))
awake.nom.masc have.1sg start.pfv subj loose.1sg.ipfv
the.acc.masc world.acc.masc
being awake.masc I have started (0.7) losing my senses
3 να νυστάζω, .hh μπαίνει μέσα, δίνει τα θέματα, λέει το μόνο
((noise. . . . . . . . . . . .))
falling asleep, .hh [she] comes in, [she] gives the tests, [she]
says the only [thing]
((11 lines omitted))
15 Dimitris → =Γράφω εγώ σίγουρος >εντωμεταξύ< τα γράφω
write.1sg.pres I sure.nom.masc meanwhile them write.1sg.
pres
Meanwhile I write [being] sure.masc I write
16 στο πρόχειρο τα καθαρογράφω της δίνω ένα γραπτό- πρώτη
in rough I write them up I give her the exam paper- it’s the
first
17 φορά δίνω στη ζωή μου τέτοιο γραπτό, .hh (1.2) τέσσερα.
((noise. . . .)) ((noise. . . . .))
time in my life that I give such an exam paper, .hh (0.2) four.
98 chapter 6
In a multi-unit turn Dimitris refers to himself via the first person singular verbs
and pronouns έχω αρχίσει να:: χάνω ([éxo arçísi na:: xáno] ‘I have started losing’),
να νυστάζω ([na nistázo] ‘falling asleep’), γράφω ([ɣráfo] ‘I write’), καθαρογράφω
([kaθaroɣráfo] ‘I write up’), δίνω ([ðíno] ‘I give’), εγώ ([eɣó] ‘I’), μου ([mu] ‘my’).
At lines 2 and 15, he uses the masculine adjectives ξύπνιος ([ksípɲos] ‘awake’)
and σίγουρος ([síɣuros] ‘sure’) to describe himself. These adjectives are gen-
der membership categories, which categorize the speaker as man. Similar
to Extract 1, in this conversational episode, sociocultural gender is not fore-
grounded interactionally. The speaker tells a story in which he is the main
protagonist. He took exams and he failed. In delivering the story, he describes
himself. He woke up early in the morning and went to the examination room,
being self-confident. Although he was sure that he had succeeded in passing
the exams, he actually failed. The speaker selects and deploys the masculine
adjectives together with other elements to deliver his story. At the same time,
these adjectives index speaker’s masculine gender and, thus, presuppose it as
an aspect of his identity.
In both Extracts 1 and 2, gender is not relevant interactionally but it is still
made available in interaction because it is morphologically codified. The avail-
ability of gender in interaction is a phenomenon that deserves our attention
as analysts in spite of the fact that participants show no demonstrable orienta-
tion to it (Schegloff 1997; 2007b). Sociocultural gender is constructed in rou-
tine ways via the compulsory use of grammatical gender in the composition
of turns given the role of grammatical gender in indexing sociocultural gender
referentially. Items marked by grammatical gender constitute obligatory gen-
der membership categories that ascribe female or male sex to speakers, cat-
egorize them as women or men, and are associated with participants’ covert
assumptions or presuppositions about the speaker as woman or man. When
participants accomplish social actions such as rejecting an invitation or telling
a story, they assume feminine or masculine gender as part of their identity and
they construct gender as a given aspect of their sociocultural world.
The relation between grammatical gender and collective self-reference is
examined in the next three extracts.4 According to Lerner and Kitzinger (2007,
526–527), these collectivities can be ‘organizational’, e.g. when one speaks on
behalf of an enterprise, ‘relational’, e.g. when one speaks on behalf of a couple,
or ‘circumstantial’, e.g. when one speaks on behalf of persons who happened
to be together somewhere.
In Extract 3, at lines 10 and 13, participants refer to a circumstantial collec-
tivity that includes the speaker, Katerina, and her co-participant, Vaso. In the
4 See Pavlidou (2008; 2012a) for the study of collective self-reference in Greek interaction.
lines preceding this extract, Vaso, Yana and Katerina refer to the holidays that
Yana had in the summer after her graduation.
Extract 3 [Ι.14.Α.19.1]: Summer holidays

1 Vaso [↑A::χ,] ελπίζω κι εγώ φέτος να είμαι το καλ- φέτος
((in a laughing tone . . .
A::h, I hope that this year I will also be in the sum- this year
2 το καλοκαίρι να ’μαι σ’ αυτή τη φάση (.)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
this summer [that I will] be in this situation (.)
3 [και η Κατερίνα:¿ ]
. . . . . . . . . . . . . . . . .))
and Katerina as well¿
4 Yana [Το εύχομαι και για] σας κορίτσια:.
I wish this for you too girls.
5 (1.0)
((noise during the silence))
6 Katerina Ποιο?
What?
7 (.)
8 Katerina [ Να: τελειώ]σουμε °ή
SUBJ finish.1pl.pfv or
To have finished or
9 Yana [>Θα είναι στ’ αλήθεια,<]
Truly it will be,
10 Katerina → με[τά να ’μαστε ξέγνοιαστες?>]
afterwards subj cop.1pl carefree.nom.fem.pl
to be carefree.fem afterwards?
11 Yana [η καλύτερη περίοδος σας] με[τά.]
the best time of your life later.
12 Vaso [Να] εί-
subj cop
To have
13 → >να ’χουμε τελειώσει< και να ’μαστε °ξέγνοιαστες¿=
subj have.1pl finish.pfv and subj cop.1pl carefree.nom.
fem.pl
finished and to be carefree.
14 Katerina =Μετά °αρχίζουν οι έγνοιες.
((noise. . . .))
Then the problems begin.
100 chapter 6
At lines 1, 2 and 3, Vaso wishes to experience something similar next summer.

At line 4, Yana expresses the same wish about her co-participants and refers
to them via the feminine address term κορίτσια: ([korít͡sça:] ‘girls’) and the
second person plural pronoun σας ([sas] ‘you’). Instead of a relevant second
pair part, Katerina delivers the first pair part of an insert sequence at line 6.
Insert sequences occur after the first pair part of the basic minimal adjacency
pair, or base first pair part, and before the base second pair part (Schegloff
2007a, 97–114). The insert sequence in this episode initiates a repair addressing
speaker’s trouble in understanding the base first pair part, that is, Yana’s wish.
The repair is not completed by Yana, and Katerina reformulates the first pair
part of the insert sequence at lines 8 and 10, partly overlapping with Yana’s
turn. She uses the first person plural verbs να: τελειώσουμε ([na: teʎósume] ‘to
have finished’) and να ’μαστε ([ná maste] ‘to be’) to refer to herself and her
co-participant Vaso. She describes this collectivity via the feminine adjective
ξέγνοιαστες ([kséɣɲastes] ‘carefree’). At lines 12 and 13, Vaso delivers a second
pair part that completes the repair initiated by Katerina. She repeats the first
person plural verbs >να ’χουμε τελειώσει< ([na xume teʎósi] ‘to have finished’)
and να ’μαστε (‘to be’) to refer to the same collectivity, and the feminine adjec-
tive ξέγνοιαστες (‘carefree’) to describe the collectivity. The feminine adjective
is a gender membership category, which ascribes female sex to the speaker and
her co-participant and categorizes them as women. The gendered categoriza-
tion of referents is not relevant for interaction at this particular moment. Yet,
it occurs in a compulsory way, when participants deal with trouble in under-
standing the talk in interaction, and is tacitly presupposed by participants.
In the next extract, Natasa and her husband, Vasilis, describe the difficulties
that they encountered because of the bad weather, and they refer to a rela-
tional collectivity, i.e. a couple.
Extract 4 [Ι.14.Α.18.1]: Weather

1 Polikseni Πώς τα πάτε παιδιά με τον^ γκαι↑ρό?=
Hey guys how do you cope with the weather?
2 Vasilis =Τι να τα πάμε. να. εδώ:, ήμασταν απ’ το
So-so. here, we have been here since
3 Σαββάτο[: °ε: ]
Saturday uh:
4 Natasa → [Αποκλει]σμέ[:νοι,]
blocked.part.nom.masc.pl
Blocked.masc,
5 Vasilis → [ απο]κλεισμένοι, >μέχρι την^

blocked.part.nom.masc.pl until the.
acc.fem
blocked.masc, until
6 Ντετάρτη το βράδυ.<
Wednesday.acc.fem the.acc.neuter night.acc.neuter
Wednesday night.
At line 1, Polikseni addresses a question to Vasilis and Natasa, and selects them
as next speakers via the address term παιδιά ([peðʝá] ‘hey guys’). The second
pair part made relevant by this first pair part, i.e. the answer, is built in a col-
laborative manner (Lerner 2004). Vasilis delivers an answer at lines 2 and 3,
and employs the first person plural verbs πάμε ([páme] ‘we cope’) and ήμασταν
([ímastan] ‘we were’) to refer to himself collectively. He speaks on behalf of
a relational collectivity that includes himself and his wife, Natasa. At line 4,
Natasa continues the second pair part that Vasilis started and she uses the
masculine participle αποκλεισμέ:νοι ([apoklizmé:ni] ‘blocked’) to describe the
collectivity. The second pair part is brought to completion by Vasilis at lines
5 and 6. Vasilis employs the same masculine participle that Natasa employed
to describe the same relational collectivity. In this conversational episode,
nothing special is happening with respect to gender. However, inferences
are produced in automatic and systematic ways, because participants empoy
the masculine participle for reference to male plus female persons. The male
categorization of referents imposed by the masculine participle is a tool that
participants employ to activate the inference of human/universal and achieve
generic reference. Thus, in this conversational episode, participants tacitly pre-
suppose the stereotypic association of man with the norm as an unquestioned
assumption about the way the social world is organized.
In Extract 5, Dimitris, Tania, Zoi and Melita discuss the relation between
religion and sexism. Tania and Melita disagree on whether Christianity is a
more sexist religion than Islam. They refer to themselves collectively, as mem-
bers of a religion, that is, they refer to an organizational collectivity.
Extract 5 [Ι.20.Α.28.6]: Christians versus Muslims

1 Tania [Παιδιά:,] [τώρα] >ας μη πιάσουμε<
((in a laughing tone. . . . .
Hey guys, now let’s not start talking
2 τη:- [το θέμα της θρησκείας,]
. . . . . . . . . . . . . . . . . . . . . . . . . . .))
about the- the topic of religion,
102 chapter 6
3 Zoi [Καλά ναι, γιατί κι ε]γώ: [εκνευρίζομαι. ]

Okay yes, because I get upset as well.
4 Dimitris [Διαφωνώ μ’ αυτό.] δεν
I disagree with this. there is no
5 [ε: >°τίθεται θέμα.<]
uh: question of that.
6 Τania [Γιατί θεωρώ] ότι ο: χριστιανισμός είναι κάπως
((in a laughing tone. . . . . . . .
Because I think that in some ways Christianity is
7 σεξιστική θρησκεία.
. . . . . . . . . . . . . . .. . . .))
a sexist religion.
8 (1.1)
9 Zoi [Ναι:. ]
Yes.
10 Melita [Όχι μό]νο ο χριστιανισμός.
Not only Christianity.
11 πόσο [μάλλον ο <↑μωαμεθα]νισμό:ς.>=
Islam is a lot more.
12 Τania [ Πολλέ:ς. (αλλά . . . )]
Many [religions]. (but)
13 Tania =[Γι ’ αυ]
That’s why
14 Zoi =[(Εγώ]..[ . . . . . . . . . . . . . . . . . . .)]
(I . . . .)
15 Melita [Κορυφαία σεξιστική:] θρησκεία.
A major sexist religion.
16 Tania Ναι. αλλά γι’ αυτήν που:: πιστεύουμε.
Yes. but regarding the [religion] that we believe in.
17 Melita [Είμαστε πολύ πιο ε]λεύθερη θρησκεία από ά:λλες.
We have a religion that is freer than others.
18 Dimitris [( . . . . . . . . . . . . . . . . . .)]
19 (0.7)
20 Melita Πολύ:.
A lot [more].
21 Τania Κοίτα, αν ακολουθούσα[με κατά γρά:μμα ] αυτό που
((noise . . .
Look, if we followed entirely what
22 Dimitris [°Δεν^ ντο συζητάμε.]

There is no doubt about it.
23 Tania λέει η ↓θρησκεία οι γυναίκες είναι::: (1.1) πάντα:=
. . . . . . .))
religion dictates women are (1.1) always
24 Μelita = >Μ δεν^ γκζέρω.< εγώ εδώ στην εστία που: σας λέω
Mm I don’t know. I am telling you, here in the dorm where
25 συναναστρέφομαι με άλλες θρησκείες, (.)
((noise. . . . .))
I mix with other religions, (.)
26 → βλέπω ότι είμαστε πολύ ελεύθεροι.
see.1sg.pres conj cop.1pl.pres very free.nom.masc.pl
I see that we are very free.masc.
27 (0.5)
Tania asserts that Christianity is a sexist religion at lines 6–7, and Melita
argues that Islam is even more sexist at lines 10–11. At line 17, Μelita asserts
that Christianity, contrary to Islam, is a religion that allows its believers to
be freer. She refers to herself and other Christians via the first person plural
είμαστε ([ímaste] ‘we are’). Tania re-refers to the same organizational collectiv-
ity at line 21 via the first person plural verb ακολουθούσαμε ([akoluθúsame] ‘we
followed’). At lines 24, 25 and 26, Melita produces a turn composed by two
tcus. In the first tcu >Μ δεν^ γκζέρω.< ([m ðen gzéro] ‘Mm I don’t know.’)
Melita disagrees with Tania. In the second tcu εγώ εδώ στην εστία που: σας λέω
συναναστρέφομαι με άλλες θρησκείες, (.) βλέπω ότι είμαστε πολύ ελεύθεροι. ([eɣó eðó
stin estía pu: sas léo sinanastréfome me áles θriscíes, (.) vlépo óti ímaste polí
eléfθeri] ‘I am telling you, here in the dorm where I mix with other religions, I
see that we are very free.masc.’) she gives an account for her disagreement, by
drawing on her personal experience. Melita keeps reference to the same orga-
nizational collectivity, i.e. all Christians, by employing the first person plural
verbs είμαστε (‘we are’), and describes the collectivity via the masculine adjec-
tive ελεύθεροι (‘free’). She differentiates the collectivity of Christians from the
collectivity of Muslims, and she selects the masculine adjective as a tool for
accomplishing this differentiation: us, Christians, are different (very free) than
the others, Muslims. In addition, the speaker employs the masculine adjec-
tive as a vehicle for achieving generic reference. Therefore, the inference of
human/universal via male sex is part of the routine meanings tacitly produced
in this conversational episode.
104 chapter 6
In Extracts 4 and 5, female speakers refer to collectivities that include them-

selves and categorize the members of the collectivities as male. In both cases,
the information of male sex codified by the masculine grammatical gender
is employed as a tool for referring to male plus female persons or persons of
indefinite sex, and accomplishing social actions that are not related to par-
ticipants’ orientation to gender. Therefore, the gender membership categories
employed by participants in these conversational episodes do interactional
work which is not “clearly or unambigously gendered” (Kitzinger 2007, 46).
However, the use of the masculine grammatical gender for generic reference
invokes tacitly the stereotypical association of man with the norm. This stereo-
type is presupposed by participants as part of their commonsense knowledge
and serves as a sort of background that is ‘seen’ but not noticed.
Finally, speakers may refer to themselves as part of an abstract group, by
employing the indefinite or generic second person singular. In this way, speak-
ers can make generalizations about collective experience or personal opin-
ions that are assumed to be common among the members of a community
(Kitagawa and Lehrer 1990; Malone 1997). According to Schegloff (1996b,
442–443), indefinite self-reference can introduce the relevance of speaker’s
and co-participants’ experience in interaction, or it can invoke the “personal
and knowledgeable I” (Schegloff 1988, 12–13). In the following extract, I analyze
grammatical gender in generic second person self-reference.
In Extract 6, Zoi tells a story about her trip to Paris on New Year’s Eve. She
describes how the French behaved in public spaces and how insecure she and
her partner felt.
Extract 6 [Ι.20.Α.28.3]: France on New Year’s Eve

1 Ζoi >Τώρα όταν είσαι< σε μια ξένη χώρα, (.) .hh και
Now when you are in a foreign country, (.) .hh and
2 τους βλέπεις όλους αυτούς να: πίνουν να κάνουν και να ράνουν,
((noise. . . . . . . . . . . . . . . .))
you see all these people drinking and doing stuff,
3 (.) .hh (.) πιχί αν- αν συνέβαινε αυτό >στη Θεσσαλονίκη<
((lighter sound during the silence))
((noise. . . .))
(.) .hh (.) for example if- if this happened in Thessaloniki
4 θα ’λεγες ε ’ντάξει. κοίτα διασκεδάζουν. ↑ε κάνουν
you would say uh okay. look they are having fun. uh they do
5 βλακείες ε:, (.) τώρα εκεί όταν^ ντους βλέπει:ς,
((Dimitris sniffs))
stupid things uh:, (.) now when you see them there,
6 Εvagelia Α:γχώνεσαι [°λίγο.]

You stress out a bit.
7 Ζoi [ Α:]γχώνεσαι. .hh
You stress out .hh
8 >εντωμεταξύ< μη σε κλέ:ψουν, μες το μετρό πί:τα, >δηλαδή<
meanwhile that they will rob you, in the metro packed out,
that is
9 μπορεί να σε πιάσουν να σε κάνουν ό,τι θε:ς,
((noise . . .))
they might touch you [they might] do anything to you,
10 Dimitris Χίλια [>δυο ρε παιδί μου τέλος] πάντων.<
Anyway anything [could happen].
11 Zoi [ °Δεν^ ντο συζητά:ω. ]
There is no doubt about it.
12 (0.5)
13 Ζoi Κι ούτε φυσικά μπορείς να εμπλακείς και σε καβγά και να
And of course you cannot get involved in a fight and
14 να πεις και τίποτα. (0.7) εκεί: είναι όλοι Γάλλοι
say something. (0.7) they are all French there
15 → και είσαι ένα:ς (.) ξένος. (1.0) έτσι. είναι δέκα άτομα και είσαι δύο.
((lighter sound))
and cop.2sg.pres one.nom.masc foreigner.nom.masc adv
cop.3pl.pres ten people.nom.neuter and cop.2sg.pres
two
and you are a.masc foreigner.masc. (1.0) like that. they are ten
people and you are two.
16 (1.2) ’ντάξει δεν^ μπάθαμε έτσι κάτι:, >°απλά εγώ
(1.2) okay nothing happened to us, I was just
17 σοκαρίστηκα απ’ ντον^ γκόσμο,°< (.) .hh
shocked by the people, (.) .hh
In her recounting of the trip to Paris, Zoi refers to all non-French tourists
including herself and her partner via the second person singular verbs and
pronouns είσαι ([íse] ‘you are’), βλέπεις ([vlépis] ‘you see’), θα ’λεγες ([θá leʝes]
‘you would say’), βλέπει:ς ([vlépi:s] ‘you see’), α:γχώνεσαι ([a:ŋxónese] ‘you stress
out’), σε ([se] ‘you’), θε:ς ([θe:s] ‘you want’), (ούτε) μπορείς να εμπλακείς ([úte borís
na eblacís] ‘you cannot get involved’), να πεις ([na pís] ‘you say’). At line 15, she
characterizes the abstract group of referents as foreigners and categorizes the
106 chapter 6
group as male by using the masculine noun ξένος ([ksénos] ‘foreigner’) together
with the masculine indefinite article ένα:ς ([éna:s] ‘one’). At line 16, she refers to
herself and her partner via the first person plural δεν^ μπάθαμε ([ðen báθame]
‘we didn’t hurt ourselves’ meaning ‘nothing happened to us’). In this way, she
shows that she understands herself as part of the abstract group of persons
that she referred to in her previous tcu. The masculine items that the speaker
selects to describe the abstract group of persons are part of the tools that the
speaker employs for the delivery of the story at that particular moment. At the
same time, these items constitute gender membership categories that catego-
rize referents as male and, therefore, allow generic reference to be achieved
via the inference of human/universal. Similar to Extracts 4 and 5, in Extract 6,
gender is not foregrounded interactionally. Yet, the stereotypical association of
man with the norm is tacitly presupposed.
6.2.2 Reference to Recipient(s) and Grammatical Gender

In this section, I examine grammatical gender in reference to recipient(s). In
Extract 7, Tania, Zoi, Melita, Evagelia and Dimitris refer to the credits they got
in the national education exams.
Extract 7 [Ι.20.Α.28.1]: Credits

1 Dimitris [Πόσα μόρι]α?
How many credits?
2 (.)
3 Tania Εγώ: ε δεκαεφτά: εξακόσια: [ (έβγαλα). ]=
Me uh (I got) seventeen [thousand] six hundred.
4 Dimitris [>Τη χτύπησες τη]=
You succeeded in entering
5 Εvagelia → =[ Α:: εσύ ] είσαι κα[λή::. ]=
interj you.sg cop.2sg.pres good.nom.fem
A::h you are good.fem.
6 Dimitris =[φιλολογία.<]
the school of philology.
7 Zoi [>Α εσύ ’σαι πιο]=
interj you.sg cop.2sg.pres more
Ah you are more
8 Μelita =[Α:. ((γελά . . . . . . . . . . . .]. . . . . . . . . . . . .))]=
A::h. ((she laughs))
9 Ζoi → =[[ικανή (από μένα).<]
competent.nom.fem prep me
competent.fem (than me).
10 Dimitris =[Δεν μπορούσες [να (. . .)?]]

Couldn’t you (. . .)?
At line 1, Dimitris asks Tania how many credits she got, Tania delivers an
answer at line 3, and co-participants produce a set of minimal post-expansions
at lines 4, 5, 6, 7, 8 and 9. Post-expansions are sequences that occur after the
base second pair part (Schegloff 2007a, 115–168). Minimal post-expansions or
sequence-closing thirds are designed to move for sequence closing. In their
minimal post-expansions, participants refer to Tania via the second person
singular verbs and pronouns χτύπησες ([xtípises] ‘you succeeded in entering’),
εσύ είσαι and εσύ ’σαι ([esí íse/esí se] ‘you are’), and assess her via the feminine
adjectives καλή:: ([kalí::] ‘good’) and ικανή ([ikaní] ‘competent’). Participants
employ these adjectives to attribute specific features to the recipient and
evaluate her competence in achieving a high score in the national exams.
Additionally, the feminine adjectives are gender membership categories that
ascribe female sex to recipient and categorize her as woman. The deployment
of these gender membership categories is not related to participants’ orienta-
tion to gender. However, the recipient’s feminine identity is indexed via the
feminine grammatical gender in a compulsory way and presupposed as part of
the commonsense knowledge that participants share.
In Extract 8, Dimitris reconsiders the choices that he would make regarding
his professional career. At lines 5 and 6, Melita and Tania ask Dimitris what he
would do if he was about to finish school now, and initiate a sequence that is
brought to completion by Tania’s minimal post-expansion at line 17.
Extract 8 [Ι.20.Α.28.1]: Choosing a profession

1 Dimitris [Πάντως] βλέπω ότι όλοι: >θα
Anyway I see that you would all
2 συνεχίζατε στο πανεπιστήμιο.
continue with your undergraduate studies.
3 [εγώ δε θα συνέ]χιζα στο πανεπιστήμιο.<
I wouldn’t continue with my undergraduate studies.
4 Εvagelia [((giggle . . . . . . .))]
5 Melita Αλλά τι θα ’κα[νες?]
((cup sound))
But then what would you do?
6 Tania [Τεΐ?]
[Would you choose] tei [Technological Educational
Institution]?
((4 lines omitted))
108 chapter 6
11 Dimitris Στρατό και μία ιδιωτική σχολή για δύο χρόνια.

Army and a private school for two years.
12 (.)
13 Εvagelia Ε [ τι: <ιδιω]τική σχολή?>
Uh what kind of private school?
14 Dimitris [Πραγματικά.]
Indeed.
15 Dimitris >Αθλητικής δημοσιογραφίας.<
Sport media.
16 (0.7)
17 Τania → <Ω[ραίος::.> ]
Nice.nom.masc
Nice.masc.
18 Μelita [Γιατί ιδιωτι]κή [(βασικά)?]
Why a private one (by the way)?
At line 11, Dimitris replies that he would go to the army and choose a private
school. At line 13, Evagelia asks Dimitris which private school he would choose
and Dimitris delivers an answer to the question at line 15. At line 17, Tania
produces a minimal post-expansion, assessing Dimitris’s choice as nice. She
uses the grammatically masculine adjective ωραίος:: ([oréos::] ‘nice’) to accom-
plish the action of assessment. Yet, besides assessing the recipient in a positive
way, the speaker also categorizes him as man. In Extracts 7 and 8, the gendered
categorization of recipients occurs in the service of the social action of assess-
ment. Speakers assess aspects of recipients’ activities or conduct, which are
not related to their gender identity, for example their exam-score or the profes-
sion they would choose. In both cases, grammatically feminine and masculine
adjectives are employed as vehicles for accomplishing the action of assess-
ment. Yet, in both cases, these items categorize recipients as women and men
respectively. Even though participants do not orient explicitly to gender, they
presuppose it tacitly as part of the taken-for-granted world.
In Extract 9, four female participants, Natasa, Anna, Marina and Elsa,
describe the decorative items that Marina and Chrysanthi have in their
apartment.
Extract 9 [Ι.20.Α.29.3]
1 Νatasa [↑Αυτά από πού τα παίρνετε εδώ] τα ωραία?=
Where do you buy these nice things over here?
2 Anna =>(. .[ . . . . . . . )<]
3 Marina [Τα βρήκαμε] στο δρόμο.

We found them on the street.
4 (.)
5 Natasa Πλάκα °κάνεις.=
You must be joking.
6 Marina =Όχι. κι αυτό το βρή[καμε στο δρό:μο:,]
No. we also found this one on the street,
7 Natasa [ Ά:ντε ρε::]::.
((noise))
Hey come on.
8 (0.6)
9 Elsa °Καλά, σοβαρά: μιλά:ς?=
Well, are you serious?
10 Νatasa =Θέλω [κι εγώ τέ]τοια πρά[γματα::.]
I also want such things.
27 Μarina °Και φουστανάκι.=
And a tiny dress.
28 Νatasa → =Τι [τυχερές που είστε. πάντα ονειρευόμουνα >να ’χω ένα]
interj lucky.nom.fem.pl pron.rel cop.2pl.pres always
dream.1sg.imperf subj have.1sg one.acc.neuter
You are so lucky.fem. I always dreamt of having
29 Chrysanthi [ .h Απλά είναι μαύρο °κι αυτό. και δεν φαίνεται. ]
.h This one is also black. and it hides the other.
30 Νatasa τέτοιο στο σπίτι μου.< ((γελά . . .)) δεν^ γκζέρω
((laughing . . .
such [a thing] in my house. ((she laughs)) I don’t know.
31 για[τί αλλά μ’ α[ρέ]σει. ((γελά . . . . . . . . . . . . . . . . . . . . . . . . . .]. . .))
. . . . . . . . . . . . . . . . . . . . . . . . ))
why but I like it. ((she laughs))
Αt line 1, Natasa initiates a sequence that comes to a possible completion at

line 7 and is further expanded. In the post-expansion, Natasa assesses her
recipients, Chrysanthi and Marina, as lucky for having such decorative items
(line 28). She addresses her recipients via the second person plural verb είστε
([íste] ‘you are’), and assesses them via the feminine adjective τυχερές ([tiçerés]
‘lucky’). Similar to the previous two conversational episodes, the deployment
of the gender membership category in this conversational episode is not
related to speaker’s orientation to gender. Still, it is linked with speaker’s covert
assumption about the recipients addressed as women.
110 chapter 6
The generic use of the masculine in reference to recipients is examined in

Extract 10. Melita, Tania, Zoi and Dimitris experienced difficulties in succeed-
ing in exams they gave for specific courses as undergraduate students. In the
lines preceding this extract, Melita refers to a particular incident that hap-
pened to her. Although she was very well prepared for the exams for a specific
course and she was absolutely sure that she had succeeded, she actually failed.
Extract 10 [Ι.20.Α.28.4]: Failing in the exams

1 Μelita Αλλά έπρεπε να πάω να δω το [γραπτό.]
But I should have checked my exam paper.
2 Τania [Έπρεπε.]
You should.
3 (0.6)
4 Τania Ναι. σε τέτοιες περιπτώσεις: (0.8)
Yes. in such cases (0.8)
5 Μelita → Μα δε σας έχει τύχει? να είστε σί:γουροι ότι
conj neg you.gen.pl has happen.pfv subj cop.2pl
sure.nom.masc.pl conj
But didn’t this happen to you? to be sure.masc that
6 → [το περάσατε. σί:γουροι. ]
it.acc pass.2pl.past sure.nom.masc.pl
you passed it. sure.masc.
7 Ζoi [Ε: με το πεντακόσια >ένα.<] ↓ κάθε φορά που το ’δινα,
U:h [it happened] with [the course with the code] five hun-
dred one. every time I was taking exams,
8 έλεγα ’ντάξει.
I said okay.
((6 lines omitted))
15 Dimitris Γραπτό δεν είδα. τηλέφωνο πήρα.
I didn’t see my exam paper. I called [them].
16 (0.9)
At lines 5–6, Μelita asks her co-participants, Tania, Zoi and Dimitris, if they
had a similar experience, and selects them as next speakers via the second
person plural verb and pronoun είστε ([íste] ‘you are’) and σας ([sas] ‘you’).
She employs the masculine adjective σί:γουροι ([sí:ɣuri] ‘sure’) to describe
her recipients. The speaker uses the masculine grammatical gender to refer
to all co-participants, two females and one male, and, thus, accomplish the
social action that the turn is designed to do, i.e. a question addressed to all co-
participants. Note, however, that more things are happening in interaction.
By using the masculine gender for generic reference, the speaker presupposes
the stereotype of man as the norm and invites her recipients to make the infer-
ence of human/universal. This inference is part of the routine, unremarkable
meanings produced in interaction.
In the conversational episodes examined in this section, sociocultural gen-
der is constructed in tacit and systematic ways due to the compulsory use of
grammatical gender in the composition of turns. In addition, the social hier-
archy between women and men is a feature of the sociocultural world that
participants presuppose when they use the masculine gender for generic
reference.
6.2.3 Reference to Third Person(s) and Grammatical Gender

In this section, I move to the study of grammatical gender in third person ref-
erence. As I mentioned in section 5.2.2, third person reference in interaction
is organized on the basis of the principles of recognition and minimization.
Moreover, third person reference in interaction is distinguished into locally
initial and locally subsequent depending on the position or the ‘slot’ in which
reference is done, that is, whether the referent is mentioned for the first time
or second, third time, etc. (Schegloff 1996b, 450). Full noun phrases, such as
proper names or descriptions, are usually deployed as initial reference forms,
while pronouns are usually deployed as subsequent reference forms.
In Extract 11, Εvagelia, Dimitris, Tania, Zoi and Melita try to identify a class-
mate they had in the first year of their undergraduate studies.
Extract 11 [Ι.20.Α.28.1]: Our classmate

1 Ζoi → Εμείς που ήμασταν στο ίδιο τμήμα θυμάστε μί:α, .hh
we pron.rel cop.1pl.past in.the.acc.neuter same.acc.
neuter class.acc.neuter remember.2pl.pres one.nom.
fem
We who have been in the same class do you remember one.
fem [girl], .hh
2 <με γυαλιά, μ’ αγορέ μαλλιά που μιλούθε °λίγο έτσι:¿>
((lisps . . . . . . . . . . .))
with glasses, with boyish hair who was speaking like this?
3 (0.5)
4 Dimitris (Τσίδα?)
(Tsida?)
5 Tania → Ψηλή? χοντρή?=
((lighter sound))
tall.nom.fem fat.nom.fem
Tall.fem? fat.fem?
112 chapter 6
6 Ζoi =[ Ναι::. ]
Yes.
7 Μelita =[Εσύ τι γρά]μμα είσαι
In which class were you
8 Dimitris =Α:XA,=
A:ha,
9 Tania → =H[: που πήγε κλα]σικό.=
the.nom.fem pron.rel go.3sg.past classic.acc.neuter
The.fem [one] who studied classic literature.
10 Dimitris → [η (προ)- η Άννα?]
((lighter sound))
the.nom.fem the.nom.fem Anna.nom.fem
The.fem (pro)- Anna.fem?
11 Zoi → =Ναι ρε. αυ[τή. ]
yes PARTICLE this.nom.fem
Hey yes. that one.fem.
12 Dimitris → [Η Ά](ννα)? =
the.nom.fem Anna.nom.fem
Anna.fem?
13 Zoi =[>(Όχι όχι.)<]
(No no.)
14 Tania → =[ Η::]:: [αχ::. ]
the.nom.fem interj
The.fem ah::.
15 Ζoi [Που ήταν πολύ] φυτό:.
Who was a real nerd.
16 (.)
17 Τania → Αυτή πρέπει να είχε πάρει και::- να είχε περάσει στους
she.nom must subj have.3sg.past take.pfv and subj
have.3sg.past pass.pfv in.the.acc.masc.pl
She must have taken and- must have been one of the
18 δέκα πρώτους >ξέρω γω< κάτι τέτοιο, όταν: [μπήκαμε.]
first ten [students] when we entered (the school), something
like that so to speak.
19 Μelita [ Υποτρο]φία:?=
[With a]
scholarship?
20 Τania → =Υποτροφία. μπρά°βο. .h (.) πώς τη λένε [ ρε? ]
scholarship.fem interj adv she.acc call.3pl.pres PARTICLE
[With a] scholarship. right. .h (.) hey what’s her name?
21 Εvagelia [Δεν^] γκζέρω.

I don’t know.
22 (0.5)
23 Ζoi → [Αυτή στάνταρ] [θα έχει πάρει πτυχίο.]
she.nom certainly fut have.3sg.pres get.pfv degree.acc.
neuter
She must have graduated for sure.
In a question to her co-participants at lines 1 and 2, Zoi refers to a girl that used
to be in the same class as theirs. This question is a possible first pair part in
a pre-sequence, that is, in a sequence occuring before the base first pair part
(Schegloff 2007a, 28–57), which aims to ensure that co-participants know and
are able to identify the third person that Zoi refers to. Zoi employs the femi-
nine indefinite pronoun μί:α ([mí:a] ‘one’) to introduce the female third per-
son. She describes the referent via the prepositional phrases <με γυαλιά, μ’ αγορέ
μαλλιά ([me ʝaʎá, maɣoré maʎá] ‘with glasses, with boyish hair’) and the rela-
tive clause που μιλούθε °λίγο έτσι:¿ ([pu milúθe °líɣo ét͡si:¿] ‘who was speaking
like this’). The gap after the first pair part shows possible trouble encountered
by co-participants in identifying the referent. At lines 4 and 5, Dimitris and
Tania initiate an insert repair sequence to resolve this trouble. Tania employs
the feminine adjectives ψηλή ([psilí] ‘tall’), χοντρή ([xodrí] ‘fat’) to describe the
referent and Zoi confirms her recipients’ understanding at line 6. At lines 9 and
10, Dimitris and Tania deliver second pair parts to Zoi’s first pair part (line 1),
showing that they have been able to identify the referent. Dimitris uses a recog-
nitional reference form, that is, the feminine proper name Άννα ([ána] ‘Anna’),
to identify the referent. At line 11, in a sequence-closing third, Zoi confirms that
this is the correct name of the referent and uses the feminine nominal demon-
strative αυτή ([aftí] ‘this’) to maintain reference to third person. However, the
pre-sequence is further expanded by Dimitris, who challenges this common
understanding and repeats the feminine proper name with rising intonation.
Tania tries to remember the referent’s proper name and uses the feminine
article η:::: ([i::::] ‘the’) at line 14. She gives additional information about the
referent at lines 17 and 18, keeping reference via the feminine personal pro-
noun αυτή (‘she’). At line 20, Tania asks her co-participants about the referent’s
proper name and re-refers to this specific person via the feminine personal
pronoun (clitic form) τη ([ti] ‘she’). Her co-participants do not deliver a second
pair part and, at line 23, Zoi moves to the first pair part that the pre-sequence
projected. She assumes that their classmate must have graduated and main-
tains reference to her via the feminine personal pronoun αυτή (‘she’).
In this conversational episode, participants make an effort to arrive at an
intersubjective understanding of the referent’s identity. The noun phrase μί:α,
114 chapter 6
.hh <με γυαλιά, μ’ αγορέ μαλλιά που μιλούθε °λίγο έτσι:¿> (‘one [girl] .hh with glasses,
with boyish hair that was speaking like this’) is an initial reference form which
consists of the feminine pronoun μί:α, and the feminine pronouns and articles
η, αυτή, τη are subsequent reference forms. Speakers use the feminine gram-
matical gender to keep track of the third person referred to in this sequence.
Linguistic items marked by feminine gender are taken to be co-referential, that
is, feminine subsequent reference forms select as their referent the same per-
son depicted initially by the feminine intitial reference form (cf. section 2.4).
These feminine items in addition to the feminine adjectives ψηλή and χοντρή
that speakers employ to describe the third person are gender membership
categories that ascribe female sex to the referent, categorize her as woman,
and assume feminine gender as part of the referent’s identity. According to
Schegloff (2007c, 125), subsequent reference forms are designed as ‘referring
simpliciter’, that is, doing reference and nothing more. In this episode, par-
ticipants employ feminine linguistic items as subsequent forms to maintain
reference to third person. Yet, these items make reference to third person func-
tionally complex, because they are grammatically feminine and, thus, index
and presuppose referent’s feminine gender. While participants attempt to
build an intersubjective understanding of the referent’s identity, they also con-
struct an intersubjective understanding of the referent as woman.
In Extract 12, Tania describes the first gynecologist that she visited in
Thessaloniki.
Extract 12 [Ι.20.Α.28.6]: Gynecologist

1 Τania Εγώ >εδώ στη Θεσσαλονίκη< πήγα πρώτη φορά
I here in Thessaloniki the first time I went
2 → σε γυναικολόγο ε ι- που: τον είχε [η μάνα] μου:, (0.5) και
((noise. . . . .))
prep gynecologist.acc.masc pron.rel he.acc have.3sg.
past the.nom.fem mother.nom my and
to a gynecologist.masc uh i- whom.masc my mother had,
(0.5) and
3 Dimitris [Πότε? ]
When?
4 Tania την^ ↑γκζεγέννησε. (0.5) απίστευτο? και [πήγα εδώ πέρα,]
((in a laughing tone)) ((in a laughing tone. . . . . . . . . . . .
((noise. . . .))
he delivered her baby. (0.5) unbelievable? and I went over
here,
5 Melita [ °Απίστευτο. ]
Unbelievable.
6 Tania → μετά από τόσα χρόνια, ((γελάκι)) αυτός ακόμα είναι
. . . . . . . . . . . . .)) ((laughing))
after prep so.many.acc.neuter years.acc.neuter he.nom
still cop.3sg.pres
after so many years, ((giggle)) he is still
7 → γιατρός, ήταν νέος τότε °βέβαια,° .hh >τώρα πια είναι<
((in a laughing tone)) ((laughing . . . .
((lighter sound and noise))
doctor.nom.masc cop.3sg.past young.nom.masc then cer-
tainly now no.longer cop.3sg.pres
a doctor.masc, of course then he was young.masc, .hh now
[he] is
8 → γέρος, .hh ((γελάκι)) .hh ένας τρελό::ς: γυναικολόγος
. . . . . . .)) ((in a laughing tone. . . . . . .
old.nom.masc one.nom.masc crazy.nom.masc gynecolo-
gist.nom.masc
old.masc, .hh ((giggle)) .hh a.masc crazy.masc
gynecologist.masc
9 → ο οποίος έχει: αφίσες του Τσε: όταν είσαι εκεί
. . . . . . . . . . . . . . . . . . . . . . . . . . .)) ((laughing . . . . . .
the.nom.masc who.nom.masc have.3sg.pres posters.acc.
fem the.gen.masc Che when cop.2sg.pres there
who.masc has posters of Che when you are there
10 στη::[ν τέτοια και περι]μένεις. ((γελά[. . . .)) ναι.] ναι:. .hh
. . . . . . . . . . . . . . . . . . . . . . . . . . . .)) ((laughing))
in that one [the waiting room] and wait. ((she laughs)) yes.
yes. .hh
11 Melita [((she laughs . .))]
12 Dimitris [( . . . . . . . )]
In her telling, Tania uses the noun γυναικολόγο ([ʝinekolóɣo] ‘gynecologist’) for
non-recognitional initial reference to third person and keeps track of the refer-
ent via the personal pronouns τον ([ton] ‘him’) (clitic form) and αυτός ([aftós]
‘he’), the noun phrase ένας τρελό::ς γυναικολόγος ([énas treló::s ʝinekolóɣos] ‘a
crazy gynecologist’) and the relative pronoun ο οποίος ([o opíos] ‘who’). She
also uses the noun γιατρός ([ʝatrós] ‘doctor’) and the adjectives νέος ([néos]
‘young’), γέρος ([ʝéros] ‘old’) and τρελό::ς ([treló::s] ‘crazy’) to describe the gyne-
cologist. All linguistic items mentioned above are inflected for the masculine
116 chapter 6
grammatical gender and, thus, constitute gender membership categories,

which categorize the referent as man. The masculine gender is not relevant for
what the speaker is doing in interaction but is indexed via the masculine gram-
matical gender and, thus, tacitly presupposed.
In Extracts 11 and 12, participants refer to third person in the service of social
actions that are not related to sociocultural gender. However, the compulsory
use of grammatical gender in the composition of turns has implications for the
routine meanings produced in these conversational episodes. The information
of feminine or masculine gender is part of the covert assumptions that partici-
pants share about referents.
6.2.4 Grammatical Gender and the Relevance of Sociocultural Gender

So far I have analyzed conversational episodes in which sociocultural gender is
not a relevant category for what participants do in interaction. In this section,
I examine conversational episodes in which participants orient to their own or
other peoples’ gender identities. The question I explore is what happens when
participants orient to gender, that is, when the information of gender that is
made available linguistically, is also foregrounded interactionally.
In the next extract, Melita, Evagelia, Zoi, Tania and Dimitris make plans
about their career after graduation from school.
Extract 13 (Ι.20.A.28.1): Μilitary service

1 Melita =Σε περίπτωση που τελειώσετε τη σχολή τι σκέφτεστε
If you finish school what are you thinking
2 °να [ κάνετε]:?=
of doing?
3 Dimitris [((giggle?))]
4 Evagelia =((giggle))
5 Melita =Ι[διωτ- ]
Priv-
6 Dimitris [>(Το μα][λάκα.)]=
((in a laughing tone))
(Playing dumb.)
7 Evagelia → [Να γί]νουμε [<πωλή:τριες, σερβιτό:ρε::ς,>]
Becoming saleswomen, waitresses,
8 Tania [°(Το μαλάκα λέει.) ((γελά ]. . .))
(Ηe says that he will play dumb.)
((she laughs))
9 (0.7)
((sound during the silence))
10 Tania [ Ντελίβερι:, ]
Delivery.
11 Ζoi [>Εσύ θα πας στο στρατό<] °φαντάζομαι.
I guess that you will go into the army.
12 Dimitris Βασικά. >πρέπει οπωσδήποτε.<
Basically. definitely I have to.
13 (1.6)
14 Ζoi °Είναι κι αυτό που έχουνε (δυ[στυχώ:ς).]
(Unfortunately) they also have to do this.
15 Τania [ Πο:]πο,
Wow,
16 [<αίσχο::ς.> ((γελά . . . . . . . . . [. . . . . . . . . .] . . . . . . . . . . . .]=
[That is a] disgrace. ((she laughs))
17 Ζoi [>Ε δεν^ μπειράζει. ας έχουμε και κάτι:.]
Uh that’s fine. let us have something as well.
18 Melita → [>↑(Κι εδώ.)< ας είμαστε]=
and here particle cop.1pl.
pres
Here as well. let us be
19 Τania =[. . . . . . . . . . . .[. . . . . . . . .]. . . . . . . . .]. . .)) ε ναι.
Eh yes.
20 Μelita → =[και κάπου τυχερές. ]
and somewhere lucky.nom.fem.pl
lucky.fem in certain things as well.
21 Ζoi [ Ε ναι. αυτό. ]
Eh yes. exactly.
At lines 1–2, Melita asks her co-participants about their plans after graduation,
and at lines 5, 6, 7 and 10, Dimitris, Evagelia and Tania deliver their answers to
the question. Dimitris replies with a joke, saying that he is thinking of playing
dumb, and Evagelia introduces a collectivity via the first person plural verb
να γίνουμε ([na ʝínume] ‘we become/becoming’). This collectivity includes
the speaker and other female co-participants, as it is shown by the feminine
nouns πωλή:τριες ([polí:tries] ‘saleswomen’) and σερβιτό:ρε::ς ([servitó:re::s]
‘waitresses’) that the speaker employs to describe the collectivity. At line 11,
118 chapter 6
Zoi self-selects as next speaker and introduces military service as what she and
her co-participants know that Dimitris will do after graduation. She addresses
Dimitris by employing the second person singular pronoun εσύ ([esí] ‘you’)
and the second person singular verb θα πας ([θa pas] ‘you will go’), and by refer-
ring to military service, which is a category-bound activity (Sacks 1972b) ste-
reotypically associated with men in Greece. Greek males between the ages of
19 and 45 are required by law to perform military service for at least 9 months.
Thus, by referring to military service Zoi limits the range of next speakers who
can produce a relevant second pair part to Dimitris only. At line 12, Dimitris
confirms that he needs to go to the army. Referring to military service, which
is a category-bound activity, introduces the relevance of the category ‘man’ in
interaction (Schegloff 2007b), as the following turn shows (line 14).
After a gap, Zoi self-selects again as next speaker and produces the first
pair part of a post-expansion (line 14). She asserts that unfortunately men are
obliged to fulfill their military service duties. She refers to this obligation by
employing the deictic pronoun αυτό ([aftó] ‘this’). The anaphoric use of the
pronoun serves as a ‘tying structure’ (Sacks 1995, 540), connecting Zoi’s turn
with the preceding sequence. Zoi refers to men as a group via the third per-
son plural verb έχουνε ([éxune] ‘they have’) and expresses her sympathy or
empathy (Ruusuvuori 2012) for men via the lexical choice δυστυχώς ([ðistixós]
‘unfortunately’). At lines 15–16, Tania supports Zoi’s affective stance by assess-
ing men’s obligation to perform military service in a negative way. However,
at line 17, Zoi takes a different affective stance: she mitigates Tania’s negative
assessment (first tcu) and expresses a wish/desire (second tcu). She intro-
duces the wish via the particle ας and refers to a collectivity, which includes
herself, her female co-participants and all women in general, via the first per-
son plural έχουμε ([éxume] ‘we have’). Similar to English first and third person
plural pronouns we and they which can refer to organizational collectivities to
which the speaker belongs (Sacks 1995, 713–714), the first and third person plu-
ral verbs in these turns refer to two different gendered groups ‘we women’ vs.
‘they men’. Zoi wishes that she and all women had at least some sort of privi-
lege, implying that usually men are in a more privileged position than women.
Melita agrees with Zoi at lines 18 and 20, by employing a similar format. She
introduces a wish via the particle ας [as], refers to self collectively via the first
person plural verb είμαστε ([ímaste] ‘we are’), and describes the collectivity via
the feminine adjective τυχερές ([tiçerés] ‘lucky’). She wishes that women were
lucky in some ways, implying that women are not lucky in general or that they
experience a greater number of difficulties compared to men. Thus, Melita
produces an affiliative response (Lindström and Sorjonen 2012), that is, she
supports Zoi’s lack of empathy for men’s obligation to perform their military
duty, and endorses Zoi’s desire for more justice or equality for women. The
grammatically feminine item τυχερές, which ascribes female sex to the collec-
tivity and categorizes referents as women, foregrounds gender interactionally.
Tania and Zoi agree with Melita at lines 19 and 21.
In the next extract, Melita, Tania, Zoi, Evagelia and Dimitris discuss the
differrences between girls and boys in the period of adolescence. In the lines
preceding this extract, Tania and Melita tell stories about female friends who
helped boys with their school homework and were sexually harassed by them.
Melita’s story is utilized as a resource for generating further topic talk (Jefferson
1978) on the sexual behavior of boys in childhood and adolescence.
Extract 14 [I.20.A.28.6]: Adolescence

1 Μelita Παιδάκι ήταν >°ρε παιδί μου.< το παιδάκι
It was a little kid. the kid
2 [που (έκανε μάθημα).°]
that [she] was teaching.
3 Dimitris [>°Ε ναι. να γυρίσεις πίσω] στα χρόνια.
Uh yes. you should turn back in time.
4 να δεις πώς [ήσουνα κι εσύ.°<]
and see how you behaved as well.
5 Zoi [ Για να το ] καταλά- °εννοεί ρε παιδί=
In order to underst- he means
6 Tania °Α ναι.
Ah yes.
7 (0.8)
8 Μelita Ξες. είναι εκείνη η ηλικία που αν θέλει να ανακαλύ:ψει,=
You know. it is in that age that if he wants to discover,
9 Εvagelia =[Και να δει πώς εί:ναι:.]
And see how it is.
10 Τania =[Ε σιγά. oι γυναί:]κες, [τα κορί]τσια μάλλον δε:ν^
Uh that is unlikely. women, or girls do not
11 Μelita [°Ναι:. ]
Yes.
12 Τania μπερνάμε τέτοια φάση:, (1.6) για
go.1pl through such a phase. (1.6) so that we [are in a position to]
13 [ να καταλά:βουμε:. ]
understand.
14 Dimitris [Κοίταξε. ίσως περνάτε αλλά] δεν^ ντην εκδηλώνετε
Look. maybe you.PL go through such a phase but you.pl do not
show it
120 chapter 6
15 όσο την εκδηλώνουμε εμείς °ας πούμε.° εμείς είμαστε

in the same way as we do so to speak. we are
16 [πιο: έτσι]
more like this
17 Tania [ Τόσο: ] ωμά: ↓δεν^ ντη- δεν^ μπιστεύω ότι ντην^
adv bluntly neg pron.acc.fem neg believe.1sg.pres conj
pron.acc.fem
I don’t- I don’t believe that
18 → μπερνά[ει καμία:.]
go.through.3sg.pres anyone.nom.fem
anyone.fem goes through [this phase] so bluntly.
19 Ζoi? [((she coughs))]
20 (0.5)
21 Μelita Εγώ νομίζω ότι: στ’ αγό:ρια είναι <πιο φυσιολογικό> κι
((Ζoi sniffes?. . . . . . . . . . . . . .))
I think that with boys it is more normal and
22 αναμενόμενο να εκδηλώσουν τέτοια συμπεριφορά και λες
expected to have this kind of behavior and you say
23 να το αγόρι μου έγινε άντρας και:: .hh ↑ενώ στο κορί:τσι:,
here it is the boy who became a man and .hh while the girl,
24 (0.9) δε θέλει να δείξει ότι έγινε γυναί:κα.
(0.9) does not want to show that she became a woman.
25 (0.5)
26 Μelita °Εγώ αυτό πιστε:ύω.° δηλαδή το κορίτσι μεγαλώνει το στήθος του
That’s what I believe. that is, for the girl her breasts grow
27 → ξέρω γω:, °στις τυχερές,° [((γελά)) και: ]
know.1sg.pres I in.the.acc.fem.pl lucky.acc.fem.pl and
I don’t know, with the.fem lucky.fem ones ((she laughs)) and
28 Τania [[((she laughs]. . . .]=
29 Ζoi [((she laughs. . . .))]=
30 Τania =[. . . . . . . . . . . .))]
31 Μelita =[Και: νιώθει κά]πως ά:βολα και λοιπά::, και
And she feels kind of uncomfortable etc., and
32 [ α: α:, ↑ενώ το α]γό:ρι να βγάλει μουστάκι
a:h a:h, but the boy will grow a moustache
33 Τania [Ναι ρε. εμείς ντρεπόμασταν.]
Yes. we felt ashamed of [growing up].
Melita foregrounds the age of the boy as an important factor in the incident
reported (lines 1–2), by using the formulation παιδάκι ([peðáci] ‘little kid’).
Dimitris agrees with Melita at lines 3–4 and prompts her to think about her
own behavior when she was a kid. Melita foregrounds age again as the key
reason for the boy’s behavior at line 8. Τania disagrees with Dimitris and
Melita (lines 10, 12 and 13) and argues that girls behave differently. Initially,
she refers to women by deploying the gender membership category γυναί:κες
([ʝiné:ces] ‘women’). Then, she switches to the gender membership category
κορίτσια ([korít͡sça] ‘girls’), foregrounding both age and gender. She uses the
first person plural verb μπερνάμε ([bernáme] ‘we go through’) to refer to herself
collectively. The collectivity includes the speaker, her female co-participants
and all women in general. In her turn, Tania differentiates girls from boys and
employs linguistic items lexically marked for female sex to foreground femi-
nine gender interactionally. Dimitris disagrees with Tania on whether these
gender differences exist or not at lines 14, 15 and 16. He refers to his female
co-participants and all women via the second person plural περνάτε ([pernáte]
‘you.PL go through’) and εκδηλώνετε ([ekðilónete] ‘you.PL show/manifest’), and
he refers to himself collectively via the first person plural verbs and pronoun
εκδηλώνουμε ([ekðilónume] ‘we show/manifest’), είμαστε ([ímaste] ‘we are’),
and εμείς ([emís] ‘we’). Thus, by employing first person plural, he introduces
a collectivity that is different from the collectivity of women and includes
men only.
Tania disagrees with Dimitris, at lines 17–18, and insists that no woman
experiences her sexuality in such a blunt (or brutal) way. She employs the
grammatically feminine indefinite pronoun καμία: ([kamía:] ‘anyone’) for
non-recognitional reference to women. At the same time, by deploying this
gender membership category, the speaker foregrounds sociocultural gender
interactionally. In the following turns, Melita highlights gender differences in
adolescence and employs gender membership categories to foreground gen-
der interactionally. These categories include linguistic items in which sex is
lexically specified, such as αγό:ρια ([aɣó:rʝa] ‘boys’), αγόρι ([aɣóri] ‘boy’), άντρας
([ándras] ‘man’), κορί:τσι: ([korí:t͡si:] ‘girl’), γυναί:κα ([ʝiné:ka] ‘woman’), κορίτσι
([korít͡si] ‘girl’), αγό:ρι ([aɣó:ri] ‘boy’), γυναίκα ([ʝinéka] ‘woman’), or grammati-
cized, such as τυχερές ([tiçerés] ‘lucky.fem’).
In Extract 15, Panayotis together with his two female friends, Danai and
Adriana, make plans for a party at Panayotis’ house.
Extract 15 (Ι.15.Α.24.1): Organizing a party

1 Panayotis [↑Θα κάνω πρόγραμμα] ε. θα βάλω
Well I will make a dj program uh. I will ask
122 chapter 6
2 τους σκύλους τους άλλους

the.masc dogs.masc the.masc other.masc guys
3 [να ψήνουν, >ο Στάθης κι αυτοί θα ψήνουν,]
to grill, Stathis.masc and the rest.masc will grill,
4 Αdriana [Αυτά μπορούμε να °τα (πάρουμε).]
We can (take) these [ones].
5 Panayotis θα κάνω εγώ πρόγραμμα [θα ’χουμε< .h δυο χιλιάδες ]
I will make a dj program we will have .h two thousand
6 Adriana [Εμείς θα ’μαστε στην^ γκουζί:να,]
((in a laughing tone. . . . . . . . . . . .))
We will be in the kitchen,
7 Panayotis σιντί ας πούμε. ’ντάξει. θα φέρουν όλοι.
cds so to speak. okay. everyone will bring [one].
8 [.hh >θα φέρεις< εσύ, θα πάρουμε]
you will bring one, we will take
9 Danai → [Εμείς στην^ γκουζίνα, οι κατίνε][:ς, ]=
((in a laughing tone . . . . . . . . . . . . . . . . . .
we in.the.acc.fem kitchen.acc.fem the.nom.fem.pl
gossipers.nom.fem
We will be in the kitchen, the gossipers.fem,
10 Αdriana [Στην^]=
In the
11 Danai =[τάχα μα]γειρε[ύουν, ]
. . . . . . . . . . . . . . . . . . . . . .))
adv cook.3pl.pres
pretending to cook,
12 Αdriana =[γκουζίνα,]
kitchen,
13 Panayotis [Θα [φέ]ρεις ε:]: ε θα φέρεις εσύ ρε παιδί μου
You will bring uh:: uh you will bring
14 Αdriana [με ποδιέ:ς,]
with aprons
15 Panayotis σιντί, θα φέρεις εσύ, θα φέρει ο Θάνος που έχει και
cds, you will bring, Thanos.masc will bring who also has
At lines 1, 2 and 3, Panayotis refers to his male friends, by using the grammati-
cally masculine reference forms τους σκύλους (tus skílus] ‘the dogs’),5 τους άλλους
5 The noun denotes metaphorically a person who listens to low quality folk music.
([tus álus] ‘the others’), ο Στάθης ([o stáθis] ‘Stathis’), αυτοί ([aftí] ‘they’). He
associates referents with the stereotypically masculine activity of grilling in
parties. At line 6, Adriana uses the first person plural pronoun εμείς ([emís] ‘we’)
to introduce a new collective referent, which includes the speaker and is dif-
ferent from the referents already introduced by the previous speaker (Pavlidou
2012a). She locates this new collectivity in the kitchen, which is linked with the
stereotypically feminine activity of cooking. The speaker makes the category
‘woman’ interactionally relevant by referring to this category-bound activity. At
lines 9 and 11, Danai keeps reference to the collectivity of women, by repeating
the first person plural pronoun εμείς, and locates referents in the kitchen. She
describes referents by using the verb phrase τάχα μαγειρεύουν ([táxa maʝirévun]
‘pretending to cook’), which refers to the stereotypically feminine activity of
cooking, and the noun phrase οι κατίνε:ς ([i katíne:s] ‘the.fem gossipers.fem’).
This noun phrase consists of a grammatically feminine head noun and modi-
fier, which categorize referents as women. In addition, the noun κατίνες is a
pejorative term denoting uneducated women who like to gossip (Dictionary of
Standard Μodern Greek 1998). Danai’s deployment of these gender member-
ship categories is related to her orientation to her own and her co-participant’s
feminine identity. In this conversational episode, the category ‘woman’ is fore-
grounded interactionally and a collective feminine identity is constructed via
referring to cooking and deploying the first person plural pronoun εμείς and
the grammatically feminine noun κατίνες, which is also semantically marked
for female sex.
In Extracts 13, 14 and 15, linguistic items marked by grammatical gender are
part of the tools that participants employ for foregrounding gender interac-
tionally and constructing gender identities. In all three cases, the information
of feminine gender, which is made available in interaction, becomes relevant
for what participants do in interaction.
6.2.5 The Non-match between Grammatical Gender and Referents’ Sex

There are certain cases in interaction in which the match between grammati-
cal gender and referent’s sex is not found in person reference. These are cases
in which the masculine grammatical gender is used for reference to female
referents only. In this section, I examine how the relation between grammati-
cal gender and culture manifests itself in interaction when there is a mismatch
between grammatical gender and sex.
In Extract 16 (part of this conversational episode was examined in Extract 8),
Dimitris asks his female co-participants about the profession they would choose
if they were about to finish highschool now (lines 2, 4). At lines 87–88, Dimitris
summarizes the answers delivered by his co-participants in the lines omitted.
124 chapter 6
Extract 16 [Ι.20.Α.28.1]: After school

1 Ζoi? [((giggle))]
2 Dimitris [ Άμα]:: ήσασταν τώρα τρίτη λυκείου τι θα κάνατε?
If you were now in sixth grade of highschool what would you
do?
3 (0.5)
4 Dimitris Και >σκεφτόσασταν το μέλλον σας?<
And you were thinking about your future?
86 Εvagelia Μ’ αρέσει ως επάγγελμα αλλά: [>°τι να πεις.<]
I like it as a profession but what can you say.
87 Dimitris [ Πάντως ] βλέπω ότι
Anyway I see that
88 → όλοι: >θα συνεχίζατε στο πανεπιστήμιο.
all.nom.masc fut continue.2pl.imperf in.the.acc.neuter
university.acc.neuter
you would all.masc continue studying at the university.
89 [εγώ δε θα συνέ]χιζα στο πανεπιστήμιο.<
I wouldn’t continue studying at the university.
90 Εvagelia? [(( giggle . . . .))]
91 Μelita Αλλά τι θα ’κα[νες?]
((cup sound . .))
But what would you do?
Dimitris refers to his co-participants via the second person plural verb θα
συνεχίζατε ([θa sineçízate] ‘you would continue’), and describes them via the
masculine adjective όλοι: ([óli:] ‘all’). That is, he uses the masculine gram-
matical gender for reference to female persons only, based on the inference
of human/universal associated with the use of the masculine gender. The mis-
match between grammatical gender and sex does not create any problems in
interaction and remains unnoticed by participants.
A similar case is found in Extract 17. Natasa delivers a story telling about a
movie screening at a film festival that she attended together with her female
friend Amalia. Natasa uses the masculine gender in reference to female per-
sons only.
Extract 17 [I.20.A.29.1]: Movies

1 Νatasa .hh και: πήγαμε με την Αμαλία. και- (.) ↑δεν βρίσκαμε
.hh and we went together with Amalia.fem. and- (.) we
didn’t find
2 ώρες ρε παιδί μου να μας βολεύουνε και: βρήκαμε μια

convenient hours and we found [a movie]
3 Γαλάζιο Φόρεμα. λέμε μ: .h α: ωραίο ακούγεται
the Blue Dress. we say mm: .h a:h it sounds good
17 Natasa και τελικά παιδιά είχαμε πάει σε μια ταινία .h (.) που: (.)
and hey guys finally we went to a movie .h (.) in which (.)
18 όλοι οι θεατές ήτανε: (.) τέτοιοι >ρε παιδί μου< γκέι,
((laughing. . .))
all the people in the film audience were (.) such gay,
19 ? °Α[:::. ]
A:::h.
20 Natasa [τς όλοι ] μα όλοι, και εμείς εκεί μέσα σαν τα χαζά
((in a laughing tone..
all.nom.masc conj all.nom.masc and we there inside
like the.acc.neuter.pl fool.acc.neuter.pl
all but all of them, and we were in there like fools so
21 → ας πούμε, .h άσχετοι τελείως .h=
. . . . . . . . . . . . . . . . . . . . . . . . . . . .))
so.to.speak.1pl disconnected.nom.masc.pl completely
so to speak, .h [ feeling] completely out of place.masc .h
22 Chrysanthi =και το [κοινό ήταν γκέι?]
the people in the audience were gay as well?
At line 1, Natasa uses the first person plural πήγαμε ([píɣame] ‘we went’) to
refer to a collectivity that includes herself and her friend Amalia. At line 21,
she describes the collectivity via the masculine adjective άσχετοι ([ásçeti] ‘out
of place’), which produces the inference of human/universal. The non-match
between the masculine gender and the female sex of referents is not consid-
ered to be a problem by participants. On the contrary, it is used as a tool for
the interactional work that the speaker is doing at that particular moment:
describing referents in a story.
In both Extracts 16 and 17, speakers refer to self or recipient, that is, to indi-
viduals who are physically present in context and can be easily identified given
the knowledge that participants share about the ‘here-and-now’ of interaction.
Second, speakers refer to more than one referent. This seems to motivate the
use of the masculine gender in these extracts. Speakers utilize the inference of
human/universal associated with the masculine gender in generic reference
to achieve reference to more that one female referents. Third, in both extracts,
gender is not relevant interactionally. One could assume that the lack of
126 chapter 6
articipants’ orientation to gender ‘neutralizes’ somehow the indexical effect

p
associated with grammatically masculine items and allows the use of these
items in reference to female persons only. However, as the analysis of the next
extract shows, the mismatch between grammatical gender and referents’ sex
can also be found in cases in which participants orient to gender.
In Extract 18, participants talk about gender differences in adolescence
(see Extract 14), and focus on the guilt that women experience regarding their
bodies.
Extract 18 [Ι.20.Α.28.6]: Adolescence (continuing)

1 Τania Αλλά αυτό που λες ότι νιώθεις ντροπή για το σώμα σου
But what you say that you feel ashamed of your body
2 καθώς μεγαλώνει:ς,=
as you grow up,
3 Μelita Μ::,
Μm,
4 Τania Μας έχουν γεμίσει ενοχές. ((γελάκι))
((in a laughing tone . . . . .))
They have filled us with guilt. ((giggle))
5 (0.6)
6 Ζoi [>Αυτό με το (σουτιέν] το< νο:-) ↑κι εγώ
This thing with the (bra the nu-) I also
7 Dimitris [ Γιατί ρε παιδιά? ]
Why guys?
8 Ζoi το ’νιωθα. [( . . . . . .)]=
felt it.
9 Tania [Ναι: ρε.]=
Yes.
10 Evagelia =((she coughs))=
11 Tania =[ Είναι ντροπή ] η περίοδος, [είναι]=
Menstruation is a shame,
12 Zoi =[>Εγώ πρώτη φορά<]
The first time I
13 Evagelia [Ναι:.]=
Yes.
14 Tania =[ ντροπή το: στή:]θο:ς, [είναι ντρο]πή
breasts is a shame, it is a shame
15 Evagelia =[↓όταν σου ’ρχεται περίοδος.]
when you have your menstruation.
16 Melita [ Είναι:
]
It is
17 Tania η τρίχα:,=
[to have] hair,
18 Μelita =Εί[ναι η κοινω]νία τέ:τοια:.
((noise starts))
Society is like that.
19 Τania [((giggle ))]
20 (.)
21 Μelita → Και θα είσαι τυχερός άμα είσαι σε μία κ- ε: οικογένεια::
and fut cop.2sg.pres lucky.nom.masc if cop.2sg.pres in
a.acc.fem family.acc.fem
And you will be lucky.masc if you are in a k- u:h family
22 φυσιολογική. .hh σκέψου να ’ναι ο πατέρας σου κανένας
[that is] normal. .hh imagine if your father was
23 <αυστηρό:[:ς,> °άντε-°] (0.5) °παναγία μου.
strict, get away- (0.5) my goodness.
24 Τania [ °Ναι ναι. ]
Yes yes.
At lines 1, 2 and 4, Τania refers to herself as part of the larger category of

women via the generic second person singular verbs and pronouns λες ([les]
‘you say’), νιώθεις ([ɲóθis] ‘you feel’), σου [su] ‘your’, μεγαλώνει:ς ([meɣalóni:s]
‘you grow up’), and the first person plural pronoun μας ([mas] ‘us’). This col-
lectivity includes herself, her female co-participants and all women in general.
By introducing this gender-relevant topic and referring to herself as part of a
gendered collectivity, the speaker orients to her gender identity. In the next
turns, co-participants make claims about the guilt that is usually associated
with female bodies. At lines 21, 22 and 23, Melita comments on the role that
the family and a strict father play in this process. She uses second person sin-
gular verbs for generic self-reference, i.e. θα είσαι ([θa íse] ‘you will be’), είσαι
([íse] ‘you are’), σκέψου ([sképsu] ‘imagine’), and she describes the referents via
the masculine adjective τυχερός ([tiçerós] ‘lucky’). Here reference is limited to
female persons only given that feminine identity is relevant for what partici-
pants do in interaction. Still, the masculine grammatical gender is used. The
interactional relevance of gender in this conversational episode does not seem
to create any trouble with respect to the non-match between the masculine
grammatical gender and referents’ female sex.
In the three extracts examined in this section, female participants ‘see’ and
identify themselves via grammatically masculine linguistic items. The stereo-
typical assumption about man as the norm allows participants to infer the
128 chapter 6
category of human/universal and refer to more than one female referent. In

these extracts, masculine linguistic items operate as indexes of ‘humanness’
and ‘plural number’. They presuppose more than one referent in context and
produce the inference of human/universal via the metonymic link between
male sex and the norm. This inference is part of the routine meanings pro-
duced in interaction when participants use the masculine gender to refer to
female persons only and do various social actions.
6.2.6 Conclusions
The analysis presented so far shows that the compulsory use of grammatical
gender in the composition of turns has specific implications for the sociocul-
tural world that participants jointly build in interaction. Items marked by gram-
matical gender are obligatory gender membership categories which structure
and constrain participants’ practices. They enforce participants to categorize
referents—self, recipient and other—as women or men on the basis of social
hierarchy in automatic, compulsory and systematic ways, when they accom-
plish various social actions. These actions may be related or not to participants’
orientation to gender. In both cases, the gendered categorization of referents
and the inference of human/universal via the codification of male sex are part
of the routine, unremarkable and subtle meanings produced in interaction in
automatic ways, and are tacitly presupposed by participants.
When participants orient to gender (cf. section 6.2.4), they utilize the infor-
mation of referent’s gender that is made available in interaction via gram-
matical gender in order to construct the bipolar gender order. These social
processes associated with the use of grammatical gender take place even when
participants do not attend to gender as an interactionally relevant category. For
instance, the gendered categorization of referents occurs when participants
do an assessment refer to third person in a locally initial or subsequent posi-
tion or refer to more than one referent (cf. sections 6.2.1, 6.2.2, 6.2.3 and 6.2.5).
In the above cases, grammatically masculine and feminine items are gendered
linguistically but not interactionally. This fact has implications for the routine
meanings produced in interaction, given the indexical effect associated with
grammatically feminine and masculine items. Although participants show no
orientation to sociocultural gender, they assume gender as an aspect of refer-
ent’s identity. Overall, grammatical gender contributes to the process of ‘gen-
dering’ (Pavlidou 2011, 415) or ‘genderizing’ (Eckert and McConnell-Ginet 2003,
254) interaction. As participants take turns in interaction and accomplish vari-
ous aspects of their daily social life, they construct gender bipolarity and gen-
der asymmetry in routine and covert ways because of grammar.
6.3 Indications of the Cognitive Role of Grammatical Gender in

Interaction
In section 6.2, I examined conversational episodes in which the covert assump-

tions associated with the use of grammatical gender remain at the ‘background’
of interaction, as part of what is seen but not noticed by participants. However,
there are rare cases in interaction in which these covert assumptions associ-
ated with the use of grammatical gender are brought to the surface of the talk
and the cognitive role of grammatical gender is uncovered. In these cases, the
use of grammatical gender creates trouble in the interpretation of referent’s
sex, which is resolved through repair.
6.3.1 Repair: Introductory Remarks

Repair refers to a set of systematically organized practices through which par-
ticipants manage troubles in speaking, hearing and understanding (Hayashi,
Raymond, and Sidnell 2013; Kitzinger 2012). Through repair participants make
sure that “interaction does not freeze in place where trouble arises, that inter-
subjectivity is maintained or restored, and that the turn, and sequence and
activity can progress to possible completion” (Schegloff 2007a, xiv). The par-
ticular segment of talk, to which the repair is addressed, is called the trouble-
source or the repairable, and there are no restrictions with respect to what
segments of talk can be repairable items (Schegloff, Jefferson, and Sacks 1977).
Trouble may be caused by recipient’s failure to hear the speaker or by speaker’s
errors of grammar, word meaning and pronunciation (Schegloff 2007a, 100).
In Conversation Analysis, a distinction is made between initiating the repair
and solving it or ‘carrying it to completion’ (Schegloff 2007a, 101). A repair may
be self-initiated or other-initiated, depending on whether it is initiated by
the speaker of the trouble-source or by someone other than the speaker of
the trouble-source. It is divided into self-repair and other-repair, depending
on whether it is carried through by the speaker of the trouble-source or by
someone other than the speaker of the trouble-source. The difference between
self-initiated repair and other-initiated repair is illustrated with the following
examples. Arrow ‘a’ indicates the trouble-source, arrow ‘b’ indicates the initia-
tion of repair, and arrow ‘c’ indicates the repair solution.
Self-initiated repair [Ι.20.Α.28.5]

1 a, b, c → Τania Και::: μετά ήρ- >πέρασε η αδερφή μου,< και μείναμε μαζί.
And afterwards cam[e]- my sister dropped by, and we stayed
together.
130 chapter 6
In the above extract, the speaker initiates repair by cutting off her talk and
effects the repair by replacing a verb uttered in error (ήρ- [ír]) with another
verb (πέρασε [pérase]). However, in the next extract, repair is initiated by the
recipient, who is not able to identify the third person that the speaker refers to.
Other-initiated repair [Ι.14.Α.17.1]

1 Αreti Χθες το βράδυ ξες ποιος με πήρε τηλέφωνο?=
Do you know who called me last night?
2 Yanis =°Ποιος.
Who?
3 a → Areti Ο Χρήστος ο Νίκου (στις) τρεις η ώρα το βρά:δυ.=
((cup sound. . . . . . . . . . . . . . . . . . .
Christos Nikou (at) three o’ clock after midnight.
4 b → Yanis = >°Ποιος είν’ ο Χρήστος ο Νίκου.< =
. . . . . . . . . . . .)) ((noise. . . . . . .))
Who is Christos Nikou?
5 c → Αreti =Ο: ↑φίλος μου αυτός από: που: είχαμε βγει
This friend of mine from- with whom we went out
6 τις προάλλες για καφέ?=
for coffee the other day?
7 Yanis = >Τι ήθελε?<
What did he want?
The trouble-source is located in Areti’s news announcement at line 3. Yanis

initiates repair at line 4, by delivering the first pair part of an insert repair
sequence. Areti provides the repair solution at lines 5 and 6, by delivering the
second pair part of the insert sequence.
In general, in interaction there is a preference for self-repair (Schegloff,
Jefferson, and Sacks 1977). Speakers employ a set of practices for doing self-
repair that Sidnell (2010, 114) defines as ‘technology’ of repair. These practices
involve a) marking the onset and completion of repair, b) locating the repair-
able item(s), and c) performing some operation upon them (Sidnell 2010, 114).
Self-repair is usually initiated by a cut-off that interrupts the talk in progress,
and it can be indicated by sound stretches or other delaying productions (e.g.
‘um’) (Kitzinger 2012, 239). Self-repair may occur in the same tcu as the trou-
ble source or later than the same tcu, in the transition space after the possible
completion of a turn.
As Kitzinger (2012, 232) notes, self-initiated repair in same tcu is the “most
common type of repair”. In same tcu self-repairs, the repairable item is located
through the repetition of some part of the prior talk (‘framing’, Schegloff 2004
cited in Sidnell 2010, 115) before (‘pre-framing’) or after the trouble source
(‘post-framing’). Same tcu self-repairs may be accompanied by silences and
delays or the use of apologetic terms (e.g. ‘sorry’), repeats of the trouble-
source, multiple tries or self-talk (Kitzinger 2012, 239–241). As Schegloff (2013)
shows, self-repair in same tcu involves several operations, such as replacing
an item by another, inserting an item into the prior talk, deleting an item from
the prior talk, re-ordering words or phrases, searching names or places, paren-
thesizing, aborting, sequence jumping, recycling or reformatting. Self-repair
can be employed for various interactional uses. Besides fixing a problem in
speaking, hearing or understanding, self-repair can also “be used in the service
of the action the speaker means to be doing with the talk” (Kitzinger 2012, 242),
and, thus, “reveal the work involved in designing talk for the action the speaker
intends it to do” (Kitzinger 2012, 243). In Drew, Walker and Ogden’s words (2013,
93), self-repairs give analysts “access to the work of constructing a turn—they
bring to the interactional surface the work in which speakers engage in order
to construct the action”. Similar to same tcu self-repairs, self-repairs in the
transition space after the possible completion of a turn involve operations
such as replacing and ‘framing’ of repairable items (Kitzinger 2012, 244–246).6
Other-initiated repair is delivered in a sequence (insert or post-expansion),
which “suspends the ongoing action in which the participants are engaged”
(Kitzinger 2012, 249). Repair can be initiated by a party other than the speaker
via the deployment of open class forms, such as what?, huh?, pardon?, sorry?,
interrogatives, such as who?, when?, where?, or repeats of the trouble-source
turn and candidate understandings, such as resayings of the trouble-source
(Kitzinger 2012, 249). In both self-initiated and other-initiated repair, the pro-
gressivity of the talk is interrupted so that intersubjective understanding is
sustained. As I mentioned in section 5.3, progressivity is defined as the moving
from some element to a hearably-next-one with nothing intervening. In repair,
the new element intervening is inspected by participants for its import, that is,
for the ways in which it reaffirms or changes the understanding of the turn so
far or the understanding of the previous turn.
Sometimes speakers “correct” other participants’ talk without employing
the technology of repair. As Kitzinger (2012, 256) notes, in these cases trouble
in speaking, hearing or understanding is resolved without having the trou-
ble “rising to the surface of the talk as overt repair”. Jefferson (1987) defines
this form of ‘covert’ correction as embedded correction and distinguishes it
6 Self-repair can also occur in third turn after intervening talk from another between repair-
able and repair or when speakers display some sort of misunderstanding of prior turn (third
position self-repair) (Kitzinger 2012, 246–249).
132 chapter 6
from exposed correction in which the progressivity of the talk is interrupted.

Corrections consist of the production of “an item, which stands as a correction
of an item in another’s talk” (Jefferson 1987, 87). The following two extracts
(found in Jefferson 1987, 87 and 93) illustrate cases of exposed and embedded
correction.
Exposed correction [gj: fn]

Pat: . . . the Black Muslims are certainly more provocative
than the Black Muslims ever were.
→ Jo: The Black Panthers.
Pat: The Black Panthers. What’d I
Jo: You said the Black Muslims twice.
Pat: Did I really?
Jo: Yes you di:d, but that’s alright I forgive you,
Embedded correction [gts: ii: 60: st]

Ken: Well- if you’re gonna race, the police have said this to us.
→ Roger: That makes it even better. The challenge of running from
the cops!
Ken: The cops say if you wanna race, uh go out at four or five in the morning
on the freeway . . .
According to Jefferson (1987, 90), in exposed correction the ongoing activity

is interrupted and correcting becomes “the current interactional business”.
Correcting can be accompanied by ‘accountings’, such as complaining or for-
giving, which is found in the first extract above. Exposed correction involves
the following series:
1. A speaker produces some object (X).

2. A subsequent speaker produces an alternative (Y).
3. Prior speaker produces an alternative (Y), that is, s/he accepts the prof-
fered correction, or the same object (X), that is, s/he rejects the proffered
correction.
In contrast, in embedded correction the XYY or XYX series is embedded

into the ongoing talk with no attendant activities, or, in Jeffferson’s words
(1987, 95), “the utterances are not occupied by the doing of correcting, but by
whatever talk is in progress”.
In general, repair can provide analysts with indications for the ways in
which participants interpret talk in interaction. According to Levelt (1983),
self-corrections show speakers’ ability to monitor the speech that they pro-
duce, for instance their ability to control whether what they say is appropriate
for their communicative purpose or suitable given the context of the speech
event. Given the role that repair plays in resolving trouble in participants’
understandings, we can theorize repair as a ‘window’ to participants’ cognition
(cf. Schegloff 1991, section 5.3), and, thus, to the cognitive role of grammatical
gender in interaction.
In the data I examined, I found overall 6 cases of repairing grammatical gen-
der: 4 self-initiated self-repairs, 1 exposed correction and 1 embedded correc-
tion. The small number of repairs found in the corpus is not a surpise. Since
grammatical gender is inherent in language structure and, thus, used in an
automatic, systematic and obligatory manner in the composition of turns, it
is expected that speakers will rarely notice the cognitive role that this feature
of grammar plays. Yet, these rare cases of repair are of special interest for the
study of the relation between grammatical gender and cognition, because they
provide indications of the cognitive role of grammatical gender, as I show in
the next section.
6.3.2 Grammatical Gender as an Item of Repair

Ι start my analysis by focusing on cases of self-initiated self-repair in which
grammatical gender constitutes the trouble-source, and which occur in the
same turn as the trouble-source, i.e. in the same tcu or in the transition space
after the completion of the tcu. The self-repairs found involve two operations,
replacing and repeating.
In extract 19, at line 12, Areti initiates a story telling about the events that
take place outside the Big Brother tv show studio. She does a self-repair at
line 27.
Extract 19 (Ι.14.Α.17.1): Big Brother

1 Yanis =Ναι. κι είναι επί δέκα λεπτά: (0.6) °δεκαπέντε άνθρωποι,
Yes. and for ten minutes (0.6) there are fifteen people,
2 κοιτάνε.° (1.6) το δωμά:τιο. (2.4) °ησυχία,° και [ λέμε ]
they stare. (1.6) at the room. (2.4) silence, and we say
3 Spiros [((giggle))]
4 Yanis ρε παιδιά: να το αλλάξουμε >°να δούμε καμιά ταινία?
hey guys should we change it and watch a movie?
5 =°°κάτσε κάτσε. άστο άστο άστο τώρα.< περί[μενε. ]
wait wait. leave it leave it now. wait.
6 Spiros [((he laughs]. . . .
134 chapter 6
7 Areti [’Ντωμεταξύ ↑δεν έχει] <τίποτα.> [ κοι↑μού]:[νται.]

((in a laughing tone . . . . . . . . . . . . . . . . . . . . . . . . . . . .))
Μeanwhile nothing is happening. they are sleeping.
8 Spiros [. . . . . . . . . . . . . . . . . . . . .))]
9 Νiki? [((giggle))]
10 Yanis [>Δεν] έχει
There is
11 ρε παιδιά< τίποτα¿=
nothing guys.
12 Αreti =Παιδιά τι ↑πράμα έξω απ’ το Μπιγκ Μπράδερ¿
Hey guys what is this thing outside the Big Brother studio.
13 (0.6)
14 Αreti Ε[ντελώς ] ξεφτιλιζέ.
Totally humiliating.
15 Yanis [>Τίποτα.<]
Nothing.
16 (.)
17 Αreti Εντωμεταξύ ρε παιδί μου::, (.) °ε° τον^ ντελευταίο καιρό:, .h
Ιn the meanwhile, (.) uh lately, .h
((simultaneous talk))
18 >αυτοί-< επειδή βγαίνουν συνεχώς αυτοί: συνεχώς.
they- because they continuously get out of the game
continuously.
19 κάθε δύο μέρες, ε ↑διώχνουν και κά:ποιον, .hh και ↑πάντοτε:
every two days, uh they kick someone out, .hh and always
20 με το που:: βγαίνουν έξω, >πηγαίνουν εκεί τους ρω-<
as soon as they get out, they go there he ask- them
21 ο Μικρούτσικος τους ρωτάει κάτι <βλακείες> ας πούμε, όλοι
Mikroutsikos asks them this nonsense so to speak, they all
22 βγαίνουν και λένε α:: τς, .h ε: η- ε:μ:: θα το ξαναέκανα, κι
get out and they say a::h ts .h uh i- e:m:: I would do it again, and
23 ήταν φοβερή εμπειρία, τέλος πάντων, τώρα τι εμπειρία
it was an amazing experience, anyway, now
24 δε μπορώ °να φαντα- (φτε-) φανταστώ,° .h και >ρε παιδιά<
I cannot imag- imagine what kind of experience this is, .h and
guys
25 → ↑έρχοντ’ οι γονείς τους, οι οποίοι (.)
((noise))
come.3pl.pres the.nom.masc.pl parents.nom.masc their
the.nom.masc.pl who.nom.masc.pl
their parents come over, who (.)
26 → <τρέχουν κλαίγοντας> στα παιδιά, τους λένε

run.3pl.pres crying to.the.acc.neuter kids.acc.neuter
to.them say.3pl.pres
run to their kids crying, they tell them
27 → παιδί μου, είμαι πε↑ρήφανος, .h (.) περήφανος ή περήφανη
child my cop.1sg.pres proud.nom.masc proud.nom.masc
or proud.nom.fem
my child, I am proud.masc .h (.) proud.masc or proud.fem
28 για σένα. ↑και λέω για ποιο λόγο (είναι)/(να ’ν’)
of you. and I say for what reason (are they)/(should they be)
29 >περή[φανοι.< =αν ο γονέ]ας είναι περήφανος για
proud of. if a parent.masc is proud of
30 Νiki [ Γιατί. τι έκανε. ]
Why. what did s/he do.
31 Αreti το παιδί του, επειδή πήγε στο Μπιγκ Μπράδερ, ένας γονέας
his child because s/he joined the Big Brother show, a parent
32 που το παιδί του ξέρω γω σπούδασε, τι πρέπει να είναι.
whose child let’s say studied at the university, how should he
feel.
In the story telling, Areti refers to the tv show players and describes how they
get out of the studio when they leave the show and how their parents welcome
them. At lines 25–26, she refers to the parents of these players who wait for
their kids outside the studio, and run to them crying. Areti employs a direct
speech report construction (Aikhenvald 2011) to describe the parents’ behavior
(lines 26, 27, 28). According to Aikhenvald (2011, 400), speech reports involve
the “author” of the original speech and the “reporter”. The speech report con-
tent corresponds exactly (more or less) to what the author of the original
speech said. Areti uses the exact words that the original speakers, i.e. the par-
ents, used in the first place to express themselves παιδί μου, είμαι πε↑ρήφανος, .h
(.) περήφανος ή περήφανη για σένα. ([peðí mu, íme pe↑rífanos, .h (.) perífanos í
perífani ʝa séna.] ‘my child I am proud.masc .h proud.masc or proud.fem of
you’). Direct speech report is linked with speakers’ shifts in their role as anima-
tor and author, that is, in their role as the person whose voice is actually being
used to produce a strip of speech, and the person who actually constructed
the phrase said (Clift and Holt 2007; Goffman 1981). In the direct speech report
in this extract, the parents are the authors of the strip of speech reported and
Areti is the animator. The self-repair to be examined occurs within this direct
speech report.
136 chapter 6
Areti refers to the parents by using the masculine non-recognitional initial

reference form οι γονείς τους ([i ɣonís tus] ‘their parents’) and the masculine
subsequent reference form οι οποίοι ([i opíi] ‘who’). She keeps reference to the
parents by using the reporting third person plural verb λένε ([léne] ‘they say’),
which introduces the speech report content παιδί μου, είμαι πε↑ρήφανος (‘my
child, I am proud.masc’). Thus, Areti shifts from third person to first person
singular and introduces an ‘I’ that corresponds to one of the two parents, who
is the ‘author’ of this particular bit of talk. She describes the referent by using
a gender membership category, i.e. the grammatically masculine adjective
πε↑ρήφανος (‘proud.masc’), which categorizes the referent as male.
Before moving towards the completion of the tcu, after a breathe-in and a
micro-pause, Areti repeats the masculine adjective and inserts the feminine
one περήφανος ή περήφανη (‘proud.masc or proud.fem’). The trouble-source
here is the linguistic item marked by grammatical gender, i.e. the gender mem-
bership category. The self-repair occurs in the same tcu with the trouble-
source. It is pre-framed, that is, located through the repetition of the adjective
περήφανος before the repair. It involves the insertion of a gender membership
category. Through her self-repair the speaker categorizes the referent as either
male or female and guides recipients to identify the referent with the father
or the mother whom she introduced earlier in her turn via the formulation οι
γονείς τους (‘their parents’). In this way, the self-repair shows that the speaker
understood the referents denoted by grammatically masculine and feminine
items as male and female respectively. Thus, the self-repair can be interpreted
by the analyst as an indication of the correlation between grammatical gender
and the interpretation of referent’s sex.
What guides the speaker to notice what usually remains unseen and auto-
matic in interaction? Schegloff (2007a, 2) argues that “one basic and omnire-
levant issue for the participants for any bit of talk-in-interaction is ‘why that
now’ (Schegloff and Sacks 1973, 299), and the key issue in that regard is what
is being done by that (whatever the ‘that’ is)”. The same question can be raised
with respect to the self-repair delivered in this conversation episode. Why
does the speaker need to return to the linguistic item marked by the mascu-
line grammatical gender and insert a linguistic item marked by the feminine
gender? The problem that the speaker tries to resolve by repairing the gender
membership category seems to be related with the identification of the refer-
ent who is the author of the direct speech report. Initially, the speaker refers to
both parents but in the direct speech report she shifts reference to one party
of the group only. She categorizes the third person as male, and, thus, excludes
the other party of the group, who is female. Therefore, the self-repair aims to
ensure that reference to the same third person that was originally introduced
is kept in interaction. Another reason why categorizing referent as female
becomes important at this particular moment in interaction is the fact that the
speaker refers to parents crying. Crying is a category-bound activity, that is, an
activity stereotypically associated with the members of the category ‘mother’.
The rest of the repairs to be examined in this section involve the generic
use of the masculine. In extract 20, Melita, Tania, Zoi and Evagelia discuss the
problems that partners encounter in living together in the same house.
Extract 20 (Ι.20.Α.28.5): Cleaning the house

1 Μelita Το: ίδιο >μπορείς να< το:::: σκεφτείς και σ’ ένα: ζευγάρι.
You can think the same [thing] also with respect to a couple.
2 (0.7)
3 Μelita °Ένα ζευγάρι.
A couple.
4 (0.5)
5 Μelita Ναι μεν έχουν θάρρος και::: οικειότητα και λοιπά:, ↑αλλά δεν^
They may feel comfortable with each other et cetera, but
6 μπρέπει να χαθεί ο σεβασμός στον άλλον, και [το ότι]
((noise. . . .))
they should not loose respect for each other, and [keep in
mind] that
7 Τania [ Ναι.]
Yes.
8 Μelita πρέπει να κάνεις εσύ κάποια πράγματα κι αυτός
you have to do certain things and he must do
9 κάποια άλλα.
others.
10 (0.5)
11 Μelita >Για να μη< μαλώνεις.
So that you do not fight.
12 (0.9)
13 Μelita Δηλαδή να υπάρχει μί:α:: (2.3) ένα τυπικό ρε παιδί μου στο
((lighter sound . . .))
In other words there should be one (2.3) [there should be]
rules
138 chapter 6
14 στο σπίτι:.
in the house.
15 (0.7)
16 Μelita °Κατάλαβες?° >καταλάβατε [τι εννοώ?<] [((giggle))]
Do you.sg understand? do you.PL see what I mean?
17 Tania [ Μ:. ]
Mm:.
18 Ζoi [Ναι αλλ-] εγώ
Yes but- I
19 νομίζω κατάλαβα. αλλά εκεί είναι και πιο δύσκολο.
think I understood. but in that case it is more difficult.
20 °να κρατηθεί.
to keep it [that way].
21 (0.6)
22 Μelita Μ::.
Mm.
23 (0.9)
24 Ζoi Και όπως επίσης, δε: θα μπορέσεις- (0.6) όχι δε θα μπορέσεις.
And in addition, you will not be able- (0.6) it’s not that you
won’t be able [to do it].
25 θα μπορέσεις. ο άλλος δεν^ γκζέρω πως θα το δεχτεί, (.)
you will be able. I don’t know how the.masc other.masc is
going to deal with it, (.)
26 όπως ας πούμε εσύ θα πεις στην συγκάτοικό σου, ξέρεις τι?
for example you ’ll tell your flatmate.fem, you know what?
27 έλα να: καθαρίσουμε πιχί το σπίτι. έτσι? θα κάνω εγώ αυτό.
come on let’s clean the house for instance. right? I will do this.
28 εσύ [ το ] άλλο. (.) hh όταν θα το πεις στο:ν άλλον που
you will do that. (.) if you tell the other.masc with whom
29 Τania? [°Μ.]
Mm.
30 Ζoi θα μένεις μαζί σου, (.) >μπορεί να-< να γελά:σει, >μπορεί να
you live together, (.) he might laugh, he might
31 πει< (.) [ τι? ]
say (.) what?
32 Εvagelia [°Να] σε βρί:σει.=
He might insult you.
33 Τania? =((she laughs[..))]
34 Ζoi [Να] κάνω τα τζάμια, >πλάκα τώρα,< (0.6) δεν-

[That I should] clean the windows, are you kidding now, (0.6)
not-
35 [>δηλαδή α<]
in other words a
36 Εvagelia → [ Kαλά ] ναι. °ποιος το κάνει?=ποιος ά[ντρας?]
((in a laughing tone . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . .))
well yes who.nom.masc it.acc do.3sg.pres who.nom.masc
man.nom.masc
Yeah right. who.masc does it? which.masc man?
37 Ζoi [Και αυ]τό είναι
But this
38 είναι πολύ μεγάλη φθορά όμως για τη σχέση.
ruins the relationship a lot.
Melita initiates a sequence about the stereotypical roles of men and women
who are couples living under the same roof (line 1). In the following lines, she
argues that there should be certain rules with respect to how partners share
space, so that they do not loose respect for each other (δεν^ μπρέπει να χαθεί
ο σεβασμός στον άλλον, και [το ότι] πρέπει να κάνεις εσύ κάποια πράγματα κι αυτός
κάποια άλλα [ðen brépi na xaθí o sevazmós ston álon, ce to óti prépi na kánis
esí kápça práɣmata ci aftós kápça ála] ‘they should not loose respect for each
other, and [keep in mind] that you have to do certain things and he must do
others’, Δηλαδή να υπάρχει μί:α:: (2.3) ένα τυπικό ρε παιδί μου στο στο σπίτι: [ðilaðí
na ipárçi mía:: éna tipikó re peðí mu sto sto spíti:] ‘In other words there should
be one (2.3) [there should be] rules in the house’). At lines 24, 25, 26, 27, 28,
30, 31, 34 and 35, Ζoi argues that usually men are unwilling to clean the house
and undervalue such tasks (>μπορεί να-< να γελά:σει, >μπορεί να πει< (.) [τι?] [borí
na na ɣelá:si, borí na pi ti?] ‘he might laugh, he might say (.) what?’). Evagelia
agrees with Zoi at line 36. Her affiliative response consists of two tcus and one
increment. In the first tcu Καλά ναι. ([kalá ne.] ‘Yeah right.’), Evagelia agrees
with the previous speaker’s turn. In the second tcu °ποιος το κάνει? ([pçós to
káni?] ‘who.masc does it?’), she asks a rhetorical question, which further shows
her agreement, and then she delivers the increment =ποιος ά[ντρας?] ([pçós
ándras?] ‘which.masc man?’) with latching. In the increment, she does a self-
repair which concerns the preceding tcu. In her rhetorical question, Evagelia
uses the masculine interrogative pronoun ποιος ‘who’ for generic reference. In
the transition space after the grammatical and pragmatic completion of the
140 chapter 6
tcu, the speaker inserts the noun άντρας (‘man’), which is both grammatically
and lexically marked as male, and she pre-frames the repairable item, i.e. by
repeating the item ποιος.
Through her self-repair the speaker aims at making the third person’s male
sex explicit and, thus, solving or anticipating a possible problem with respect
to how repicients understand the turn. The generic use of the masculine in the
second tcu could guide recipients to interpret referent’s sex as indefinite or
not exclusively male. This interpretation would cancel the kind of agreement
that Evagelia’s rhetorical question is designed to do: Evagelia agrees with Zoi
that men do not participate in cleaning the house.
The analysis of this self-repair provides support for the argument presented
in previous chapter about the relation between the generic use of the mas-
culine and the inference of human/universal. More specifically, it shows that
speakers understand the masculine grammatical gender as a tool for achieving
generic reference to humans.
However, things are a bit more complicated when it comes to the generic
use of the masculine in interaction, as the following repairs show. The self-
repair to be examined in Extract 21 occurs at line 257. It comes after a series
of reciprocal or exchange sequences (Schegloff 2007a) in which participants
judge the similarity in appearance between co-participants and public figures,
such as singers, actors, etc. According to Schegloff (2007a, 195), reciprocal or
exchange sequences “refer to episodes in which a sequence which has just
been initiated by A to B (that is, in which A is the first pair part speaker and
B is the second pair part speaker) is then (after it has run its course) recipro-
cated—initiated by B to A”.
Extract 21 (I.20.A.28.2): Similarity in appearance between persons

1 Evagelia Στην αρχή ήταν ωραίο. °μετά,=
In the beginning [the tv series] was nice. afterwards,
2 Μelita =Γιατί ↑το ’κοψαν? [έτσι ξαφνικά.]
Why did they cut it? so suddenly.
3 Τania → [Όταν ήμουνα] στο [ γυμνάσι]ο: (.)
When I was in high school (.)
4 Εvagelia [((giggle))]
5 Τania → μου ’λέγαν [ότι μοιάζω] με μία που παίζει σ’ αυτό
they told me that I look like someone.fem who plays in this
[series]
6 Μelita? [((giggle))]
7 Τania → [και μ’ είχαν] [σταματήσει ] στο [↑δρό:]μο.=
and they stopped me on the street.
8 Ζoi [ Αχ ναι:. ]
Oh yes.
9 Dimitris [↑Η Χούλια.]
Choulia.fem.
10 Ζoi [Μια ]
One.fem
11 Ζoi =Ναι μοιά:[ζει:ς . ]
Yes you look like her.
110 (1.5)
((noise . . .
111 Τania Έτσι. >για να πάρω μια γεύση.<
Like this. in order to get an idea.
. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . .))
112 Μelita → >Η Ευαγγελία με ποια σας μοιάζει?< εμένα μου μοιάζει με
Whom.fem do you think Evagelia looks like? I think that she
looks
113 τη Νάνσυ Αλεξιάδου.
like Nancy.fem Alexiadou.fem.
114 (0.5)
115 Evagelia Ίδιες είμαστε.
((ironically))
We are the same.fem.
((noise))
157 (1.5)
158 Μelita → Εσάς με ποιους σας έχουν πει?
Whom.masc.pl did they tell you that you look like?
159 (0.6)
160 Τania °Με κανέναν. ((γελά[. . .)) είναι στη μόδα (. . . . . . . . . . . . . . )]
Like nobody.masc. ((she laughs)) it is in fashion ( . . . . .)
161 Melita [↑Τίποτα? κανείς? δεν [ γίνεται. ]]
Nothing? no one? it can’t be.
162 Ζoi [Με την αδερ]φή
Like my sister,
163 [μου, με τον μπαμπά]:: μου[: ]
like my father
164 Τania [((she laughs. . . . .))]
165 Dimitris [Με] [την Ελπίδα,]
Like Elpida.fem,
142 chapter 6
166 Εvagelia [ Με τη μα]μά μου,

Like my mum,
167 (0.8)
((lighter sound))
168 Ζoi Με την Ελπίδα λέει ο [(άλλος). ((γελά. . . . . .]=
((laughing. . . . . . . . . . .))
[Dimitris] says [she looks] like Elpida.fem. ((she laughs))
169 Τania [((γελά. . . . . . . .)) σου]=
((in a laughing tone . . .
((she laughs)) does
170 Zoi =[. . . . . . . . . . . . .))]
171 Τania =[μοιάζει στη φά]τσα?=
. . . . . . . . . . . . . . . . . . . . . . ))
[Zoi’s] face looks like [Elpida’s]?
172 Dimitris =Όχι.=
No.
249 Τania =[Ρε: ήτανε κακάσχημο:.]=
Man [the girl] was very ugly.
250 Εvagelia =[((she laughs. . . . . . . . .))]=
251 Ζoi = >°Η Χούλια.<
Choulia.fem.
252 (0.6)
253 Τania ((she laughs [. . . . . . .))]
254 Μelita [((γελά))] Ποια μιλάει? ((γελά[. . . . . . . . . .]. . .)) α:.
((laughing . . .)) ((laughing))
((she laughs)) Look who’s talking? ((she laughs)) a:h.
255 Τania [((giggle))]
256 (.)
257 Ζoi → Εσένα με ποιον σου είπαν ότι μοιάζεις? με ποια?
you.acc prep who.acc.masc.sg you.sg.gen tell.3sg.past conj
look.like.2sg.pres prep who.acc.fem.sg
Whom.masc did they tell you that you look like? whom.fem?
258 (0.6)
259 Melita ↑Μου έχουν πει με πολλές αλλά δεν σας λέω.
((in a laughing tone . . . . . . . . . . . . . . . . . . . . . . .
They told me that I look like many.fem but I won’t tell you.
In the lines preceding this extract, participants talk about ΤV series for chil-
dren. At lines 3, 5 and 7, Tania does stepwise topic transition (Jefferson 1984;
Sidnell 2010, 240–244), by asserting that when she was little other people told
her that she looks like Choulia, an actress in a Brazilian tv series for chil-
dren (μου ’λέγαν [ότι μοιάζω] με μία που παίζει σ’ αυτό [mu léɣan óti mɲázo me
mía pu pézi saftó] ‘they told me that I look like someone.fem who plays in
this series’). Zoi agrees with this assessment (lines 8, 11). Tania’s similarity to
Choulia is topicalized in the following lines that have been omitted. In these
lines, Dimitris agrees with the assessment, and Tania mentions that her mum
and her schoolmates also shared the same view. She also gives a description
of her look when she was a kid to justify this view. At lines 112–113, Melita asks
her co-participants whom they think Evagelia looks like (>Η Ευαγγελία με ποια
σας μοιάζει?< [i evagelía me pça sas mɲázi] ‘Whom.fem do you think Evagelia
looks like?), and she describes Evagelia as similar to the female singer Nancy
Alexiadou (εμένα μου μοιάζει με τη Νάνσυ Αλεξιάδου. [eména mu mɲázi me ti
nánsi aleksiáðu] ‘I think that she looks like Nancy.fem Alexiadou.fem’). Her
first pair part initiates a sequence (the lines are omitted) in which Evagelia
disagrees with Melita’s assessment. Melita initiates a new sequence at line 158,
asking Zoi and Dimitris whom other people told them that they look like (Εσάς
με ποιους σας έχουν πει? [esás me pçus sas exun pi?] ‘Whom.masc.pl did they
tell you that you look like?’). At lines 162–163, Zoi describes herself as similar to
her family members ([Με την αδερ]φή [μου, με τον μπαμπά]:: μου[:] [me tin aðerfí
mu, me ton babá:: mu:] ‘Like my sister, like my father’), and at line 165, Dimitris
describes Zoi as similar to the female person, Elpida ([Με] [την Ελπίδα,] [me
tin elpíða] ‘Like Elpida.fem’). At line 166, Evagelia describes herself as similar
to her mother ([Με τη μα]μά μου, [me ti mamá mu] ‘Like my mum’). In the
lines omitted, participants disagree with Dimitris’ assessment and comment
on actresses playing roles in these tv series. In general so far, participants
describe their female co-participants as similar to female persons, that is, simi-
larity assessment seems to be based on the sex of the person being assessed.
In this sense, assessing similarity in appearance between persons constitutes a
category-bound activity. In particular, it is an activity linked with the category
of female sex that introduces the relevance of female sex in interaction, as
I show in the following.
At line 257, Zoi addresses a question to Melita. Her turn consists of a tcu
Εσένα με ποιον σου είπαν ότι μοιάζεις? ([eséna me pçon su ípan óti mɲázis?]
‘Whom.masc did they tell you that you look like?’), which is grammatically
and pragmatically complete, and an increment με ποια? ([me pça?] ‘whom.
fem?’), which is designed as grammatical continuation of the tcu and deliv-
ers a self-repair. In the tcu, Zoi selects Melita as next speaker by using the sec-
ond person singular verb μοιάζεις (‘you look like’) and the pronouns εσένα, σου
(‘you’), and associates Melita with a third person. Zoi refers to the third person
144 chapter 6
by using a prepositional phrase, which consists of the grammatically mascu-

line relative pronoun ποιον. In the increment, the speaker does a self-repair
by replacing the masculine pronoun with the feminine one (ποια). Similar to
the self-repair examined in Extract 20, the trouble-source in this extract is the
gender membership category, i.e. the linguistic item marked by the masculine
grammatical gender. The self-repair occurs in the same turn with the trouble-
source, in the transition space after the possible completion of the tcu. It is
located through the repetition of the preposition με (‘with’), and it involves the
replacement of one gender membership category for another.
Again the question to be raised here is ‘why that now?’ Why does the speaker
need to return to the gender membership category and replace it? Could this
be a category-formed error in which the speaker chooses the wrong category
and then substitutes it for the correct category (Jefferson 1996)? There are two
reasons why this should not be the case. First, the self-repair is delivered in an
increment after the pragmatic and grammatical completion of the tcu, which
constitutes a recognizable action, i.e. a question within the course of recip-
rocal sequences of assessment. Second, in the tcu, the speaker employs the
masculine gender for generic reference to third person. In the increment, the
speaker deviates from this pattern, limiting reference to female third persons
only. The self-repair in this conversational episode is linked with the interac-
tional relevance of female sex. As I mentioned above, the action accomplished
by Zoi’s turn is related to the category-bound activity of assessing physical sim-
ilarity between persons. This activity introduces the relevance of recipient’s
female sex in interaction, and shapes the choices made by the speaker at this
particular moment in interaction.
More specifically, the gender membership category ποιον, which is origi-
nally selected by the speaker, does not align with her orientation to recipient’s
female sex. The masculine gender marking the pronoun codifies referent’s male
sex and, thus, guides the interpretation of the referent as male. Νevertheless,
such an interpretation affects the possible understanding of the action that
Zoi’s turn was designed to accomplish. In her turn, Zoi associates the female
recipient with a male referent, implying that the female recipient could be
similar to a male person. Thus, she deviates from the general pattern followed
by participants in the conversational episode: assessing female co-participants
as similar to female third persons. For this reason, Zoi’s turn runs the risk of
being interpreted as irony, mockery or insult by Melita, who is the recipient.
Zoi tries to avoid this understanding through her self-repair. Overall, the self-
repair delivered by the speaker has a double aim: to incorporate the relevance
of female sex and ensure that the action accomplished by the turn will be
interpreted as such by the addressed recipient.
The self-repair examined in this extract can be interpreted by the analyst as

an indication of the role of the masculine gender in guiding the speaker to the
interpretation of the referent as male. The speaker draws on the stereotypical
assumption of man as the norm or human/universal, when she uses the mas-
culine gender for generic reference in the beginning of her turn. However, the
self-repair delivered in the increment shows that the masculine gender limits
speaker’s thinking, by guiding the interpretation of the referent as male. The
inference of human/universal that is produced via the use of the masculine
grammatical gender does not seem to include the category of female sex in
particular. This is why the speaker needs to return to the masculine gender and
replace it with the feminine.
The self-repair to be examined in Extract 22 occurs at line 18. In the lines pre-
ceding this extract, Vaso, Yana and Katerina refer to their childhood and they
discuss whether they used to go to kids’ camp. Yana and her sister, Katerina,
who used to go, assess their experience as a good one.
Extract 22 (I.14.A.19.1): Summer camping

1 (1.0)
2 Yana Δεν είσαι άτομο που μπορείς να μείνεις [για πολύ καιρό.]
You are not the type of person who can stay at a camp for a
long time.
3 Κaterina [Όχι δεν είμαι, ]
It’s not that I’m not,
4 είχα πάει:: ήμουν πολύ μικρή όταν (.) στα Ψακού[δια]=
I went [to a kids’ camp] I was very young when (.) in Psakoudia
5 Yana [ Η]=
6 Κaterina =[>πηγαίναμε παλιότερα με τους γονείς μας.<]
we went with our parents in the past.
7 Yana =[ Κατερίνα ήτανε τετάρτη, κι ε]γώ ήμουνα:
Katerina was in fourth grade, and I was
8 έ- (.) ε:: πε- έκτη δημοτι[κού. δυο χρόνια διαφορά.]
si- (.) eh:: in fi- sixth grade. [we had a] two years difference.
9 Κaterina [ Στα Ψακούδια, πη]γαίναμε
In Psakoudia, we went
10 κάθε χρόνο:, κι έμεναν οι γονείς μας σε άλλη σκηνή.
every year, and our parents stayed in another tent.
11 είχαν τότε >αυτή τη δυνατότητα ας πούμε.< [κι εμείς]
they were able to do that then so to speak. and we
12 Vaso [ Μ:. ]
Mm.
146 chapter 6
13 Κaterina μέναμε με όλα τα παιδιά:. αλλά ήταν κι εκεί οι γονείς μου.

stayed with all the kids. but our parents were also there.
14 Vaso Ναι ναι:.=
Yes yes.
15 Κaterina =Και μας είχαν στείλει μια χρονιά σε κάτι::
=And one year they had sent us to a
16 στις κατασκηνώσεις του Ντέμου. τώρα νομίζω λέγονται
to Demou’s camp. now I think it is called
17 αλλιώς. [στην^ Μποτίδαια.]
differently. in Potidea.
18 Yana → [ Ήμασταν πιο με]γάλοι [τότε. πιο με]γάλες.=
cop.1pl.past more old.nom.masc.pl then more old.
nom.fem.pl
We were older.masc then. older.fem.
19 Κaterina [Όχι Ποτίδαια.]
Not Potidea.
20 Κaterina =Στις Φώκιες.=
=In Fokies.
21 Vaso =Α:.
Uh:.
At line 2, Yana assesses Katerina’s competence in staying at a kids’ camp for a

long period as poor (Δεν είσαι άτομο που μπορείς να μείνεις [για πολύ καιρό.] [ðen
íse átomo pu borís na mínis ʝa polí ceró] ‘You are not the type of person who
can stay at a camp for a long time’). At lines 3–4, Katerina disagrees with Yana’s
assessment (Όχι δεν είμαι, [óçi ðen íme,] ‘It’s not that I’m not’) and initiates a
story telling (είχα πάει:: ήμουν πολύ μικρή [íxa pái:: ímun polí mikrí] ‘I went [to a
kids’ camp] I was very young’), treating Yana’s first pair part as story provoca-
tion (Lerner 1992). Stories are usually initiated by a story preface sequence in
which a story is proposed by a possible teller and is accepted or preempted
by another participant (Sacks 1974). After the preface, one participant delivers
the story in a telling sequence. One or more co-participants are the story recip-
ients. Yet, sometimes two or more participants may be able to jointly initiate
and deliver a story. In initiating an assisted story telling, participants demon-
strate their shared knowledge of the story events and establish themselves
as story consociates (Lerner 1992). As Lerner (1992, 250) shows, in assisted
story initiation “a co-participant can be selected to deliver the story through
a story prompt or story ‘provocation’ ”. Yana provokes Katerina into telling a
story about a kids’ camp by making a negative assessment about her com-
petence in staying at a kids’ camp for a long time. The negative assessment
invokes Katerina’s disagreement, which is followed by an account that takes

the form of a story telling. The assisted story telling establishes both Yana and
Katerina as story consociates sharing common knowledge about the story
events. Moreover, Katerina and Yana are introduced as protagonists in the story
via the first plural verbs πηγαίναμε ([piʝéname] ‘we went’), μέναμε ([méname]
‘we stayed’), and the pronouns εμείς ([emís] ‘we’), μας ([mas] ‘our’), μας (‘us’).7
The common knowledge that Yana and Katerina share is based on their
sister relation. The relevance of the ‘sister’ identity is introduced in this con-
versational episode via Yana’s assessment. More specifically, Yana’s assess-
ment is first positioned. According to Heritage and Raymond (2005, 16), first
positioned assessments “carry an implied claim that the speaker has primary
rights to evaluate the matter assessed” and can be used as resources for index-
ing speakers’ epistemic rights. Heritage and Raymond (2005) classify first posi-
tioned assessments in English interaction into unmarked, downgraded and
upgraded. Unmarked first assessments consist of simple declarative evalua-
tions that contain no items strengthening or weakening the evaluation and
“claim unmediated access to the assessable” (Heritage and Raymond 2005, 19).
Downgraded first assessments are “designed to exhibit downgraded epistemic
access to a state of affairs” via the use of evidentials or tag questions (Heritage
and Raymond 2005, 19). Finally, upgraded first assessments are designed to
exhibit upgraded epistemic access via the use of negative interrogatives.
Following Heritage and Raymond’s (2005) classification, Yana’s assessment
displays the features of an unmarked first positioned assessment, i.e. it is a
declarative utterance asserting an evaluation of Katerina (‘You are not the type
of person who can stay at a camp for a long time’). In doing so, Yana claims
unmediated access to the assessable, i.e. Katerina’s behavior. As Raymond and
Heritage (2006, 680) argue, “the management of rights to knowledge and, relat-
edly, rights to describe or evaluate states of affairs can be a resource for invok-
ing identity in interaction”. In this conversational episode, Yana’s unmarked
first positioned assessment makes direct access to Katerina’s conduct relevant
and evokes the relevance of Yana’s identity as Katerina’s sister. This specific
identity is consequential for the design of Yana’s turn at line 18.
Story consociates can participate in the course of story delivery by con-
tinuing the story or by repairing aspects of the story and its delivery, such as
trouble in the event sequencing of the story, in the delivery of the story, in
story elaboration, and in the facts of the story (Lerner 1992). Two story conso-
ciate initiated entries are found at lines 5, 7 and 8, and at line 18. Yana repairs
7 The same clitic form is employed for genitive and accusative case in first person plural per-
sonal pronoun.
148 chapter 6
trouble in story elaboration (lines 5, 7, 8) by giving more information about the

exact age that she and Katerina had at the time of the story events. Yana repairs
trouble in the facts of the story (line 18), that is, she corrects their age again.
Her turn consists of one tcu and one increment. The tcu is: [Ήμασταν πιο με]
γάλοι [τότε. ([ímastan pço meɣáli tóte.] ‘We were older.masc then.’). The incre-
ment is: πιο με]γάλες. ([pço meɣáles] ‘older.fem.’). The increment delivers the
self-repair and changes in some respect the turn to which ensuing talk will be
addressed (Schegloff n.d.). In the tcu, Yana refers to herself and her sister, by
employing the first person plural verb ήμασταν, and describes the referents via
the masculine adjective μεγάλοι (‘old’). In the transition space after the possi-
ble completion of the tcu, Yana does a self-repair. She replaces the masculine
adjective with the feminine one μεγάλες, and pre-frames the repairable item by
repeating the adverb πιο (‘more’).
Why does the non-match between grammatical gender and referents’ sex
that usually remains unremarkable in interaction (cf. section 6.2.5) create
trouble in this conversational episode? Similar to Extracts 20 and 21, in this
extract, it is unlikely that the speaker deals with a ‘category-formed’ error
(Jefferson 1996) for two reasons. First, the self-repair comes after the pragmatic
and grammatical completion of the tcu, which does a recognizable action, i.e.
the speaker repairs trouble in the facts of the story. This means that the mas-
culine adjective was selected by the speaker together with the other elements
of the tcu to accomplish this specific action. Second, in the tcu, the speaker
uses the grammatically masculine adjective to refer to more than one female
referent, by employing the inference of human/universal associated with the
masculine gender. Yet, in the increment, she chooses to categorize referents as
exclusively female via the feminine adjective.
In repairing grammatical gender, Yana aims at incorporating the relevance
of the sister-identity that was introduced in interaction via her assessment at
line 2 and remained throughout the sequence because of the assisted story-
telling, which established Yana and Katerina as story consociates who share
common knowledge. The action that Yana’s turn is designed to accomplish
is related to her orientation to her identity as Katerina’s sister. Namely, Yana
is entitled to repair trouble in story elaboration, because she shares common
knowledge with Katerina. However, trouble is created when it comes to the
gender membership category that the speaker selects—together with other
elements in the composition of the turn—to do this sort of interactional work.
The deployment of the masculine adjective does not align with speaker’s ori-
entation to her sister-identity, which is conceptually marked for female sex.
Therefore, the speaker selects the feminine adjective in order to incorporate
what is relevant in interaction at this particular moment.
Similar to the previous extracts, the self-repair examined in this extract

uncovers the cognitive role of grammatical gender. In addition, it shows that
the inference of human/universal produced by the use of the masculine gram-
matical gender in reference to more than one female referent presupposes the
male categorization of referents and does not include the category of female
sex in particular.
The generic use of the masculine gender can also be an item of exposed
and embedded correction, as is shown in the next two extracts. In Extract 23,
Yanis assesses negatively Dora, a female player in the Big Brother tv show, and
Marina disagrees with him. The correction to be examined is found at line 25.
Extract 23 [Ι.14.Α.17.1]
1 Αreti [Ε ναι ρε [συ.] τις έχουν ξεπα↑στρέψει τις <γυναί]κες.>
Uh yes man. women have been kicked off.
2 Yanis [°(. . . . . . . . . . .)]
3 Μarina [Για↑τί μιλάς έτσι για μια <κοπέλα.>]
Why do you speak about a girl like that.
4 Yanis ↑Α[ΦΟΥ Η]ΤΑΝΕ χαζοχαρούμενη η ά[λλη ρε.]
Because she was stupid/lighthearted.
5 Αreti [°(είναι-)]
([She] is-)
6 Μarina [↑ΕΝΤΑ]:ΞΕΙ
Okay.
7 ΟΜΩΣ- h (.) ε?=
but- h (.) eh?
8 Yanis =[(Δηλαδή αυτή δεν] είναι?)=
That is to say that she is not?
9 Αreti =[ Η Γωγώ::? ]
[You mean] Gogo?
10 Μarina = >Όλες οι άλλες όμως< ήτανε: μπουρούχες να πούμε. ήταν
All.fem the others.fem were ugly and stupid so to speak.
11 ήταν η μόνη που: τέλος πάντων έδινε κάποια ζωντάνια. κάτι.=
anyway she was the only one who livened things up. [who was
doing] something.
12 Yanis =°Ναι.
Yes.
13 (.)
14 Yanis Όπως λένε κι οι φα[ντάροι, η Δώρα ήταν η κλαψομούνα.]
As soldiers [in the military service] say, Dora was the whining-
cunt.fem.
150 chapter 6
15 Μarina [ Έδειχνε ότι υπήρχε εκεί ] μέσα.

She showed that she existed in there.
16 (.)
17 Μarina Ε:?
Eh:?
18 Spyros Κλαψομούνα?
Whining-cunt.fem?
19 Yanis °(Μ.)
(Mm.)
20 Spyros ((muffled giggle))
21 (0.7)
22 Yanis °(Γιατί τη βάλανε.)=
(Why did they put her in the game.)
23 Νiki → => Ποιος είν’ ο κλαψομούνης?
((in a laughing tone . . . . . . . .))
who.nom.masc cop.3sg.pres the.nom.masc whining-cunt.
nom.masc
Who.masc is the whining-cunt.masc?
24 (0.7)
25 Yanis → [Μούνα.]
cunt.nom.fem
Cunt.fem.
26 Spyros [ Η πιο ] χάλια παιδιά ήταν η πρώτη πρώτη που έφυγε.
((in a laughing tone . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . .))
Hey guys the.fem most horrible was the very first.fem one to
leave.
At line 14, Yanis describes Dora via the feminine pejorative term κλαψομούνα
([klapsomúna] ‘whining-cunt’), that is, as a person who always complains and
does not get pleasure from anything. At line 23, Niki initiates a repair sequence
to deal with her trouble in identifying the person who is described as a whin-
ing-cunt (=> Ποιος είν’ ο κλαψομούνης? [pços in o klapsomúnis?] ‘Who.masc is
the whining-cunt.masc?’). She initiates repair via the masculine interrogative
pronoun ποιος (‘who’) and describes the referent via the masculine adjective
κλαψομούνης (‘whining-cunt’). Niki’s first pair part makes relevant a second pair
part that will deliver the repair. Instead of delivering a second pair part, the
next speaker produces a turn that is designed to correct previous speaker’s
turn (the gap after the first pair part indicates that there might be a prob-
lem with the production of the second pair part). The item of this exposed
correction is the gender membership category κλαψομούνης. Yanis corrects the
masculine adjective that Niki used in her previous turn by offering the equiva-
lent feminine adjective as an alternative. More specifically, he uses the second
root of the compound word, the feminine item μούνα [múna] that is marked
as female both grammatically and semantically. The semantic marking of sex
is due to the fact that the word μουνί ([muní] ‘cunt.NEUTER’) denotes woman
metonymically. As I mentioned in section 6.3.1, exposed correction consists
of a speaker producing an object X, a subsequent speaker producing an alter-
native object Y, and a prior speaker producing an alternative object Y or the
same object X. The first two parts of this series are found in this conversa-
tional episode. Namely, Niki employs the masculine grammatical gender and
Yanis offers the feminine grammatical gender as an alternative. The correction
sequence is not continued, because Yanis’s correction overlaps with Spyros’s
first pair part, which continues the sequence that was suspended by the
exposed correction.
The exposed correction done by the speaker shows that the speaker inter-
preted the sex of the referent introduced by the previous speaker via the gram-
matically masculine pronoun as male. For this reason, this correction can be
interpreted as an indication of the correlation between grammatical gender
and cognition. Contrary to what we have seen so far, the exposed correction in
this conversational episode does not seem to be contextually motivated. Or, if
context plays some role, this is not shown in the interaction data.
An embedded correction of grammatical gender is examined in the next
extract. The conversational topic here is the curriculum and other issues
related to participants’ undergraduate studies at the School of Philology.
Extract 24 [Ι.20.Α.28.1]
1 Μelita [Καταρχήν] εγώ <πιστεύω,> ότι ε στο πανε- το πεζ- (0.5) ε
First of all I believe, that uh in the uni- the pez- (0.5) uh
2 ↑πτου:. ((γελάκι))=
damn. ((giggle))
3 Τania =((she laughs[. . . . . . . . . . .))]
((5 lines omitted))
9 Μelita [Το <πανεπιστήμιο> δε σου μα]θαίνει παιδιά κάτι
Hey guys the university does not teach you.sg anything
10 για να διδά:ξει::ς. ↑σας έχει μάθει εσάς τ’ αρχαία ξέρω γω: .h=
with respect to teaching. did it teach you.PL [anything] about
ancient Greek for example .h
11 Εvagelia =Ου:. πάρα πολύ.=
interj [irony] a lot.
152 chapter 6
12 Dimitris =↑Το πιστεύεις ότι κι εγώ αυτό το ι- (.) ίδιο παρατηρώ >ας
Do you believe that I also observe this the i- (.) the same
[thing] so
13 πούμε ότι< δεν έχει[: ένα μά]θημα παιδα- >να μου πεις<
to speak that there is no peda- course anyway
((lighter sound))
14 Melita [( . . . . .)]
15 Dimitris έχει ένα παιδαγωγικό το οποίο είναι εισαγωγή. καπ-
there is a pedagogical [course] that is an introduction. som-
16 ο τρόπος >δηλαδή να διδάξεις το μάθημα.<=
namely the way to teach a course.
17 Melita =Εκτός του- [του τρό:που:. και το αντικείμενο ] ρε παιδί μου.
Besides the- the way. [it is about] the topic as well.
18 Dimitris [Όταν μπεις σε τάξη. έστω μια °πρακτ-]
When you.SG enter a class. some [sort of ] practi- so
be it
19 Melita στ’ αρχαία πιχί κάναμε στο σχολείο, το και το,
for example at school in the course on ancient Greek we did,
this and that,
20 να ξες να το εξηγήσεις, .h ↑τώ:ρα, είναι μόνος σου, ά:μα
((noise. . . . .))
subj know.2sg subj it.acc explain.2sg.pfv now cop.3sg.
pres on.your.own.nom.masc if
you should know how to explain it, .h now, you are on your
own.masc, if
21 [θες >να] κάτσεις να< τα διαβάσει:ς, [.h και να] είσαι και
want.2sg.pres subj sit.2sg.pfv subj them read.2sg.pfv
and subj cop.2sg.pres and
you want to sit down and read them, .h and you will be also
22 Dimitris [ Έτσι. ]
Precisely.
23 Tania [Ναι °ναι.]
Yes yes.
24 Melita τυχερός άμα το περάσεις με ένα πέ:ντε¿ [μια χα]ρά, .h
lucky.nom.masc if it.acc pass.2sg.pfv with a.neuter five
a.fem joy.fem
lucky.masc if you succeed in passing the exams with a five.
that’s fine, .h
25 Tania [°Ναι:.]
Yes.
26 Melita δε σου μαθαίνει °κά:τι.=

It doesn’t teach you.2sg anything.
27 Tania → =Και γενικά στο πανεπιστήμιο πρέπει
and in.general in.the.acc.neuter university.acc.neuter
must
And in general at the university you must
28 μόνη σου °να [ διαβάσεις.° ]
on.your.own.nom.fem subj read.2sg.pfv
read on your own.fem.
29 Melita → [°Μόνη σου. α]κριβώ(ς).°
on.your.own.nom.fem exactly
On your own.fem. exactly.
30 απλά μπορεί να σου δώσει το πανεπιστήμιο .h τη:ν ε: δυνατότητα
the university may just give you.SG .h the u:h chance
31 να σου πει διάβασε αυτό κι αυτό °άμα [ θε:ς, ]
[it may] instruct you.SG to read this and that if you.SG want,
32 Dimitris [>Ε ναι. δίνω] τώρα
Uh yes. now I give
Melita claims (lines 9–10) that students do not receive any teaching training
at the university ([Το <πανεπιστήμιο> δε σου μα]θαίνει παιδιά κάτι για να διδά:ξει::ς.
[to <panepistímio> ðe su maθéni peðʝá káti ʝa na ðiðá:ksi::s] ‘Hey guys the uni-
versity does not teach you.SG anything with respect to teaching.’). The speaker
employs the second person singular pronoun σου (‘you’) to refer to an abstract
group of referents, all students of the school of philology, including speaker
and the other co-participants. In the next tcu, she addresses a question to
her co-participants, soliciting confirmation of her claim (↑σας έχει μάθει εσάς
τ’ αρχαία ξέρω γω: .h= [↑sas éçi máθi esás tarçéa kséro ɣo: .h] ‘did it teach you.
PL [anything] about ancient Greek for example .h’). She selects them as pos-
sible next speakers via the second person plural pronouns εσάς, σας (‘you’). In
his reply at lines 12, 13, 15 and 16, Dimitris keeps reference to the same abstract
group of referents via the second person singular verb να διδάξεις ([na ðiðáksis]
‘you teach’). At lines 17, 19, 20, 21, 24 and 26, Melita refers to the same referents
via the first person plural verb κάναμε ([káname] ‘we did’, line 19). Then, she
shifts to second person singular να ξες να το εξηγήσεις ([na kses na to eksiʝísis]
‘you should know how to explain’, line 20), θες >να κάτσεις να< τα διαβάσει:ς
([θes >na kát͡sis na< ta ðʝavási:s] ‘you want to sit down and read them’, line
21), να είσαι ([na íse] ‘you are’, line 21), περάσεις ([perásis] ‘you pass’, line 24),
and categorizes the referent via the masculine second person singular pro-
noun μόνος σου ([mónos su] ‘on your own’, line 20) and the masculine adjective
154 chapter 6
τυχερός ([tiçerós] ‘lucky’, line 24). In their turns, speakers employ the second
person singular to refer to self as part of an abstract group of referents and
make assertions about the collective experience that they share with their co-
participants. In this way, they introduce the relevance of their own and their
co-participants’ experience as students at the school of philology (cf. Schegloff
1996b). Tania keeps generic second person self-reference at lines 27–28 (να
διαβάσεις [na ðʝavásis] ‘you read’). However, she categorizes referents as female
via the feminine pronoun μόνη σου ([móni su] ‘on your own’). In her next turn,
Melita agrees with Tania. She keeps generic second person self-reference and
repeats the gender membership category μόνη σου.
The change of grammatical gender in this conversational episode operates
as embedded correction, in Jefferson’s (1987) terms. The speaker employs a
linguistic item marked by the masculine grammatical gender, that is, she pro-
duces an object X. The subsequent speaker employs a linguistic item marked
by the feminine grammatical gender, that is, she produces an alternative
object Y. Finally, the prior speaker employs the linguistic item marked by the
feminine grammatical gender, that is, she produces the alternative object Y.
The XYY is embedded into the ongoing talk with no attendant activities. The
categorization of referents as exclusively female becomes important for what
participants do in interaction, because of the relevance of speaker’s personal
experience that has been introduced in interaction via the generic second per-
son singular. This experience seems to be hidden by the abstraction and indefi-
niteness that the generic use of the masculine carries. The deployment of the
feminine pronoun limits reference to female persons only and, thus, aims at
foregrounding speaker’s personal experience.
In the extracts examined in this section, the codification of referent’s sex
via grammatical gender becomes important for what participants do in inter-
action. In these cases, the cognitive role of grammatical gender is uncovered
in interaction. In the next section, I show that indications of the cognitive
role of grammatical gender can also be found in cases in which participants
do not ‘notice’ the routine meanings associated with the use of grammatical
gender.
6.3.3 Other Indications

Besides the overt or direct indications of the cognitive role of grammatical
gender found in repairs, interaction also provides covert or less direct indica-
tions. These are divided into four kinds: i) speakers’ next turns, which show
their understanding of prior turn, ii) turns’ recipient-design, iii) the mismatch
between grammatical gender and referents’ sex, and iv) the membership cat-
egorization device.
Next turns are understood by co-participants as displaying their speaker’s

understanding of the just-prior turn (cf. section 5.3). Sometimes next turns
display how recipients interpret the sex of the referent that was introduced or
described by items marked by grammatical gender in the speaker’s previous
turn. For example, this is shown in Extract 13. This extract was discussed in sec-
tion 6.2.4. Here I cite it again up to line 14. This time the analysis focuses on the
relation between the action accomplished by Evagelia’s turn at line 7, and the
action accomplished by Zoi’s turn at line 11.
Extract 13* (Ι.20.A.28.1): Μilitary service

1 Melita =Σε περίπτωση που τελειώσετε τη σχολή τι σκέφτεστε
If you finish school what are you thinking
2 °να [ κάνετε]:?=
of doing?
3 Dimitris [((giggle?))]
4 Evagelia =((giggle))
5 Melita =Ι[διωτ- ]
Priv-
6 Dimitris [(Το μα][λάκα.)]=
(Playing dumb.)
7 Evagelia → [Να γί]νουμε [<πωλή:τριες, σερβιτό:ρε::ς,>]
subj become.1pl saleswomen.nom.fem wait-
resses.nom.fem
Becoming saleswomen, waitresses,
8 Tania [°(Το μαλάκα λέει.) ((γελά . . . . .]. . .))
(Ηe says that he will play dumb.) ((she laughs))
9 (0.7)
((sound during the silence))
10 Tania [ Ντελίβερι:, ]
Delivery.
11 Ζoi [>Εσύ θα πας στο στρατό<]°φαντάζομαι.
I guess that you will go into the army.
12 Dimitris Βασικά. >πρέπει οπωσδήποτε.<
Basically. definitely I have to.
13 (1.6)
In her reply to Melita’s question, Evagelia refers to self and other female co-par-
ticipants via the first person plural. By employing the grammatically feminine
156 chapter 6
nouns πωλή:τριες (‘saleswomen’) and σερβιτό:ρε::ς (‘waitresses’), she categorizes

the collectivity as female. Zoi self-selects as next speaker at line 11 and intro-
duces military service as what Dimitris will do after graduation. By referring
to the military service, that is, an exclusively masculine activity, she limits the
range of next speakers who can produce a relevant second pair part to Dimitris
only. The use of the feminine grammatical gender in Evagelia’s turn is related
to the fact that Zoi treats Dimitris as an individual different from the collectiv-
ity introduced by Evagelia in her previous turn.
After Melita’s question, which is addressed to all co-participants, Evage-
lia refers to self collectively. This collectivity is expected to include all co-
participants who have been selected as possible next speakers by Melita.
However, the speaker categorizes the collectivity as exclusively female via the
use of gender membership categories. This gendered categorization has con-
sequences in terms of how the conversational episode proceeds. By referring
to an exclusively masculine activity, Zoi differentiates Dimitris from the collec-
tivity that Evagelia referred to in her previous turn. Thus, in her next turn, Zoi
displays her understanding of Evagelia’s turn as a second pair part, which refers
to speaker and female co-participants only. In this way, Zoi’s next turn provides
an indirect indication of the cognitive role of the feminine grammatical gender.
In Extract 25, Anna, Elsa, Marina and Natasa make plans for going to the free
pass movies at a film festival.
Extract 25 [I.20.A.29.1]: Movies for free

1 Elsa → [Ωραία. >πότε] θέλετε να πάμε
well when want.2pl.pres
subj go.1pl
Good. when do you want to
go
2 → στα τζάμπα όλοι μαζί?<
in.the.acc.neuter free all.nom.masc together
[to the movies] all.masc together for free?
3 Marina [Εγώ παιδιά- Δευτέρα Τρίτη] Τετάρτη έχουμε μάθημα
I hey guys- on Monday Tuesday Wednesday we have a class
4 Natasa [Να κανονίσουμε μια μέρα.]
We should arrange this one day.
5 Marina έντεκα δύο. πώς θα-
from eleven to two. how will-
6 Anna °Ναι.=
Yes.=
Elsa asks her co-participants when they would like to go to the free pass movies
(lines 1–2). She selects them as next speakers via the second person plural verb
θέλετε ([θélete] ‘you want’), and describes them and herself via the masculine
adjective όλοι ([óli] ‘all’). Thus, she employs the masculine gender to refer to
more than one female referent. The next turns-answers produced by Marina
and Natasa show that female participants interpreted successfully the social
action that Elsa’s turn was designed to accomplish and that they were able to
‘see’ themselves through the inference of human/universal produced by the
masculine grammatical gender. In this way, next turns in this conversational
episode indicate implicitly the correlation betweeen the masculine gram-
matical gender and the interpretation of referents as male-human-gender
indefinite.
The same applies to Extract 10, which was analyzed in section 6.2.2. Melita
addresses a question to her female and male co-participants (lines 5–6),
employing the masculine grammatical gender to refer to them, and Zoi deliv-
ers an answer (lines 7–8). Her next turn shows that she understood herself as
one of the recipients to be addressed and selected as possible next speaker. In
other words, she understood the referent described by the masculine adjective
as not exclusively male.
Indications of the correlation between grammatical gender and the
intepretation of referent’s sex can also be found in the recipient-design fea-
ture of turns. As I mentioned in the beginning of this chapter, turns are always
recipient-designed, that is, speakers design their turns in such a way as to make
them appropriate and relevant for recipients. If we go back to Extracts 7, 8 and
9, we can make the following remark. In these extracts, speakers assess recipi-
ents and use gender membership categories, which match recipients’ sex. In
this way, speakers make their turns appropriate for the specific recipients they
address and show indirectly their interpretation of recipients as women or
men. Extracts 10 and 16 are also interesting from this perspective, because they
implicitly uncover the conceptual processes associated with the generic use of
the masculine. In these extracts, speakers employ linguistic items marked by
the masculine grammatical gender to achieve reference to females plus males,
and, thus, show their understanding of the masculine gender as a tool for acti-
vating the inference of human/universal. The same understanding is uncov-
ered by the use of the masculine gender for reference to female referents only
that was examined in Extracts 16, 17, 18 and 25.
When the generic use of the masculine is found in collective and generic
or indefinite self-reference and reference to recipients, it is easy for analysts
to claim that the inference of human/universal is activated, given that refer-
ents are present at context. Yet, this is not the case in non-recognitional third
158 chapter 6
person reference, as it is not always clear whether referents are exclusively male
or not (cf. section 4.2.1). For instance, if interlocutors refer to a stereotypically
masculine activity, it is very likely that the referents denoted by the masculine
reference form will be interpreted as exclusively male, rather than as male plus
female. In the following two extracts, I show that the membership categoriza-
tion device can provide a partial solution to the ‘ambiguity’ that accompanies
the generic use of the masculine in non-recognitional third person reference.
In Extract 26, Yanis gives advice to his friend Spyros, who is about to join
the army.
Extract 26 [Ι.14.Α.17.1]: Preparing for the army

1 Yanis Μπαταρίες, (1.1) τσιγάρα πάρε <από έξω,> (0.6)
You should buy batteries, (1.1) cigarettes outside [the military
camp], (0.6)
2 Spyros Ναι:? γιατί?
Yes? why?
3 (1.0)
4 Yanis Ε γιατί περιμένει πολύ ουρά ρε παιδί μου.
Uh because there is a long waiting queue.
5 (.)
6 Spyros Καμιά κούτα ξερω γω?
Let’s say a box [of cigarettes]?
7 Yanis Ναι.
Yes.
8 (0.6)
9 Spyros >Κούτα είχες πάρει εσύ?<
Did you take a box?
10 Yanis Εγώ δεν είχα πάρει. >εγώ έκανα τράκα.< ((ρουφά τη μύτη του))
I didn’t. I was bumming cigarettes off others. ((he sniffs))
11 (.)
12 Niki ((he laughs))
((Yanis sniffs))
13 Niki → Από ποιον έκανες τράκα? εσύ ήξερες κανένα όταν^ μπήγες?
prep who.acc.masc do.2sg.past cop.acc.fem you
know.2sg.past anyone.acc.masc when go.2sg.past
Whom.masc did you bum a cigarette off? did you know any-
one.masc when you joined the army?
14 (.)
15 Yanis Όχι.
No.
16 (.)
17 Niki → Δεν έφευγες με γκανέναν?
neg leave.2sg.imperf with anyone.acc.masc
Didn’t you leave with anyone.masc?
18 Yanis Τς. °μόνος.
Ts [negation]. [I left] alone.masc.
19 (1.3)
20 Yanis → Αλλά γνώρισα πολλούς.
but meet.1sg.past many.acc.masc
But I met many.masc.
Niki shifts topic at line 13, by asking Yanis whether he knew anyone when he
joined the army, and what he was doing with regard to smoking. She addresses
him via the second person singular verbs έκανες ([ékanes] ‘you did’), ήξερες
([íkseres] ‘you knew’) and the pronoun εσύ ([esí] ‘you’). She employs the mas-
culine interrogative pronoun ποιον ([pçon] ‘who’) and the masculine indefi-
nite pronoun κανένα ([kanéna] ‘anyone’) for generic reference. At line 17, she
addresses another question to Yanis. She employs the second person singular
verb έφευγες ([éfevʝes] ‘were leaving’) to refer to Yanis and the masculine indefi-
nite pronoun γκανέναν ([ganénan] ‘anyone’) for generic reference. In his reply
at line 20, Yanis employs the masculine adjective πολλούς ([polús] ‘many’) as a
non-recognitional reference form. In this conversational episode, participants
refer to a category-bound activity, that is, military service, which is stereotypi-
cally associated with men in the Greek context. Referring to the army can intro-
duce the relevance of the category ‘man’ in interaction and, thus, favors the
interpretation of referents as exclusively male. Note that this interpretation is
not shown in the data of interaction. It is an assumption that we analysts can
make, taking membership categorization device into account.
A similar example is found in the following extract. However, in this conver-
sational episode, interaction data provide an indication of the interpretation
of referent’s sex in speaker’s next turn. Melita describes her life in Athens in
the first year of her undergraduate studies and how she was sexually harassed
on the street.
Extract 27 [Ι.20.Α.28.1]: Athens

1 Μelita Και: ήμουνα ρε παιδί μου ξέρεις τώρα <πρώτο έτος.> .hh
And you know now I was in first grade. .hh
2 δεν είχα συγγενείς τίποτα Αθήνα. είχε περάσει μόνο
((noise. . . . .))
I didn’t have any relatives in Athens nothing. there was only
160 chapter 6
3 μία συμμαθήτριά μου Αθήνα:, η οποία έμενε Αιγάλεω.

one.fem classmate of mine studying in Athens, who was stay-
ing in Egaleo.
4 καμία σχέση. δύο ώρες να πας °δηλαδή.°=
very far away. that is, you needed two hours to reach her.
5 Ζoi =Μ[::. ]
Mm.
6 Μelita [Και ήμουν τό]σο παιδιά μό:νη °μου° και τόσο στο
And hey guys I was so lonely and [I was] in the
7 κέντρο των εξελίξεων.=
((laughing . . . . . . . .))
thick of things.
8 Ζoi/Tania =((they laugh))
9 Μelita Φοβόμουνα να πάω σούπερ μάρκετ
I was afraid to go to the super market
10 βρά:[δυ. ξέρω γω από]γευμα. [°σε τέτοια] φάση.° καλά.
at night. in the evening so to speak. the situation was like that.
okay.
11 Ζoi [ Α::, ] [ °Ωραία. ]
A::h, Fine.
12 Μelita δεν υπήρχε περίπτωση να βγω: και να μην έρθει κάποιο
neg exist.3sg.imperf occasion.nom.fem subj go.out.1sg.
pfv and subj neg come.3sg.pfv someone.nom.neuter
there was no chance that I would go out without meeting a
13 → πρεζόνι, κάποιο:: α:νώμαλος. ήτανε [το στάνταρ. ]=
junkie.nom.neuter someone.nom predator.nom.masc
cop.3sg.past the.nom.neuter standard
junkie, a.masc sexual predator.masc. that was for sure.
14 Τania [((she laughs))]=
15 Εvagelia =Είχες [και [ τα τυ]χερά] °σου.=
((in a laughing tone . . . . . . . . . . . . .))
You also had your chance.
16 Τania [((she laughs))]
17 Μelita [ Ειλικρινά. ]
Honestly.
At lines 9–10, Melita describes the fear that she felt when she was in public
spaces at night and, thus, she refers to a stereotypically feminine behavior. At
lines 12–13, she gives an account for her fear: she used to run into people who
threatened her safety. She refers to these people via the non-recognitional
reference forms κάποιο πρεζόνι ([kápço prezóni] ‘a junkie’) and κάποιο:: α:νώμαλος
([kápço:: anómalos] ‘a sexual predator’). The grammatical marking of male
sex in the second form, plus its lexical meaning, activate a cultural stereotype
according to which men—who have some sort of disorder—usually harasss
women. This stereotype, in combination with the stereotype of the insecure
and vulnerable woman that has been previously introduced in interaction,
guides recipients to interpret referent as exclusively male. This interpretation
is displayed in Evagelia’s next turn at line 15. Evagelia responds with an ironic
comment, characterizing the incident that Melita describes as luck. Thus, she
interprets the behavior of the third person as sexual desire for Melita. Based on
the norm of compulsory heterosexuality, men are the ones expected to have
sexual desire for women.
However, it is not always easy to resolve the ambiguity associated with the
generic use of the masculine in third person reference in interaction. In the
next extract, Dimitris, Melita and Tania describe the problems of the public
sector in Greece. They use masculine non-recognitional third person reference
forms at lines 5, 8 and 9.
Extract 28 [Ι.20.Α.28.1]: Public sector

1 Μelita =Να σου πω κάτι? γιατί να μη κάνουμε κάτι ούτως ώστε
Can I tell you something? why don’t we do something that
2 να γίνει και το <δημόσιο καλύτερο?>
will improve the public sector?
3 [(να μου) πεις,]
[anyway,]
4 Tania [Ναι. αυτό θα] σου ’λεγα κι εγώ.=
((lighter sound . . .))
Yes. I would also say the same thing.
5 Melita → =ο καθένας μόνος του >δεν μπορεί< να το πετύ[χει αλλά ίσως]
the.nom.masc each.one.nom.masc on.his.own.nom neg
can.3sg.pres subj it.acc achieve.3sg.pfv but maybe each
one.masc on his own cannot achieve it but maybe
6 Dimitris [Άντε. θα κάνει]ς
Come on. you will do
κάτι εσύ. θα ’χει αντί[κτυπο αυτό που θα κάνεις?]
something yourself. what you will do will it have an impact?
7 Melita [ °Αυτό λέω. ναι. ]
That’s what I am saying. yes.
162 chapter 6
8 Dimitris → γιατί αυτός που θα ’ναι εκεί πάνω, που θα περιμένεις >εσύ<
because he.nom pron.rel will cop.3sg there up conj fut
wait.2sg you
because the one.masc who will be up there [in a position of
power], while you are waiting
9 → να κάνεις, >αυτός θα τα παίρνει.<
subj do.2sg he.nom fut them get.3sg.pres
to do something, he will be profiting [from the public sector]
Μelita asserts that the public sector cannot improve if people do not under-
take a collective form of action. At line 1, she refers to self collectively via the
first person plural κάνουμε ([kánume] ‘we do’). This collectivity includes her-
self, co-participants and all members of the community in general. At line 5,
she employs the grammatically masculine referring expression ο καθένας μόνος
του ([o kaθénas mónos tu] ‘each one on his own’) to maintain reference to
the same collectivity, that is, a gender indefinite collectivity. However, things
become fuzzier when it comes to the deployment of the masculine items at
lines 8–9. Dimitris employs the masculine pronoun αυτός ([aftós] ‘he’) twice
to refer to any individual who holds a position of power in the public sector.
Given that superior social position is usually associated with men, referents
may be interpreted as exclusively male, rather than male plus female.
Finally, an indirect indication of the correlation between grammatical gen-
der and the interpretation of referent’s sex is provided by the speaker’s switch
from feminine to masculine gender in Extract 29. This switch looks similar to
self-repair, because it involves the replacement of a feminine item with the
equivalent masculine. However, the mechanism of repair is not employed.
Melita, Evagelia, Tania and Zoi describe the difficulties of living together with
another person. Melita finds it difficult to share space with someone else after
having lived on her own for so long.
Extract 29 [Ι.20.Α.28.5]: Home alone

1 Μelita [Ε να] θες και συ παράδειγμα να κυκλοφορείς
Uh for example you might also want to
2 °ελεύθερη στο [σπίτι σου. δεν μπο]ρείς.=
((noise. . .))
move freely inside your house. you cannot [do it].
3 Evagelia [ <Εννοείται.> ]
That goes without saying.
4 Tania =Ναι. και μόνο γι’ αυτό το λόγο:,
Yes. and for that reason only,
5 Melita Μμ.
Mm.
6 (1.5)
7 Tania Όχι. εμείς τέτοια προβλήματα δεν είχαμε ποτέ.
((noise . . .
No. we never had such problems.
8 (.)
. . .
9 Melita Και τώρα που μένω: μό:νη μου:, στην εστία:, (1.2)
. . . . . . . . . . . .))
And now that I live on my own.fem, in the dorm, (1.2)
10 → >τέλος πάντων. μόνη μου. πώς μπορείς να το θεωρείς μόνη σου
anyhow on.my.own.nom.fem how can.2sg.pres subj it.acc
think.2sg.ipfv on.your.own.fem
anyhow. on my own.fem. how can you think of yourself as
being on your own.fem
11 στην εστία.< έχεις τόσα άτομα. ↑αλλά στο χώρο σου
in the dorm. you have so many people [around]. but in your
space
12 → είσαι μόνος σου εν πάση περιπτώσει. (.) φαντάζομαι ότι πολύ
cop.2sg.pres on.your.own.masc anyway reckon.1sg.pres
conj very
you are on your own.masc anyway. (.) I reckon that [it would
be] very
13 δύσκολα θα συγκατοικούσα με κάποιον. γιατί μαθαίνεις
hardly fut live.together.1SG.IMPERF with someone.acc.
masc because learn.2sg.pres
difficult [ for me] to share the house with someone else.
because you learn to
14 να είσαι μό:νος σου. [κάνει]ς ό,τι θες.
((noise. . . . . . . .))
subj cop.2sg on.your.own.masc do.2sg.pres pron.rel
want.2sg.pres
be on your own.masc. you do whatever you want.
15 Zoi [ Μ:. ]
Mm:.
16 Zoi °Παιδιά είναι πολύ καλό αυτό.
Hey guys that’s very good.
164 chapter 6
At line 9, Melita refers to self via the first person singular verb μένω: ([méno] ‘I
stay’), and describes herself via the feminine pronoun μό:νη μου ([móni mu] ‘on
my own’). At line 10, she maintains individual self-reference (μόνη μου ‘on my
own’) but in the following tcus she shifts to generic second person singular
(μπορείς να το θεωρείς μόνη σου [borís na to θeorís móni su] ‘you can think of your-
self as being on your own’) and she refers to herself plus all other students liv-
ing in the dorm. Melita employs the same feminine pronoun and then replaces
it with the masculine pronoun (μόνος σου ‘on your own’). She keeps using the
masculine gender till the end of her turn (γιατί μαθαίνεις να είσαι μό:νος σου [ʝatí
maθénis na íse mó:nos su] ‘because you learn to be on your own’). What is the
problem here? The speaker seems to be restricted by the use of the feminine
grammatical gender at this particular moment in interaction that categorizes
the referent as exclusively female. She uses the masculine gender in order to
activate the inference of human/universal and ensure that generalization or
indefiniteness is expressed. The switch of grammatical gender found in this
conversational episode uncovers indirectly the conceptual processes linked
with the use of the masculine grammatical gender.
6.3.4 Conclusions
The analysis presented in sections 6.3.2 and 6.3.3 shows that grammatical gen-
der guides speakers to interpret referent(s) as female or male. Overall, indi-
cations of the cognitive role of grammatical gender in person reference in
interaction are found when participants recognize or notice the codification
of referent’s sex via grammatical gender, as well as when the codification of sex
via grammatical gender passes unnoticed. Indications of the first kind can be
described as overt and indications of the second kind as covert.
Overt indications are found in self-repairs and exposed and embedded cor-
rections of grammatical gender. Repairs involve switching from one grammati-
cal gender to another and address problems in the interpretation of referent’s
sex encountered by speaker or recipient. The source of trouble is related to
aspects of context, which make the codification of sex via grammatical gen-
der important for what participants do in interaction. Speakers suspend the
progressivity of the turn (in self-repairs) or the sequence (in exposed correc-
tion) in order to modify the interpretation of referent’s sex that is denoted or
described by linguistic items marked by grammatical gender. In addition, self-
repairs of grammatical gender ensure that the action that the speaker’s turn is
designed to accomplish will be interpreted as such by recipients. Self-repairs
occur in an increment after the pragmatic and grammatical completion of a
tcu or within the same tcu, and exposed and embedded corrections occur in
next turn. In other words, repairs of grammatical gender occur after the item
marked by grammatical gender was produced and the sex of referent was inter-
preted by interlocutors. In this way, repairs can be interpreted by analysts as
indications of the role of grammatical gender in guiding speakers’ thinking of
referent(s) as male or female, and for the conceptual processes that take place
during the production and understanding of turns in interaction. The majority
of the repairs examined in section 6.3.2 involve the generic use of the mascu-
line (see Extracts 20, 21, 22, 23, 24). More specifically, the self-repair examined
in Extract 20 shows that the masculine gender can produce the inference of
human/universal. This inference is based on the male categorization of refer-
ents, as self-repairs and corrections in Extracts 21, 22, 23 and 24 show.
Covert indications of the cognitive role of grammatical gender are found
in i) next turns, ii) turn’s recipient-design, iii) the membership categoriza-
tion device, iv) the mismatch between grammatical gender and referents’ sex,
v) and the switch of grammatical gender. These indications point to the cogni-
tive role of grammatical gender only indirectly. They can be inferred by us ana-
lysts, based on the knowledge that we have about the way in which interaction
is organized and turns are designed.
CHAPTER 7
What Can We Conclude?
To sum up, the present study has shown how the interrelation between the
cultural and cognitive aspects of grammatical gender manifests itself in inter-
action. In the last chapter, I present a series of reflections on the complex
interplay between grammatical gender, culture and cognition, based on the
empirical analysis presented in the previous chapter. These reflections con-
cern three issues: i) the relation between language and cognition in interaction
(section 7.1), ii) the covert reproduction of sexism in interaction (section 7.2),
and iii) the approach of gender performativity through the lens of the present
study (section 7.3). Certain reflections on interdisciplinarity in language and
gender research are also presented in the final section.
7.1 ‘Thinking for Speaking’ in Interaction
The investigation of the cognitive role of grammatical gender in interaction

concerns just one of the various aspects of the relation between language and
cognition in interaction in general. As Ι showed in chapter 6, interaction oper-
ates as a sort of ‘natural lab’ in which analysts can explore the relation between
language system and speakers’ cognition.
The indications found for the cognitive role of grammatical gender are com-
patible with Slobin’s (1996, 2003) hypothesis on the role of language in medi-
ating speakers’ thinking for speaking. More specifically, when speakers repair
grammatical gender, they seem to intervene in the mental representations
constructed in interaction, when they produce and understand utterances. In
this way, repairs indicate that grammatical gender guides speakers to attend
to referents’ sex as an aspect of social context when they think for speaking,
and, thus, they open a ‘window’ to speakers’ mental world. Therefore, the rela-
tion between language and cognition can be examined empirically not only
through experimental methods, but in naturally occurring interaction as well,
by analyzing what speakers do in their everyday social life.
In addition, repairs indicate that language mediates speakers’ cognition
without fully determining it. For instance, in Extracts 21, 22, 23 and 24 (cf. sec-
tion 6.3.2), speakers deviate from a pattern of usage, i.e. the use of the mas-
culine for gender indefinite reference or reference to female persons only,
and employ the alternative choices provided by language system, that is, the

What Can We Conclude ? 167
feminine gender. In these cases, the routine and covert meanings produced
in person reference in interaction come to the surface of the talk and are
noticed by speakers. More specifically, speakers recognize the automatic and
unconscious conceptual processes related with the use of grammatical gender
and modify them. In this sense, repairs show that language mediates speak-
ers’ thinking but does not ‘imprison’ speakers into static conceptualizations
of reality.1
Three points need to be highlighted here with respect to repairs of gram-
matical gender. First, the small number of repairs found in the corpus is justi-
fied by the fact that grammatical gender is inherent in the language system, is
used in a compulsory and systematic way, and guides speakers to unconscious
and automatic interpretations of experience (cf. chapters 2, 4). Based on that,
it should be expected that speakers rarely become aware of the cognitive role
of grammatical gender and notice the indexical effect associated with items
marked by grammatical gender. Second, the analysis of repairs shows that the
hidden and automatic meanings produced in interaction come to the surface
of the talk because aspects of context shed light on them. The information of
referent’s sex becomes relevant for what participants do in interaction and,
thus, it is foregrounded in the conversational episodes examined, and noticed
by participants. This is an important point that should be taken into account
by research on language and cognition. In exploring the ‘effect’ of language
structure on speakers’ cognition, the context in which this structure is put into
use should also be taken into consideration. In other words, we should explore
not only what speakers must say, but also when and why they say it. Third, it is
important to keep in mind that treating repairs of grammatical gender as indi-
cations of the cognitive role of grammatical gender is an interpretation made
by the analyst on the basis of the empirical analysis of interaction data. As long
as this interpretation is subjected to the limits of the theoretical and method-
ological perspectives adopted in this study, repairs cannot be said to ‘prove’ the
cognitive role of grammatical gender in an objective manner.
The findings reported in this study with respect to grammatical gender and
cognition raise an interesting question, given the scope of structural relativity
(cf. chapter 4). If the use of grammatical gender in Greek presupposes a par-
ticular kind of ‘thinking for speaking’, how does this sort of thinking differ from
‘thinking for speaking’ in languages with no gender systems? For example, is
1 This finding is opposed to linguistic determinism, which views language as a medium shap-
ing speakers’ thinking in absolute ways. If language had a deterministic impact on what
speakers actually do or on what they can think, then it would be impossible for them to
intervene in the mental representations constructed in interaction.
168 chapter 7
it possible that sex is a more salient concept for Greek speakers than Finnish
speakers? Or to phrase it differently, is it possible that the Greek language
guides speakers to a bipolar gender categorization of the world more system-
atically than the English language for example? As was shown in the analysis,
grammatical gender structures participants’ practices in interaction by cat-
egorizing referents as women or men and guiding speakers to attend to sex as
an aspect of context. Even if speakers wanted to escape from the static gen-
dered categorization of self and others, the language they speak enforces and
maintains this categorization on the basis of social hierarchy. These habits of
thinking associated with the use of grammatical gender should be absent from
languages in which sex is not grammaticized. If this is true, then genderless
languages should provide speakers with more possibilities for an egalitarian
representation of women and men via the use of gender-indefinite terms, as
Hellinger and Bussmann (2001, 20) note. Things, however, are not that simple.
As Whorf (1956) argues, covert classifications or cryptotypes may be far more
significant in terms of their impact on speakers’ cognition compared with
overt classifications or phenotypes. If Whorf’s idea is correct, then genderless
languages, in which sex is lexically specified or stereotypically assumed, may
guide speakers’ interpretation of the world in more systematic, unconscious
and significant ways than languages in which sex is grammaticized. In addi-
tion, as Hellinger and Bussmann (2001, 20) point out, it may be more difficult
to challenge sexism in genderless languages, because of the “covert male bias”
found in many nouns denoting humans (cf. section 4.3).
7.2 Sexism in Interaction: Fact or Fiction?
As the analysis shows, gender hierarchy is reinforced and sustained through

the generic use of the masculine grammatical gender in Greek interaction.
In chapter 4, I assumed that the generic use of the masculine is linked with
the inference of human/universal in certain contexts. The empirical analy-
sis of interaction data confirmed this correlation. The most direct indication
of the conceptual processes associated with the generic use of the mascu-
line was found in the case of repairs. Repairs of the masculine grammatical
gender revealed that the inference of human/universal is based on the male
categorization of referents. However, as I showed, the inference of human/
universal is not activated always in non-recognitional third person reference,
because referring to a stereotypically masculine activity or behavior may favor
the interpretation of referents’ sex as exclusively male. The relation between
the use of the masculine gender and the interpretation of referents’ sex in
non-recognitional third person reference is an issue that needs to be further

examined.
The challenging question emerging from the analysis of interaction data is
whether sexism should be treated as a problem for the conversational episodes
examined, although it is not a problem for participants. For instance, even when
speakers repair the generic use of the masculine and notice the male catego-
rization of the referent imposed by the masculine gender, sexism is not the
reason why they do so. Sexism is not foregrounded interactionally in none of
these cases. Given that, the following question arises. Should we analysts con-
sider sexism as an aspect of interaction that remains hidden and unnoticed by
participants, like power relations usually do, or should we consider it as a piece
of ‘fiction’, an analytic category that we impose externally on the data, and,
therefore, we must exclude from our analysis, following ca’s principles?
I will answer this question by drawing both on the theoretical framework of
this study and the empirical analysis. According to linguistic relativity, gram-
matical categories guide speakers to automatic and unconscious interpreta-
tions of experience (cf. section 4.2.2). In addition, part of the meanings that
is produced in interaction is not explicitly articulated. It is rather expressed
through covert assumptions and preferred interpretations (cf. section 5.2).
This means that speakers are not totally aware of all the different things hap-
pening in interaction, while they perform their everyday social life. The study
presented in this book shows that grammatical gender is designed to pass
unnoticed in interaction and produce a sense of ordinariness in Sacks’ (1984)
terms. Because it is inherent in language structure, its function cannot be eas-
ily detected. A lot of things are accomplished in interaction by language itself,
at the background of social actions.
In addition, as Kitzinger (2000, cf. section 5.3.1) points out, power is main-
tained in interaction, because participants rarely attend to it. This is not big
news for feminist scholars working in the framework of postmodern/post-
structural theories (cf. section 3.2). As we saw in chapter 3, Foucault’s (1982)
concept of power is central in this framework. In Foucault’s terms, power is
not related to a single source, and is not an attribute, held by one or more indi-
vidual. It is a relation, defined by multiplicity, with productive and repressive
aspects. It is exercised through micro-techniques that construct subjects and
‘naturalize’ the social world in which subjects live. This idea of power can help
us explain why participants in interaction keep the stereotypical assumptions
on which their world is based unquestioned. Sexism, like other power rela-
tions, is achieved in routine ways in interaction.
Moreover, the analysis of repairs of the masculine grammatical gender
point to a relation between the generic use of the masculine and the covert
170 chapter 7
reproduction of sexism in interaction. Repairs reveal that what is codified

morphologically (masculine gender-male sex) leaves its cognitive trace on
speakers’ minds and limits speakers’ thinking in specific ways. Speakers inter-
pret referent(s) as male and infer the meaning of human/universal, depending
on context. Yet, this inference is shaped by the ‘tension’ that exists between
the economy principle in person reference in interaction and the mediation
of language structure to speakers’ thinking. Speakers use grammatically mas-
culine reference forms to refer to male plus female or gender indefinite per-
sons, following the principle of economy in interaction. However, the use of
the masculine gender enforces speakers to categorize referent(s) as male, and
‘hides’ female sex. This is why speakers need to return to grammatically mas-
culine items and replace them with the feminine ones.
Overall, interaction is ‘colonized’ by the masculine grammatical gender, that
is, it is ‘gendered’ as male. This means that the concept of male sex is made
conceptually more salient than female sex for Greek speakers in interaction.
Men become more ‘visible’ than women in the cognitive universe that partici-
pants jointly build in interaction. The sociocultural world that is constructed
and maintained in interaction is a man’s world. This is an assumption that
remains unquestioned in interaction as part of the commonsense knowledge
that participants share. Discrimination against women and gender inequal-
ity is a product materialized through ‘micro’ everyday routine practices that
participants employ in interaction. In their daily lives, speakers construct the
asymmetrical bipolar gender order, because they subscribe to a silent agree-
ment about it, to remember Whorf (1956, cf. section 4.2.2).
So far so good one may think. But there is still an issue that needs to be
addressed. What happens when women ‘see’ themselves through the use of
the masculine gender and employ grammatically masculine forms to refer to
themselves plus other women? Is it possible that in this case the masculine
gender is simply understood as a medium of generalization, dissociated from
its ‘male’ indexical effect? As was shown in section 6.2.5, masculine linguis-
tic items can function as indexes of ‘humanness’ and ‘plural number’, when
they are used for reference to female persons only. However, the metonymic
association of man with the norm is still present in this use. It is presupposed
and reinforced every time women use the masculine gender in collective self-
reference or recipient reference.
Based on the discussion so far, I suggest that participants’ orientation in
interaction can be overt but also covert. Overt orientation involves partici-
pants’ recognition of and attention to specific aspects of context (e.g. rela-
tions, identities), displayed on the interaction data, while covert orientation
involves the role of language in guiding speakers to unconscious and automatic
i nterpretations of experience, when they think for speaking. Both orientations

shape the intersubjective understandings that participants jointly build in
interaction.
7.3 Reconsidering Gender Performativity
In this last section, I turn to the concept of gender performativity that was
introduced in chapter 3, and I present some preliminary remarks on how this
concept can be re-signified through the lens of the present study. I consider this
process of re-signification as an interdisciplinary process between linguistic
and non-linguistic approaches to gender. In my opinion, linguists need to take
this interdisciplinary step in order to gain a fuller understanding of the com-
plexities of the relation between language and gender. As Cameron (1996, 33)
insightfully points out, “language and gender studies, like other subfields
within sociolinguistics, has tended to neglect theoretical questions about its
‘socio’ side (in this instance, gender)”. As a consequence, Cameron (1996, 33)
claims that “linguists are cut off from insights that would be relevant to their
work, while feminists in other disciplines can continue to talk about language
in ways that are not accountable to the specialized knowledge linguistics
makes available”. In the following, I explore how the study presented in this
book contributes to the interdisciplinary dialogue between linguistic and non-
linguistic approaches.
Following Lykke (2004, 97), I understand interdisciplinarity as a process of
transgressing “borders between disciplinary canons and approaches in a theo-
retical and methodological bricolage that allows for new synergies to emerge”.
These synergies can emerge from the exchange or ‘travelling’ (Bal 2002) of con-
cepts between disciplines and their subsequent integration and transforma-
tion in different disciplinary contexts. Interdisciplinarity consists of moving in
and out of disciplines and creating hybrid spaces, in Latour’s (2000) terms, in-
between disciplines. The concept of performativity constitutes a typical case
of travelling from linguistics to feminist philosophy and back to linguistics.
The concept of performativity proposed by Butler has been central to
Austin’s speech act theory, which theorizes language as a tool for ‘doing things’
or performing actions (cf. section 3.4). Based on this idea, Butler argues that
gender is performative, in the sense that it constitutes an accomplishment, a
‘doing’ rather than a ‘being’, an outcome of repeated practices. Butler’s concept
of gender performativity informed research on language and gender. An origi-
nally linguistic concept returned to linguistics in a slightly modified version,
with a renewed feminist philosophical content. The concept was integrated
172 chapter 7
within the ‘discourse’ or ‘performance’ turn in language and gender research,

which explores the construction of gender through linguistic practices. So,
performativity re-entered language and gender research through the lens of
Butler’s work. Moreover, it got creatively integrated and expanded through lin-
guistic studies, which made the abstract concept of performativity more con-
crete and specified the linguistic practices through which gender is performed.
This process of travelling from feminist philosophy to linguistics involves a set
of ‘transgressive steps’, which blur disciplinary boundaries and open up hybrid
spaces. These spaces consist of elements of knowledge from both feminist phi-
losophy and linguistics, and re-signify the content of performativity.
How is the present study located in these transgressive steps? This study
examines in detail how participants categorize self and others on the basis of
gender bipolarity and how gender hierarchy is maintained and reinforced as
a feature of the sociocultural world at the micro-level of interaction. For this
reason, it contributes to the systematic investigation of the performative prac-
tices through which gender is constructed. It also contributes to our under-
standing of gender performativity in an additional way. The perspective on
the interplay between language, culture and cognition employed in this study
links the notion of linguistic practice with speakers’ cognition, and, in this way,
highlights the cognitive depth of gender performativity, and foregrounds the
‘material effect’ of linguistic practices.
More specifically, by ascribing sex to referents and categorizing them as
women or men grammatical gender establishes habits of thinking of the world
in terms of pre-existing substances and attributes that subjects hold. In this
way, it sustains the man/woman dichotomy as a dominant ideological schema
in our society. The bipolar categorization of humans into women/men is
enforced by language as a ‘given’ feature of reality. Naming a person via a gram-
matically masculine or feminine form is a powerful practice that interpellates
this specific person into social existence. In this sense, language is a big ally
that the dominant gender order has in maintaining itself. Is language also one
of the reasons why it is so difficult to transgress the female/male sex dichotomy
and challenge the binary conceptualization of the world? It could be.
The cognitive depth of gender performativity has implications also for
the way we think about another binary, the material-symbolic. Because lan-
guage mediates speakers’ thinking of the world, it ‘shapes’ the sociocultural
world that speakers live in. It guides speakers to attend to specific features of
experience and makes these features conceptually salient. Sex is one of these
features that Greek speakers must attend to every time they speak. There is a
correlation between this habit of thinking established by linguistic practices
and the material practices employed at the social level for gendering the world.
For instance, subordinating women is a material process that is also related

with language. Verbal violence or sexism in language may be—or, at least,
seem to be—less important for women’s lives, when it is compared with sexual
harassment at workplace or domestic violence. Yet, language plays its part in
constructing and sustaining a world in which the above practices arise: a world
in which men enjoy the privilege of universal subject and women are reduced
to the status of deviant and subordinate other. In broad terms, linguistic
practices affect and they are also affected by the materiality of the lives that
speakers lead.
Of course, there are limits with respect to what language ‘can’ do. Grammati-
cal gender may advance speakers’ binary thinking of humans but it does not
fully determine their thinking. First, as a number of studies show, all discussed
in section 3.3.1, speakers use grammatical gender to construct gender identi-
ties that deviate from the norm of compulsory heterosexuality. Thus, speakers
are not fully trapped by the bipolarities that language imposes on them, and
can find ways to be creative and subversive. Second, as the analysis of repairs
show (cf. section 6.3.2), speakers can always select one of the two forms made
available by language system, and construe experience in alternative ways.
Third, based on cognitive linguistics (cf. section 4.2.1), cognitive categories
are not static and fixed; they have fuzzy boundaries and are organized on the
basis of prototypes. This means that the concepts of sex/gender may be present
when speakers think for speaking but they are not static and are historically
and culturally defined. As Hellinger and Bussmann (2001, 8) insightfully point
out, “although the terms ‘female’ and ‘male’ contribute to the construction of
people’s everyday experience, they might perhaps be more adequately placed
on a continuum, which allows for variation, fuzzy category boundaries, and pro-
totype effects (cf. Lakoff 1987)”. Language construes experience in specific ways,
but these ways can always change due to social, cultural and historical factors.
Reconsidering performativity as a process related to cognition brings the
role of language reform to the frontline again (cf. section 3.3.2). As Cameron
(1985, 1) argues, feminists have always considered language as “an essential
part of the struggle for liberation”. However, a lot of tension is involved in
the debates on language reform as a vehicle generating or facilitating social
reform. Is language reform something to come after social change (Lakoff 1975)
or before it (Spender 1980)? Is the elimination of linguistic sexism a tool for
eliminating social sexism? Or, is it useless and ineffective as long as patriarchy
remains unchallenged? For instance, the address term Μs. was introduced by
the feminist movement as a replacement term for the term Mrs. In the long
run, it ended up denoting the non-married woman and being used together
with the terms Mrs. and Miss (Romaine 1999, 310). Thus, a feminist linguistic
174 chapter 7
activist practice ended up being assimilated by the dominant patriarchal order.

As Cameron (1985, 90) acknowledges, “in the mouths of sexists, language can
always be sexist”. Language is not ‘guilty’ for every gender inequality that exists
in the world but it is certainly ‘accountable’ for it, at least partly. It constitutes
“part of social change”, according to Thorne and Henley (1975, 30), and a tool
for re-signifying women in positive and affirmative ways.
When it comes to the use of grammatical gender systems, feminist language
planning faces a challenge. Gender neutralization and gender specification
(cf. section 3.3.2) are strategies that can work successfully in written language
to achieve a symmetrical representation of women and men, e.g. through
personal use or pressure on media, education and legislative writing to adopt
guidelines for non-sexist language use (Pauwels 2003). However, when it
comes to interaction, things are more complicated. For instance, avoiding the
generic use of the masculine by repeating the feminine form whenever the
masculine form is used does not align with the principle of economy in person
reference in interaction. This means that feminist language planning in inter-
action involves more effort, political awareness and creativity from speakers’
part. For example, using the feminine gender for generic reference occasion-
ally in interaction may be a way to break grammatical conventions and expose
the “subtle ways” (Pauwels 2003, 555) in which women are discriminated in
language. Yet, such a disruption cannot be accomplished easily on a daily basis.
The difficulties that language planning encounters when it comes to interac-
tion should not prevent us from aiming to feminist language reform in general.
For instance, the use of grammatical gender in written texts is still expected
to mediate speakers’ thinking at the time of reading, as Slobin (2003) suggests,
and contributes to the construction of the sociocultural world (e.g. see Livia
2001 for an analysis of the literary uses of grammatical gender, section 3.3.1).
Gender performativity is central in any discussion about the role of feminist
language reform. As Cameron ([1995] 2005, 16) argues,
The reason we perceive ourselves, and are perceived by others, as particu-

lar kinds of people is that we repeat the actions that define those kinds
of people until in time they come to seem like a fixed and integral part
of our nature.
This means that different linguistic practices, oral or written, give rise to dif-
ferent ways of experiencing the world and, therefore, construct different sorts
of subjectivities. The aim of this book is not to provide guidelines on exactly
how this can be done. But it certainly invites feminists to think why language
should be one of the key issues to be raised in our political agenda.
7.4 Grammatical Gender in Interaction: A Summary
The study presented in this book shows that the use of grammatical gender is
part of the practices that speakers employ to construct themselves and others
as women or men, and maintain gender bipolarity on the basis of social hierar-
chy. Grammatical gender contributes to the routine achievement of sociocul-
tural gender in interaction. This process is materialized via covert assumptions
that speakers share about referent’s gender. In addition, grammatical gender
is shown to guide speakers to interpret person as female or male, when they
think for speaking. Indications of the cognitive role of grammatical gender in
person reference are found in the interaction data, in participants’ publicly
displayed, intersubjective understandings.
References
Aikhenvald, Alexandra Y. 1999. “The Arawak Language Family.” In The Amazonian Lan-
guages, edited by R.M.W. Dixon and Alexandra Y. Aikhenvald, 65–106. Cambridge:
Cambridge University Press.
———. 2000. Classifiers: A Typology of Noun Categorization Devices. Oxford: Oxford
University Press.
———. 2004. Evidentiality. Oxford: Oxford University Press.
———. 2006. “Classifiers and Noun Classes: Semantics.” In Encyclopedia of Languages
and Linguistics, edited by Keith Brown, 463–471. Oxford: Elsevier.
———. 2011. “Speech Reports: A Cross-linguistic Perspective.” In Language at Large:
Essays on Syntax and Semantics, edited by Alexandra Y. Aikhenvald and R.M.W. Dixon,
290–326. Leiden: Brill.
———. 2012. “Round Women and Long Men: Shape and Size in Gender Choice in
Papua New Guinea and Beyond.” Anthropological Linguistics 54(1): 33–86.
Alpher, Barry. 1987. “Feminine as the Unmarked Grammatical Gender: Buffalo Girls are
no Fools.” Australian Journal of Linguistics 7: 169–187.
Althusser, Louis. 1984. Essays on Ideology. London: Verso.
Alvanoudi, Angeliki. 2008. “Η Σχέση Γλώσσας και Φύλου στις Φεμινιστικές Θεωρίες της
Αποδόμησης: Η Περίπτωση του Γραμματικού Γένους” [The Relationship between Lan-
guage and Gender in Feminist Deconstruction Theories: The Case of Grammatical
Gender]. Studies in Greek Linguistics 28: 89–99.
———. (forthcoming). The routine achievement of gender in Greek conversation
[Special issue]. Gender and Language.
Anastasiadi-Symeonidi, Anna, and Despina Chila-Markopoulou. 2003. “Συγχρονικές και
Διαχρονικές Τάσεις στο Γένος της Ελληνικής: Μια Θεωρητική Πρόταση” [Synchronic
and Diachronic Trends in Gender of Modern Greek: A Theoretical Proposition]. In
Gender, edited by Anna Anastasiadi-Symeonidi, Angeliki Ralli and Despina Chila-
Markopoulou, 13–56. Athens: Patakis.
Andermahr, Sonya, Terry Lovell, and Carol Wolkowitz. 2000. A Glossary of Feminist
Theory. London: Arnold.
Asher, R.E. 1985. Tamil. London: Croom Helm.
Austin, John L. 1962. How to Do Things with Words. Oxford: Oxford University Press.
Austin, Peter. 1981. A Grammar of Diyari, South Australia. Cambridge: Cambridge Uni-
versity Press.
Bal, Mieke. 2002. Travelling Concepts in the Humanities: A Rough Guide. Toronto: Uni-
versity of Toronto Press.
Barcelona, Antonio. 2000. “Introduction: The Cognitive Theory of Metaphor and
Metonymy.” In Metaphor and Metonymy at the Crossroads: A Cognitive Perspective,
edited by Antonio Barcelona, 1–28. Berlin/New York: Mouton de Gruyter.
178 references
Beauvoir, Simone de. [1949] 1972. The Second Sex. Translated and edited by Howard M.
Parshley. Harmondsworth: Penguin.
Bergvall, Victoria L., Janet M. Bing, and Alice F. Freed, eds. 1996. Rethinking Language
and Gender Research: Theory and Practice. London: Longman.
Berman, Ruth A., and Dan Isaac Slobin. 1994. Relating Events in Narrative: A Crossling-
uistic Developmental Study. Hillsdale, nj: Lawrence Erlbaum.
Bickel, Balthasar. 2000. “Grammar and Social Practice: The Role of ‘Culture’ in Linguis-
tic Relativity.” In Evidence for Linguistic Relativity, edited by Susanne Niemeier and
René Dirven, 161–191. Amsterdam/Philadelphia: John Benjamins.
Boas, Franz. [1911] 1991. Introduction to Handbook of American Indian Languages.
Edited by Preston Holder, 1–79. Lincoln: University of Nebraska Press.
———. 1938. “Language.” In General Anthropology, edited by Franz Boas, 124–145.
Boston, New York: D.C. Heath and Company.
Borba, Rodrigo, and Ana C. Ostermann. 2007. “Do Bodies Matter? Travestis’ Embodi-
ment of (Trans)gender Identity through the Manipulation of the Brazilian Portug-
ese Grammatical System.” Gender and Language 1(1): 131–147.
Boroditsky, Lera, Lauren A. Schmidt, and Webb Phillips. 2003. “Sex, Syntax, and Seman-
tics.” In Language in the Mind: Advances in the Study of Language and Thought, edited
by Dedre Gentner and Susan Goldin-Meadow, 61–79. Cambridge, ma: mit Press.
Bowerman, Μelissa, and Stephen C. Levinson. 2001. Language Αcquisition and Concep-
tual Development. Cambridge: Cambridge University Press.
Braidotti, Rosi. 1991. Patterns of Dissonance: A Study of Women in Contemporary Philoso-
phy. Oxford: Polity.
———. 1994. Nomadic Subjects: Embodiment and Sexual Difference in Contemporary
Feminist Theory. New York: Columbia University Press.
———. 2000. “The Relevance of Gender for the Issue of ‘Women and Science’.” In The
Making of European Women’s Studies: A Work in Progress Report on Curriculum
Development and Related Issues, Vol. I, edited by Rosi Braidotti and Esther Vonk,
187–201. ATHENA/Utrecht University: Utrecht.
———. 2002a. “The Uses and Abuses of the Sex/Gender Distinction in European Fem-
inist Practices.” In Thinking Differently: A Reader in European Women’s Studies,
edited by Rosi Braidotti and Gabriele Griffin, 285–307. London/New York: Zed
Books.
———. 2002b. Metamorphoses: Towards a Materialist Theory of Becoming. Cambridge:
Polity Press.
———. 2003. “Feminist Philosophies.” In A Concise Companion to Feminist Theory,
edited by Mary Eagleton, 195–214. Oxford: Blackwell.
Braun, Friederike. 2001. “The Communication of Gender in Turkish.” In Gender across
Languages: The Linguistic Representation of Women and Men, edited by Marlis
Hellinger and Hadumod Bussmann, 283–310. Amsterdam: John Benjamins.
References 179
Braun, Friederike, Anja Gottburgsen, Sabine Sczesny, and Dagmar Stahlberg. 1998.
“Können Geophysiker Frauen sein? Generische Personenbezeichnungen im
Deutschen.” Zeitschrift für Germanistische Linguistik 26: 265–283.
Brown, Penelope. 2007. “Principles of Person Reference in Tzeltal Conversation.” In
Person Reference in Interaction: Linguistic, Cultural and Social Perspectives, edited by
N.J. Enfield and Tanya Stivers, 172–202. Cambridge: Cambridge University Press.
Brown, Penelope, and Stephen C. Levinson. 1987. Politeness: Some Universals in Lan-
guage Usage. Cambridge/New York: Cambridge University Press.
Bucholtz, Mary, and Kira Hall. 2005. “Identity and Interaction: A Sociocultural Linguis-
tic Approach.” Discourse Studies 7(4–5): 584–614.
Bussmann, Hadumod, and Marlis Hellinger. 2003. “Engendering Female Visibility in
German.” In Gender across Languages: The Linguistic Representation of Women and
Men, edited by Marlis Hellinger and Hadumod Bussmann, 141–174. Amsterdam:
John Benjamins.
Butler, Judith. [1990] 1999. Gender Trouble: Feminism and the Subversion of Identity.
London/New York: Routledge.
———. 1993. Bodies that Matter: On the Discursive Limits of Sex. New York: Routledge.
———. 1997. Excitable Speech: A Politics of the Performative. London: Routledge.
Cameron, Deborah. 1985. Feminism and Linguistic Theory. London: Macmillan.
———. [1995] 2005. Verbal Hygiene. London/New York: Routledge.
———. 1996. “The Language-Gender Interface: Challenging Co-optation.” In Rethink-
ing Language and Gender Research: Theory and Practice, edited by Victoria L.
Bergvall, Janet M. Bing and Alice F. Freed, 31–53. London: Longman.
Chila-Markopoulou, Despina. 2003. “Γένος και Συμφωνία στη Νέα Ελληνική” [Gender and
Agreement in Modern Greek]. In Gender, edited by Anna Anastasiadi-Symeonidi,
Angeliki Ralli and Despina Chila-Markopoulou, 132–167. Athens: Patakis.
Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, ma: mit Press.
Christidis, Anastasios Ph. 1999. Γλώσσα, Πολιτική, Πολιτισμός [Language, Politics, Cul-
ture]. Athens: Νisos.
Clark, Herbert H. 1996. Using Language. Cambridge: Cambridge University Press.
Clarke, Mark A., Ann Losoff, Margaret Dickenson McCracken, and JoAnn Still. 1981.
“Gender Perception in Arabic and English.” Language Learning 31: 159–169.
Claudi, Ulrike. 1985. Zur Entstehung von Genussystemen: Überlegungen zu einigen Theo-
retischen Aspekten, Verbunden mit einer Fallstudie des Zande: Mit einer Bibliographie
und einer Karte. Hamburg: Buske.
Clift, Rebecca, and Elizabeth Holt. 2007. “Introduction.” In Reporting Talk: Reported
Speech in Interaction, edited by Elizabeth Holt and Rebecca Clift, 1–15. Cambridge:
Connell, Raewyn. 2002. Gender. Cambridge: Polity Press.
Corbett, Greville, G. 1991. Gender. Cambridge: Cambridge University Press.
180 references
Corpus of Spoken Greek of the Institute of Modern Greek Studies. Accessed June 12, 2014.
http://corpus-ins.lit.auth.gr/corpus/about/symbols.html.
Croft, William, and D. Alan Cruse. 2004. Cognitive Linguistics. Cambridge: Cambridge
University Press.
Curzan, Anne. 2003. Gender Shifts in the History of English. Cambridge: Cambridge Uni-
versity Press.
Derrida, Jacques. [1972] 1988. “Signature, Event, Context.” In Limited, Inc, edited by Ger-
ald Graff. Translated by Samuel Weber and Jeffrey Mehlman. Evanston: Northwest-
ern University Press.
Dictionary of Standard Modern Greek. 1998. Thessaloniki: Institute of Modern Greek
Studies (Manolis Triantafyllidis Foundation).
Dixon, R.M.W. 1972. The Dyirbal Language of North Queensland. London: Cambridge
University Press.
———. 1980. The Languages of Australia. Cambridge: Cambridge University Press.
———. 2002. Australian Languages: Their Nature and Development. Cambridge: Cam-
bridge University Press.
———. 2010a. Basic Linguistic Theory. Vol. 1. Methodology. Oxford: Oxford University
Press.
———. 2010b. Basic Linguistic Theory. Vol. 2. Grammatical Topics. Oxford: Oxford Uni-
versity Press.
Doleschal, Ursula. 1993. “Genus als Grammatische und Textlinguistische Kategorie. Eine
Kognitiv-funktionalistische Untersuchung des Russischen”. Ph.D. diss., University of
Vienna.
———. 1997. “O Vzaimosvjazi Grammatičeskoj Kategorii Roda i Pola” [On the Interde-
pendence of the Grammatical Category Gender and Sex]. In Verbal and Non-Verbal
Interpretation of Gender Characteristics, edited by Aleksandr Xolod, 134–155. Krivoj
Rog: Meždunarodnyj Issledovatel’skij Centr Čelovek-Jazyk-Kul’tura-Poznanie.
Doleschal, Ursula, and Sonja Schmid. 2001. “Doing Gender in Russian: Structure and
Perspective.” In Gender across Languages: The Linguistic Representation of Women
and Men, edited by Marlis Hellinger and Hadumod Bussmann, 253–282. Amster-
dam: John Benjamins.
Drew, Paul, Traci Walker, and Richard Ogden. 2013. “Self-repair and Action Construc-
tion.” In Conversational Repair and Human Understanding, edited by Makoto
Hayashi, Geoffrey Raymond, and Jack Sidnell, 71–94. Cambridge: Cambridge Uni-
versity Press.
Eakins, Barbara W., and Rollin G. Eakins. 1978. Sex Differences in Human Communica-
tion. Boston: Houghton Mifflin.
Eckert, Penelope, and Sally McConnell-Ginet. 1992. “Think Practically and Look
Locally: Language and Gender as Community-Based Practice.” Annual Review of
Anthropology 21: 461–490.
References 181
———. 2003. Language and Gender. Cambridge: Cambridge University Press.

Enfield, N.J. 2000. “On Linguocentrism.” In Explorations in Linguistic Relativity, edited
by Martin Pütz and Marjolijn Verspoor, 125–157. Amsterdam/Philadelphia: John
Benjamins.
———. 2002. “Ethnosyntax: Introduction.” In Ethnosyntax: Explorations in Grammar
and Culture, edited by N.J. Enfield, 3–30. Oxford: Oxford University Press.
———. 2007. “Meanings of the Unmarked: How Default Person Reference Does more
than just Refer.” In Person Reference in Interaction: Linguistic, Cultural and Social
Perspectives, edited by N.J. Enfield and Tanya Stivers, 97–120. Cambridge: Cambridge
University Press.
———. 2012. “Reference in Conversation.” In The Handbook of Conversation Analysis,
edited by Jack Sidnell and Tanya Stivers, 433–454. Boston: Wiley-Blackwell.
Enfield, N.J., and Tanya Stivers, eds. 2007. Person Reference in Interaction: Linguistic,
Cultural and Social Perspectives. Cambridge: Cambridge University Press.
Engelberg, Μila. 2002. “The Communication of Gender in Finnish.” In Gender across
Ervin, Susan M. 1962. “The Connotations of Gender.” WORD 18(3): 249–261.
Evans, Vyvyan, and Melanie Green. 2006. Cognitive Linguistics: An Introduction. Edin-
burgh: Edinburgh University Press.
Fausto-Sterling, Anne. 2000. Sexing the Body: Gender Politics and the Construction of
Sexuality. New York: Basic Books.
Ferguson, Charles A. 1964. “Diglossia.” In Language in Culture and Society: A Reader in
Linguistics and Anthropology, edited by Dell H. Hymes, 429–439. New York/London:
Harper and Raw. Originally published in Word 15: 325–340 (1959).
Flaherty, Mary. 2001. “How a Language Gender System Creeps into Perception.” Journal
of Cross-Cultural Psychology 32(1): 18–31.
Fodor, Jerry A. 1975. The Language of Thought. New York: Crowell.
———. 1983. The Modularity of Mind. Cambridge, ma: mit Press.
Foley, William A. 1986. The Papuan Languages of New Guinea. Cambridge: Cambridge
University Press.
———. 1991. The Yimas Language of New Guinea. Stanford: Stanford University
Press.
———. 1997. Anthropological Linguistics: An Introduction. Oxford: Blackwell.
Foucault, Michel. 1982. “Afterward: The Subject and Power.” In Michel Foucault: Beyond
Structuralism and Hermeneutics, edited by Hubert L. Dreyfus and Paul Rabinow,
208–226. Chicago: University of Chicago Press.
Frank, Francine W., and Paula A. Treichler. 1989. Language, Gender, and Professional
Writing: Theoretical Approaches and Guidelines for Non-sexist Usage. New York:
Modern Language Association.
182 references
Freed, Alice F. 2003. “Epilogue: Reflections on Language and Gender Research.” In The
Handbook of Language and Gender, edited by Janet Holmes and Miriam Meyerhoff,
699–721. Oxford: Blackwell.
Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs, N.J.: Prentice-
Hall.
Gentner, Dedre, and Susan Goldin-Meadow, eds. 2003. Language in the Mind: Advances
in the Study of Language and Thought. Cambridge, ma: mit Press.
Gleitman, Lila, and Anna Papafragou. 2005. “Language and Thought.” In Cambridge
Handbook of Thinking and Reasoning, edited by Keith J. Holyoak and Robert G.
Morrison, 633–661. Cambridge: Cambridge University Press.
Goffman, Erving. 1981. Forms of Talk. Oxford: Blackwell.
———. 1983. “The Interaction Order: American Sociological Association, 1982 Presi-
dential Address.” American Sociological Review 48(1): 1–17.
Graham, Alma. 1975. “The Making of a Nonsexist Dictionary.” In Language and Sex: Dif-
ference and Dominance, edited by Barrie Thorne and Nancy Henley, 57–63. Rowley,
ma: Newbury House.
Grice, H. Paul. 1989. Studies in the Way of Words. London/Cambridge, ma.: Harvard Uni-
versity Press.
Grönberg, Anna G. 2002. “Masculine Generics in Current Icelandic.” In Gender across
Gumperz, John J. 1996. “The Linguistic and Cultural Relativity of Conversational Infer-
ence.” In Rethinking Linguistic Relativity, edited by John J. Gumerz and Stephen C.
Levinson, 374–406. Cambridge: Cambridge University Press.
Gumperz, John J., and Stephen C. Levinson, eds. 1996a. Rethinking Linguistic Relativity.
Cambridge: Cambridge University Press.
———. 1996b. “Introduction: Linguistic Relativity Re-examined.” In Rethinking Lin-
guistic Relativity, edited by John J. Gumperz and Stephen C. Levinson, 1–18. Cam-
bridge: Cambridge University Press.
Hall, Kira, and Veronica O’Donovan. 1996. “Shifting Gender Positions among Hindi-
speaking Hijras.” In Rethinking Language and Gender Research: Theory and Practice,
edited by Victoria L. Bergvall, Janet M. Bing and Alice F. Freed, 228–266. London:
Longman.
Hanks, William F. 1990. Referential Practice: Language and Lived Space among the
Maya. Chicago: University of Chicago Press.
———. 1996. “Language Form and Communicative Practices.” In Rethinking Linguistic
Relativity, edited by John J. Gumperz and Stephen C. Levinson, 232–270. Cambridge:
———. 2007. “Person Reference in Yucatec Maya Conversation.” In Person Reference in
Interaction: Linguistic, Cultural and Social Perspectives, edited by N.J. Enfield and
Tanya Stivers, 149–171. Cambridge: Cambridge University Press.
References 183
Haraway, Donna. 1991. “Situated Knowledges: The Science Question in Feminism and
the Privilege of Partial Perspective.” In Simians, Cyborgs, and Women: The Reinven-
tion of Nature, 183–201. London: Free Association Books.
Harding, Sandra. 1991. “What is Feminist Epistemology?” In Whose Science? Whose
Knowledge, 105–137. Open up.
Haviland, John B. 2007. “Person Reference in Tzotzil Gossip: Referring Dupliciter.” In
Person Reference in Interaction: Linguistic, Cultural and Social Perspectives, edited
by N.J. Enfield and Tanya Stivers, 226–252. Cambridge: Cambridge University
Press.
Hayashi, Makoto, Geoffrey Raymond, and Jack Sidnell, eds. 2013. Conversational Repair
and Human Understanding. Cambridge: Cambridge University Press.
Heine, Bernd. 1982. “African Noun Class Systems.” In Apprehension: Das sprachliche
Erfassen von Gegenständen, i: Bereich und Ordnung der Phänomene, edited by
Hansjakob Seiler and Christian Lehmann, 189–216. Tübingen: Narr.
Hellinger, Marlis. 2001. “English: Gender in a Global Language.” In Gender across Lan-
guages: The Linguistic Representation of Women and Men, edited by Marlis Hellinger
and Hadumod Bussmann, 105–113. Amsterdam: John Benjamins.
Hellinger, Marlis, and Hadumod Bussmann. 2001. “Gender across Languages: The Lin-
guistic Representation of Women and Men.” In Gender across Languages: The Lin-
guistic Representation of Women and Men, edited by Marlis Hellinger and Hadumod
Bussmann, 1–25. Amsterdam: John Benjamins.
———, eds. 2001–2002–2003. Gender across Languages: The Linguistic Representation
of Women and Men. Amsterdam: John Benjamins.
Heritage, John. 1984. Garfinkel and Ethnomethodology. Cambridge: Polity Press.
———. 1988. “Explanations as Accounts: A Conversation Analytic Perspective.” In
Analyzing Everyday Explanation: A Casebook of Methods, edited by Charles Antaki,
127–144. London: Sage.
Heritage, John, and J. Maxwell Atkinson. 1984. “Introduction.” In Structures of Social
Action: Studies in Conversation Analysis, edited by J. Maxwell Atkinson and John
Heritage, 1–15. Cambridge: Cambridge University Press.
Ηeritage, John, and Geoffrey Raymond. 2005. “The Terms of Agreement: Indexing Epis-
temic Authority and Subordination in Talk-in-Interaction.” Social Psychology Quar-
terly 68(1): 15–38.
Hill, Jane H., and Bruce Mannheim. 1992. “Language and World View.” Annual Review of
Anthropology 21: 381–406.
Holmes, Janet, and Miriam Meyerhoff. 2003. “Different Voices, Different Views: An
Introduction to Current Research in Language and Gender.” In The Handbook of
Language and Gender, edited by Janet Holmes and Miriam Meyerhoff, 1–17. Oxford:
Blackwell.
Holton, David, Peter Mackridge, and Irene Philippaki-Warburton. 1997. Greek: A Com-
prehensive Grammar. London: Routledge.
184 references
Horrocks, Geoffrey C. 2010. Greek: A History of the Language and its Speakers. Wiley-
Blackwell.
Hougaard, Anders. 2008. “Compression in Interaction.” In Mental Spaces in Discourse
and Interaction, edited by Todd Oakley and Anders Hougaard, 179–208. Amsterdam/
Philadelphia: John Benjamins.
Huang, Yan. 1994. The Syntax and Pragmatics of Anaphora. Cambridge: Cambridge
University Press.
———. 2007. Pragmatics. Oxford: Oxford University Press.
Hymes, Dell H. 1966. “Two Types of Linguistic Relativity (with Examples from Amerin-
dian Ethnography).” In Sociolinguistics, Proceedings of the ucla Sociolinguistics Con-
ference, edited by William Bright, 114–157. The Hague: Mouton.
Imai, Mutsumi, Lennart Schalk, Henrik Saalbach, and Hiroyuki Okada. 2013. “All Giraffes
have Female-specific Properties: Influence of Grammatical Gender on Deductive
Reasoning about Sex-specific Properties in German Speakers.” Cognitive Science 1–23.
Irigaray, Luce. 1985a. The Speculum of the Other Woman. Ithaca: Cornell University
Press.
———. 1985b. This Sex which is not One. Ithaca: Cornell University Press.
———. 1985c. Speaking is never Neutral. Paris: Editions de Minuit.
———. 1993. Je, Tu, Nous: Towards a Culture of Difference. New York: Routledge.
Ιrmen, Lisa, and Astrid Köhncke. 1996. “Zur Psychologie des ‘generischen’ Maskuli-
nums.” Sprache und Kognition 15: 152–167.
Jefferson, Gail. 1978. “Sequential Aspects of Storytelling in Conversation.” In Studies in
the Organization of Conversational Interaction, edited by Jim Schenkein, 219–248.
New York: Academic Press.
———. 1984. “On Stepwise Transition from Talk about a Trouble to inappropriately
Next-positioned Matters.” In Structures of Social Action: Studies in Conversation
Analysis, edited by J. Maxwell Atkinson and John Heritage, 191–221. Cambridge:
———. 1987. “On Exposed and Embedded Correction in Conversation.” In Talk and
Social Organization, edited by Graham Button and John R.E. Lee, 86–100. Clevedon,
uk: Multilingual Matters.
———. 1996. “On the Poetics of Ordinary Talk.” Text and Performance Quarterly 16(1):
1–61.
Joseph, Brian, and Irene Philipaki-Warburton. 1987. Modern Greek. London: Croom Helm.
Kati, Dimitra. 2013. Ο Γνωσιακός Ρόλος της Γλώσσας: Η Συμβολή των Επιστημών της Γλώσσας
και του Νου [The Cognitive Role of Language: The Contribution of the Studies on
Language and Mind]. Athens: Nisos.
Kati, Dimitra, Marianna Kondyli, and Kiki Nikiforidou. 1999. “Εισαγωγή” [Introduc-
tion]. In Language and the Mind, edited by Dimitra Kati, Μarianna Κοndyli and Kiki
Νikiforidou, 11–42. Αthens: Αlexandreia.
References 185
Kitagawa, Chisato, and Adrienne Lehrer. 1990. “Impersonal Uses of Personal Pronouns.”
Journal of Pragmatics 14(5): 739–759.
Kitzinger, Celia. 2000. “Doing Feminist Conversation Analysis.” Feminism Psychology
10(2): 163–193.
———. 2005. “Speaking as a Heterosexual: (How) Does Sexuality Matter for Talk-in-
Interaction.” Research on Language and Social Interaction 38(3): 221–265.
———. 2007. “Is ‘Woman’ always relevantly Gendered?” Gender and Language 1(1):
39–49.
———. 2012. “Repair.” In The Handbook of Conversation Analysis, edited by Jack Sidnell
and Tanya Stivers, 229–256. Boston: Wiley-Blackwell.
Klein, Josef. 1988. “Die Benachteiligung der Frau im Generischen Maskulinum—eine
Feministische Schimäre oder Psycholinguistische Realität?” In Vorträge des Ger-
manisten-tages Berlin 1987. Vol. 1: Das Selbstverständnis der Germanistik, edited by
Norbert Oellers, 310–319. Tübingen: Niemeyer.
Konishi, Toshi. 1993. “The Semantics of Grammatical Gender: A Cross-cultural Study”.
Journal of Psycholinguistic Research 22(5): 519–534.
Koniuszaniec, Gabriela, and Hanka Blaszkowska. 2003. “Language and Gender in Pol-
ish.” In Gender across Languages: The Linguistic Representation of Women and Men,
edited by Marlis Hellinger and Hadumod Bussmann, 259–285. Amsterdam: John
Benjamins.
Köpcke, Klaus-Michael, and David A. Zubin. 2003. “Metonymic Pathways to Neuter-
Gender Human Nominals in German.” In Metonymy and Pragmatic Inferencing,
edited by Klaus-Uwe Panther and Linda L. Thornburg, 149–166. Amsterdam/
Philadelphia: John Benjamins.
Köpcke, Klaus-Michael, Klaus-Uwe Panther, and David A. Zubin. 2010. “Motivating
Grammatical and Conceptual Gender Agreement in German.” In Cognitive Founda-
tions of Linguistic Usage Patterns, edited by Hans-Jörg Schmid and Susanne Handl,
171–194. Berlin/New York: Walter de Gruyter.
Kramarae, Cheris, and Paula A. Treichler. 1985. A Feminist Dictionary. London: Pandora
Press.
Kulick, Don. 1998. Travesti: Sex, Gender, and Culture among Brazilian Transgendered
Prostitutes. Chicago/London: University of Chicago Press.
Lakoff, George. 1987. Women, Fire and Dangerous Things: What Categories Reveal about
the Mind. Chicago/London: The University of Chicago Press.
Lakoff, Robin. 1975. Language and Woman’s Place. New York: Harper and Row.
Langacker, Ronald. W. 1993. “Reference-Point Constructions.” Cognitive Linguistics 4(1):
1–38.
Latour, Bruno. 1993. We Have never been Modern. Translated by Catherine Porter. Cam-
bridge, ma.: Harvard University Press.
Lee, Penny. 1996. The Whorf Theory Complex. Amsterdam/Philadelphia: John Benjamins.
186 references
Lerner, Gene H. 1992. “Assisted Story Telling: Deploying Shared Knowledge as a Practi-
cal Matter.” Qualitative Sociology 15(3): 247–271.
———. 2004. “Collaborative Turn Sequences.” In Conversation Analysis: Studies
from the First Generation, edited by Gene H. Lerner, 225–256. Amsterdam: John
Benjamins.
Lerner, Gene H., and Celia Kitzinger. 2007. “Extraction and Aggregation in the Repair
of Individual and Collective Self-reference.” Discourse Studies 9(4): 526–557.
Levelt, Willem. J.M. 1983. “Monitoring and Self-repair in Speech.” Cognition 14: 41–104.
———. 1989. Speaking: From Intention to Articulation. Cambridge, ma: mit Press.
Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press.
———. 2000. Presumptive Meanings: The Theory of Generalized Conversational Impli-
cature. Cambridge, ma: mit Press.
———. 2003. Space in Language and Cognition: Explorations in Cognitive Diversity.
———. 2004. “Deixis.” In The Handbook of Pragmatics, edited by Laurence Horn and
Gergory Ward, 97–121. Oxford: Blackwell.
———. 2006. “Cognition at the Heart of Human Interaction.” Discourse Studies 8(1):
85–93.
———. 2007. “Optimizing Person Reference: Perspectives from Usage on Rossel
Island.” In Person Reference in Interaction: Linguistic, Cultural and Social Perspec-
tives, edited by N.J. Enfield and Tanya Stivers, 29–72. Cambridge: Cambridge Univer-
sity Press.
Lindström, Anna, and Marja-Leena Sorjonen. 2012. “Affiliation in Conversation.” In The
Handbook of Conversation Analysis, edited by Jack Sidnell and Tanya Stivers, 350–
369. Boston: Wiley-Blackwell.
Livia, Anna. 1997. “Disloyal to Masculinity: Linguistic Gender and Liminal Identity in
French.” In Queerly Phrased: Language, Gender and Sexuality, edited by Anna Livia
and Kira Hall, 349–368. Oxford/New York: Oxford University Press.
———. 2001. Pronoun Envy: Literary Uses of Linguistic Gender. Oxford/New York:
Oxford University Press.
Livia, Anna, and Kira Hall, eds. 1997. Queerly Phrased: Language, Gender and Sexuality.
Oxford/New York: Oxford University Press.
Lucy, John A. 1992. Language Diversity and Thought: A Reformulation of the Linguistic
Relativity Hypothesis. Cambridge: Cambridge University Press.
———. 1996. “The Scope of Linguistic Relativity: An Analysis and Review of Empirical
Research.” In Rethinking Linguistic Relativity, edited by John J. Gumperz and
Stephen C. Levinson, 37–69. Cambridge: Cambridge University Press.
Lykke, Nina. 2004. “Women’s/Gender/Feminist Studies: A Post-Disciplinary Disci-
pline?” In The Making of European Women’s Studies, Vol. V, edited by Rosi Braidotti,
Edyta Just and Marlise Mensink, 91–101. Utrecht: ATHENA.
References 187
———. 2010. Feminist Studies: A Guide to Intersectional Theory, Methodology and Writ-
ing. New York: Routledge.
Lykke, Nina, and Rosi Braidotti, eds. 1996. Between Monsters Goddesses and Cyborgs:
Feminist Confrontations with Science, Medicine and Cyberspace. London: zed.
Lyotard, Jean-François. 1984. The Postmodern Condition: A Report on Knowledge. Trans-
lated by Geoff Bennington and Brian Massumi, foreword by Fredric Jameson. Man-
chester: Manchester University Press.
Mackridge, Peter. 1985. The Modern Greek Language: A Descriptive Analysis of Standard
Modern Greek. Oxford: Oxford University Press.
Makri-Tsilipakou, Μarianthi. 1984. “Απόπειρα Περιγραφής της Νεοελληνικής Προσφώ‑
νησης” [An Αttempt to Describe the Forms of Address in Modern Greek]. Studies in
Greek Linguistics 4: 219–239.
———. 1989. “The Gender of Άνθρωπος: An Exercise in False Generics.” In Proceedings
of the 3rd Symposium on the Description and/or Comparison of English and Greek,
Aristotle University of Thessaloniki, School of English, Department of Theoretical
and Applied Linguistics, 61–83. Thessaloniki: University Studio Press.
Malone, Martin J. 1997. Worlds of Talk: The Presentation of Self in Everyday Conversation.
Cambridge: Polity Press.
Maltz, Daniel N., and Ruth A. Borker. 1982. “A Cultural Approach to Male/Female Mis-
communication.” In Language and Social Identity, edited by John J. Gumperz, 195–
216. Cambridge: Cambridge University Press.
Marcato, Gianna, and Eva-Maria Thüne. 2002. “Gender and Female Visibility in Italian.”
In Gender across Languages: The Linguistic Representation of Women and Men,
edited by Marlis Hellinger and Hadumod Bussmann, 186–217. Amsterdam: John
Benjamins.
Martyna, Wendy. 1983. “Beyond the He/Man Approach: The Case for Nonsexist Lan-
guage.” In Language, Gender and Society, edited by Barrie Thorne, Cheris Kramarae
and Nancy Henley, 25–37. Rowley, ma: Newbury House.
Mathiot, Madeleine. 1979. “Sex Roles as Revealed through Referential Gender in Ame
rican English.” In Ethnolinguistics: Boas, Sapir and Whorf Revisited, edited by
Madeleine Mathiot, 1–47. The Hague/Paris/New York: Mouton Publishers.
Mayer, Mercer. 1969. Frog, Where are You? New York: Dial Press.
McConnell-Ginet, Sally. 2003. “What’s in a Name?: Social Labeling and Gender Prac-
tices.” In The Handbook of Language and Gender, edited by Janet Holmes and
Miriam Meyerhoff, 69–97. Oxford: Blackwell.
———. 2011. “Language and Gender.” In Gender, Sexuality, and Meaning: Linguistic
Practice and Politics, 37–62. New York: Oxford University Press. Originally published
in Frederick J. Newmeyer, ed., Linguistics: The Cambridge Survey. iv: Language: The
Sociocultural Context (Cambridge: Cambridge University Press, 1988).
188 references
———. 2014. “Gender and its Relation to Sex: The Myth of ‘Natural’ Gender”. In The
Expression of Gender, edited by Greville G. Corbett, 3–38. Berlin: De Gruyter
Mouton.
McElhinny, Bonnie. 2003. “Theorizing Gender in Sociolinguistics and Linguistic
Anthropology.” In The Handbook of Language and Gender, edited by Janet Holmes
and Miriam Meyerhoff, 21–42. Oxford: Blackwell.
McKay, Donald G., and David C. Fulkerson. 1979. “On the Comprehension and Produc-
tion of Pronouns.” Journal of Verbal Learning and Verbal Behavior 18(6): 661–673.
Miller, Casey, and Kate Swift. [1981] 1988. The Handbook of Non-Sexist Writing for Writ-
ers, Editors and Speakers. The Women’s Press.
Μills, Anne E. 1986. The Acquisition of Gender: A Study of English and German. Berlin:
Springer-Verlag.
Mühlhäusler, Peter, and Rom Harré. 1990. Pronouns and People: The Linguistic Construc-
tion of Social and Personal Identity. Oxford: Blackwell.
Niemeier, Susanne, and René Dirven, eds. 2000. Evidence for Linguistic Relativity.
Amsterdam/Philadelphia: John Benjamins.
Nilsen, Alleen P. 1977. “Sexism as Shown through the English Vocabulary.” In Sexism
and Language, edited by Alleen P. Nilsen, Haig Bosmajian, H. Lee Gershuny and
Julia P. Stanley, 27–41. Urbana, Ill.: National Council of Teachers of English, U.S.A.
Nissen, Uwe K. 1997. “Do Sex-neutral and Sex-specific Nouns Exist: The Way to Non-
sexist Spanish?” In Kommunikation von Geschlecht [Communication of Gender],
edited by Friederike Braun and Ursula Pasero, 222–241. Pfaffenweiler: Centaurus.
———. 2002. “Gender in Spanish: Tradition and Innovation.” In Gender across Lan-
guages: The Linguistic Representation of Women and Men, edited by Marlis Hellinger
and Hadumod Bussmann, 251–279. Amsterdam: John Benjamins.
Ochs, Elinor. 1992. “Indexing Gender.” In Rethinking Context: Language as an Interactive
Phenomenon, edited by Alessandro Duranti and Charles Goodwin, 335–358. Cam-
Pastre, Geneviène. 1997. “Linguistic Gender Play among French Gays and Lesbians.” In
Queerly Phrased: Language, Gender and Sexuality, edited by Anna Livia and Kira
Hall, 369–379. Oxford/New York: Oxford University Press.
Pavlidou, Theodossia-Soula. 1985. “Παρατηρήσεις στα Θηλυκά Επαγγελματικά” [Remarks
on Feminine Occupational Terms]. Studies in Greek Linguistics 5: 201–217.
———. 2000. “Modern Greek Gender.” In The Making of European Women’s Studies: A
Work in Progress Report on Curriculum Development and Related Issues, Vol. I, edited
by Rosi Braidotti and Esther Vonk, 42–43. ATHENA/Utrecht University: Utrecht.
———. 2003. “Women, Gender and Modern Greek.” In Gender across Languages: The
Linguistic Representation of Women and Men, edited by Marlis Hellinger and Hadu-
mod Bussmann, 175–199. Amsterdam: John Benjamins.
References 189
———. [2002] 2006a. “Γλώσσα-Γένος-Φύλο: Προβλήματα, Aναζητήσεις και Ελληνική

Γλώσσα” [Language-Grammatical Gender-Social Gender: Problems, Queries and
the Greek Language]. In Language-Grammatical Gender-Social Gender, edited by
Theodossia-Soula Pavlidou, 15–64. Thessaloniki: Institute of Modern Greek
Studies.
———. 2006b. “The Lexical Representation of Women and Men in Modern Greek.”
Paper presented at the 4th International Gender and Language Association Confer-
ence, Valencia, Spain, November 8–10.
———. 2008. “Εμείς και η Συνομιλιακή Συγκρότηση (Έμφυλων) Συλλογικοτήτων” [We
and the Construction of (Gendered) Collectivities]. In Light and Warmth: In Mem-
ory of A.-Ph. Christidis, edited by Maria Theodoropoulou, 437–453. Thessaloniki:
Center of the Greek Language.
———. 2011. “Gender and Interaction.” Ιn The Sage Ηandbook of Sociolinguistics, edited
by Ruth Wodak, Barbara Johnstone and Paul Kerswill, 412–427. London: Sage.
———. 2012a. “Collective Aspects of Subjectivity: The Subject Pronoun Eμείς (‘We’) in
Modern Greek.” In Subjectivity in Discourse, edited by Nicole Baumgarten, Inke Du
Bois and Juliane House, 33–65. Bingley, uk: Emerald.
———. 2012b. “The Corpus of Spoken Greek: Goals, Challenges, Perspectives.” lrec
Proceedings, Workshop 18 (Best Practices for Speech Corpora in Linguistic Research),
23–28.
Pavlidou, Theodossia-Soula, and Angeliki Alvanoudi. 2013. “Grammatical Gender and
Cognition.” In Major Trends in Theoretical and Applied Linguistics: Selected Papers
from the 20th istal, Vol. 2, 109–123. London: Versita de Gruyter.
Pavlidou, Theodossia-Soula, Angeliki Alvanoudi, and Eleni Karafoti. 2004. “Γραμματικό
Γένος και Σημασιακό Περιεχόμενο: Προκαταρκτικές Παρατηρήσεις για τη Λεξιλογική
Αναπαράσταση των Φύλων” [Grammatical Gender and Semantic Content: Prelimi-
nary Remarks on the Lexical Representation of Social Gender]. Studies in Greek Lin-
guistics 24: 543–553.
Pauwels, Αnne. 2003. “Linguistic Sexism and Feminist Linguistic Activism.” In The
Handbook of Language and Gender, edited by Janet Holmes and Miriam Meyerhoff,
550–570. Oxford: Blackwell.
Pawley, Andrew. 2002. “Using He and She for Inanimate Referents in English: Questions
of Grammar and World View.” In Ethnosyntax: Explorations in Grammar and Cul-
ture, edited by N.J. Enfield, 110–137. Oxford: Oxford University Press.
Peirce, Charles S. 1940. The Philosophy of Peirce: Selected Writings. Edited by Justus
Buchler. London: Routledge and Kegan Paul.
Penelope, Julia. 1990. Speaking Freely: Unlearning the Lies of the Fathers’ Tongues. New
York: Pergamon Press.
Pomerantz, Anita. 1984. “Agreeing and Disagreeing with Assessments: Some Features
of Preferred/Dispreferred Turn Shapes.” In Structures of Social Action: Studies in
190 references
Conversation Analysis, edited by J. Maxwell Atkinson and John Heritage, 57–101.

Potter, Jonathan, and Hedwig te Molder. 2005. “Talking Cognition: Mapping and Mak-
ing the Terrain.” In Conversation and Cognition, edited by Hedwig te Molder and
Jonathan Potter, 1–54. Cambridge: Cambridge University Press.
Pütz, Martin, and Marjolijn Verspoor, eds. 2000. Explorations in Linguistic Relativity.
Amsterdam/Philadelphia: John Benjamins.
Radden, Günter, and Zoltán Kövesces. 1999. “Towards a Theory of Metonymy.” In
Metonymy in Language and Thought, edited by Klaus-Uwe Panther and Günter Rad-
den, 17–59. Amsterdam/Philadelphia: John Benjamins.
Raymond, Geoffrey, and John Heritage. 2006. “The Epistemics of Social Relations:
Owning Grandchildren.” Language in Society 35(5): 677–705.
Romaine, Suzanne. 1999. Communicating Gender. Mahwah, N.J.: Erlbaum
Rosch, Εleanor. 1978. “Principles of Categorization.” In Cognition and Categorization,
edited by Eleanor Rosch and Barbara B. Lloyd, 27–48. Hillsdale, nj: Lawrence
Erlbaum.
Ruusuvuori, Johanna. 2012. “Emotion, Affect and Conversation.” In The Handbook of
Conversation Analysis, edited by Jack Sidnell and Tanya Stivers, 330–349. Boston:
Wiley-Blackwell.
Saalbach, Henrik, Mutsumi Imai, and Lennart Schalk. 2012. “Grammatical Gender and
Inferences about Biological Properties in German-speaking Children.” Cognitive Sci-
ence 36: 1251–1267.
Sacks, Harvey. 1972a. “An Initial Investigation of the Usability of Conversational Data
for doing Sociology.” In Studies in Social Interaction, edited by David Sudnow, 31–74.
New York: Free Press.
———. 1972b. “On the Analyzability of Stories by Children.” In Directions in Sociolin-
guistics: The Ethnography of Communication, edited by John J. Gumperz and Dell H.
Hymes, 325–345. New York: Holt, Rinehart and Winston.
———. 1974. “An Analysis of the Course of a Joke’s Telling in Conversation.” In Explora-
tions in the Ethnography of Speaking, edited by Richard Bauman and Joel Sherzer,
337–353. Cambridge: Cambridge University Press.
———. 1984. “On Doing ‘Being Ordinary’.” In Structures of Social Action: Studies in Con-
versation Analysis, edited by J. Maxwell Atkinson and John Heritage, 413–429. Cam-
———. 1995. Lectures on Conversation. Vol. I and ii. Edited by Gail Jefferson. Oxford:
Blackwell.
Sacks, Harvey, and Emanuel A. Schegloff. 1979. “Two Preferences in the Organization of
Reference to Persons in Conversation and their Interaction.” In Everyday Language:
Studies in Ethnomethodology, edited by George Psathas, 15–21. New York: Irvington
Publishers, Inc.
References 191
Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. “A Simplest Systematics
for the Organization of Turn-taking for Conversation.” Language 50(4): 696–735.
Sapir, Edward. [1949] 1970. Selected Writings in Language, Culture and Personality.
Edited by David G. Mandelbaum. Berkley: University of California Press.
Schafroth, Elmar. 2003. “Gender in French.” In Gender across Languages: The Linguistic
Representation of Women and Men, edited by Marlis Hellinger and Hadumod Buss-
mann, 87–117. Amsterdam: John Benjamins.
Schegloff, Emanuel A. 1986. “The Routine as Achievement.” Human Studies 9: 111–151.
———. 1987. “Analyzing Single Episodes of Interaction: An Exercise in Conversation
Analysis.” Social Psychology Quarterly 50(2): 101–114.
———. 1988. “Description in the Social Sciences I: Talk in Interaction.” ipra Papers in
Pragmatics 2: 1–24.
———. 1991. “Conversation Analysis and Socially Shared Cognition.” In Perspectives
on Socially Shared Cognition, edited by Lauren B. Resnick, John M. Levine and
Stephanie D. Teasley, 150–171. Washigton, D.C: American Psychological Association.
———. 1996a. “Issues of Relevance for Discourse Analysis: Contingency in Action,
Interaction and Co-participant Context.” In Computational and Conversational Dis-
course: Burning Issues—An Interdisciplinary Account, edited by Eduard H. Hovy and
Donia R. Scott, 3–38. Heidelberg: Springer Verlag.
———. 1996b. “Some Practices for Referring to Persons in Talk-in-Interaction: A Partial
Sketch of a Systematics.” In Studies in Anaphora, edited by Barbara A. Fox, 437–85.
Amsterdam: John Benjamins.
———. 1997. “Whose Text? Whose Context?” Discourse and Society 8(2): 165–187.
———. 2004. Lectures on Repair and Categorization (244b/266 Conversational Struc-
tures ii), ucla.
———. 2006a. “On Possibles.” Discourse Studies 8(1): 141–157.
———. 2006b. “Interaction: The Infrastructure for Social Institutions, the Natural Eco-
logical Niche for Language, and the Arena in which Culture is Enacted.” In Roots of
Human Sociality: Culture, Cognition and Interaction, edited by N.J. Enfield and Ste-
phen C. Levinson, 70–96. Oxford/New York: Berg.
———. 2007a. Sequence Organization in Interaction. Cambridge: Cambridge Univer-
sity Press.
———. 2007b. “A Tutorial on Membership Categorization.” Journal of Pragmatics 39:
462–482.
———. 2007c. “Conveying Who You Are: The Presentation of Self, strictly Speaking.”
In Person Reference in Interaction: Linguistic, Cultural and Social Perspectives, edited
by N.J. Enfield and Tanya Stivers, 123–148. Cambridge: Cambridge University Press.
———. n.d. “Increments.” Manuscript, ucla.
192 references
———. 2013. “Ten Operations in Self-initiated Same-turn Repair.” In Conversational

Repair and Human Understanding, edited by Makoto Hayashi, Geoffrey Raymond,
and Jack Sidnell, 41–70. Cambridge: Cambridge University Press.
Schegloff, Emanuel A., and Harvey Sacks. 1973. “Opening up Closings.” Semiotica 8(4):
289–327.
Schegloff, Emanuel A., Gail Jefferson, and Harvey Sacks. 1977. “The Preference for Self-
correction in the Organization of Repair in Conversation.” Language 53(2):
361–382.
Schmid, Sonja. 1998. “Zur Bezeichnung Weiblicher Personen im Russischen: Eine
Empirische Pilotstudie.” Wiener Slawistischer Almanach 41: 239–262.
Schulz, Muriel R. 1975. “The Semantic Derogation of Woman.” In Language and Sex:
Difference and Dominance, edited by Barrie Thorne and Nancy Henley, 64–75. Row-
ley, ma: Newbury house.
Sera, Maria D., Christian A.H. Berge, and Javier del Castillio Pintando. 1994. “Grammati-
cal and Conceptual Forces in the Attribution of Gender by English and Spanish
Speakers.” Cognitive Development 9(3): 261–292.
Sera, Maria D., Chryle Elieff, James Forbes, Melissa Clark Burch, Wanda Rodríguez, and
Diane Poulin Dubois. 2002. “When Language Affects Cognition and when it Does
not: An Analysis of Grammatical Gender and Classification.” Journal of Experimen-
tal Psychology, General 131(3): 377–397.
Sidnell, Jack. 2010. Conversation Analysis: An Introduction. Wiley-Blackwell.
Sidnell, Jack, and N.J. Enfield. 2012. “Language Diversity and Social Action: A Third
Locus of Linguistic Relativity.” Current Anthropology 53(3): 302–333.
Silverstein, Michael. 1976. “Shifters, Linguistic Categories, and Cultural Description.” In
Meaning in Anthropology, edited by Keith H. Basso and Henry A. Selby, 11–55. Albu-
querque: University of New Mexico.
Slobin, Dan Isaac. 1987. “Thinking for Speaking.” Proceedings of the Berkley Linguistics
Society 13: 435–444.
———. 1996. “From ‘Thought and Language’ to ‘Thinking for Speaking’.” In Rethinking
Linguistic Relativity, edited by John J. Gumperz and Stephen C. Levinson, 70–96.
———. 2003. “Language and Thought Online: Cognitive Consequences of Linguistic
Relativity.” In Language in the Mind: Advances in the Study of Language and Thought,
edited by Dedre Gentner and Susan Goldin-Meadow, 157–191. Cambridge, ma: mit
Press.
Speer, Susan A. 2005. Gender Talk: Feminism, Discourse and Conversation Analysis.
London: Routledge.
Speer, Susan A., and Elizabeth Stokoe. 2011. “An Introduction to Conversation and Gen-
der.” In Conversation and Gender, edited by Susan A. Speer and Elizabeth
Stokoe, 1–27. Cambridge: Cambridge University Press.
References 193
Spender, Dale. 1980. Man Made Language. London: Routledge and Kegan Paul.
Stanley, Julia P. 1977. “Paradigmatic Woman: The Prostitute.” In Papers in Language
Variation, edited by David L. Shores and Carole P. Hines, 303–321. Tuscaloosa: The
University of Alabama Press.
Stivers, Tanya, and Jack Sidnell. 2012. “Introduction.” In The Handbook of Conversation
Analysis, edited by Jack Sidnell and Tanya Stivers, 1–8. Boston: Wiley-Blackwell.
Stivers, Τanya, N.J. Enfield, and Stephen C. Levinson. 2007. “Person Reference in Inter-
action.” In Person Reference in Interaction: Linguistic, Cultural and Social Perspec-
tives, edited by N.J. Enfield and Tanya Stivers, 1–20. Cambridge: Cambridge University
Press.
Stockill, Clare, and Celia Κitzinger. 2007. “Gendered ‘People’: How linguistically Non-
Gendered Terms Can Have Gendered Interactional Relevance.” Feminism Psychol-
ogy 17(2): 224–236.
Tannen, Deborah. 1990. You just Don’t Understand: Women and Men in Conversation.
New York: William Morrow.
Taylor, John R. 2003. Linguistic Categorization. New York: Oxford University Press.
Thorne, Barrie, and Nancy Henley, eds. 1975. “Difference and Dominance: An Overview
of Language, Gender, and Society.” In Language and Sex: Difference and Dominance,
5–42. Rowley, ma: Newbury House.
Τopsakal, Maria. 1995. The Influence of the Grammatical Gender on the Perception of
Gender in Modern Greek: Comparison with German and English. ma Thesis, Aristotle
University Thessaloniki.
Triantafyllidis, Manolis. [1941] 2005. Νεοελληνική Γραμματική (της Δημοτικής) [Modern
Greek Grammar (of the Demotic)]. Reprint with corrections of the 1941 edition.
Thessaloniki: Institute of Modern Greek Studies.
———. [1953] 1963. “H ‘Βουλευτίνα’ και ο Σχηματισμός των Θηλυκών Επαγγελματικών
Ουσιαστικών” [The ‘Parliament-member-fem’ and the Formation of Feminine
Occupational Terms]. In Άπαντα Μανόλη Τριανταφυλλίδη [Collected Works of Mano-
lis Triantafyllidis], Vol. 2, 326–334. Thessaloniki: Institute of Modern Greek Studies.
Tsangalidis, Anastasios. 2011. Greek. In The Languages and Linguistics of Europe: A Com-
prehensive Guide, edited by Bernd Kortmann and Johan van der Auwera, 223–239.
Berlin: De Gruyter Mouton.
Tsokalidou, Roula. 1996. Το Φύλο της Γλώσσας: Οδηγός μη Σεξιστικής Γλώσσας για τον
Δημόσιο Ελληνικό Λόγο [The Gender of Language: A Guide to Non-sexist Language for
Public Greek Speech]. Athens: Syndesmos Ellinidon Epistimonon.
van Dijk, Teun A. 2009. “Critical Discourse Studies: A Sociocognitive Approach.” In
Methods of Critical Discourse Studies, edited by Ruth Wodak and Michael Meyer,
62–85. London: Sage.
Weber, David John. 1989. A Grammar of Huallaga (Huánuco) Quechua. Berkley/Los
Angeles: University of California Press.
194 references
Whorf, Benjamin L. 1956. Language, Thought and Reality: Selected Writings of Benjamin
Lee Whorf. Edited by John B. Carroll. Cambridge, ma: mit Press.
Widerberg, Karin. 1998. “Translating Gender.” Nora 2(6): 133–138.
Wierzbicka, Anna. 2002. “Sexism in Grammar: The Semantics of Gender in Australian
English.” Anthropological Linguistics 44(2): 143–177.
Wilkinson, Sue. 2010. “Gender, Routinization and Recipient Design.” In Conversation
and Gender, edited by Susan A. Speer and Elizabeth Stokoe, 112–134. Cambridge:
Wittig, Monique. 1992. The Straight Mind and Other Essays. Boston: Beacon Press.
Zubin, David A., and Klaus-Michael Köpcke. 1981. “Gender: A Less than Arbitrary
Grammatical Category.” In Papers from the Seventh Regional Meeting, Chicago Lin-
guistic Society. April, 16–18. Chicago: The Society.
———. 1984. “Affect Classification in the German Gender System.” Lingua 63: 41–96.
———. 1986. “Gender and Folk Taxonomy: The Indexical Relation between Grammati-
cal and Lexical Categorization.” In Νoun Classes and Categorization: Symposium on
Categorization and Noun Classification: Papers (Typological Studies in Language
Vol. 7), 139–180. Amsterdam: John Benjamins.
Index of Authors
Aikhenvald, Alexandra 1, 14–18, 27–28, 31, Croft, William 7, 53

43, 63, 135 Cruse, D. Alan 7, 53
Alpher, Barry 27 Curzan, Anne 43
Althusser, Louis 49–50
Alvanoudi, Angeliki 21, 41, 44, 46, 67–69, Derrida, Jacques 49–50
73, 95 Dirven, René 62
Anastasiadi-Symeonidi, Anna 22, 26 Dixon, R.M.W. 15, 18, 27–28, 64, 79
Andermahr, Sonya 33, 47 Doleschal, Ursula 6, 66
Asher, R.E. 28 Drew, Paul 131
Atkinson, J. Maxwell 73, 85
Austin, John L. 49–50, 171 Eakins, Barbara W. 4, 43
Austin, Peter 15 Eakins, Rollin G. 4, 43
Eckert, Penelope 5, 8, 33, 36–37, 128
Bal, Mieke 171 Enfield, N.J. 7, 10, 62, 75–77, 81–82, 86–87,
Barcelona, Antonio 54 90–91
Beauvoir, Simone de 34 Engelberg, Μila 40–41, 44, 66–67
Bergvall, Victoria L. 33 Ervin, Susan M. 68
Berman, Ruth A. 63 Evans, Vyvyan 52–53
Bickel, Balthasar 64–65
Bing, Janet M. 33 Fausto-Sterling, Anne 2
Blaszkowska, Hanka 43 Ferguson, Charles A. 13
Boas, Franz 59–60, 63 Flaherty, Mary 68
Borba, Rodrigo 8, 38–39 Fodor, Jerry A. 52
Borker, Ruth A. 4 Foley, William A. 17, 62
Boroditsky, Lera 6, 68–69 Foucault, Michel 49, 169
Bowerman, Μelissa 62 Frank, Francine W. 44
Braidotti, Rosi 2–3, 33–36, 47 Freed, Alice F. 4, 33
Braun, Friederike 66–67 Fulkerson, David C. 66–67
Brown, Penelope 10, 76–78
Bucholtz, Mary 38 Garfinkel, Harold 10, 84, 90
Bussmann, Hadumod 1–2, 8, 14, 16, 18, 27, Gentner, Dedre 62
37, 42–43, 66–67, 72, 168, 173 Gleitman, Lila 52
Butler, Judith 5, 9, 33–35, 46, 49–50, 171–172 Goffman, Erving 84, 135
Goldin-Meadow, Susan 62
Chila-Markopoulou, Despina 22, 25–26, 29 Graham, Alma 4, 43
Chomsky, Noam 52 Green, Melanie 52–53
Christidis, Anastasios Ph. 13 Grice, H. Paul 82
Clark, Herbert H. 84 Grönberg, Anna G. 43
Clarke, Mark A. 68 Gumperz, John J. 59, 62, 65
Claudi, Ulrike 27
Clift, Rebecca 135 Hall, Kira 8, 33, 38
Connell, Raewyn 40 Hanks, William F. 10, 64, 73, 75, 77–78
Corbett, Greville, G. 1, 14–15, 17, 27–28 Haraway, Donna 34, 36
196 index of authors
Harding, Sandra 32–33 Levelt, Willem J.M. 64, 132

Harré, Rom 95 Levinson, Stephen C. 59, 62, 64–65, 75–76,
Haviland, John B. 10, 77–79 79–83, 87
Hayashi, Makoto 129 Lindström, Anna 118
Heine, Bernd 17 Livia, Anna 8, 33, 38–40, 174
Hellinger, Marlis 1–2, 8–9, 14, 16, 18, 27, 37, Lovell, Terry 33, 47
42–43, 66–67, 72, 168, 173 Lucy, John A. 6–7, 59, 61
Henley, Nancy 4, 174 Lykke, Nina 34–36, 171
Heritage, John 73, 85–86, 96, 147 Lyotard, Jean-François 47
Hill, Jane H. 62
Holmes, Janet 5 Mackridge, Peter 13, 22–23
Holt, Elizabeth 135 Makri-Tsilipakou, Μarianthi 44, 67, 78
Holton, David 13, 22–23 Malone, Martin J. 104
Horrocks, Geoffrey C. 13 Maltz, Daniel N. 4
Hougaard, Anders 87 Mannheim, Bruce 62
Huang, Yan 30, 75 Marcato, Gianna 17, 42
Hymes, Dell H. 61 Martyna, Wendy 43
Mathiot, Madeleine 42
Imai, Mutsumi 68, 70 Mayer, Mercer 63
Irigaray, Luce 9, 36, 46–48 McConnell-Ginet, Sally 2, 5, 8, 33, 36–38, 128
Ιrmen, Lisa 66 McElhinny, Bonnie 37
McKay, Donald G. 66
Jefferson, Gail 86, 93–94, 119, 129–132, 142, Meyerhoff, Miriam 5
144, 148, 154 Miller, Casey 4, 43–44
Joseph, Brian 13 Μills, Anne E. 68
Mühlhäusler, Peter 95
Karafoti, Eleni 21, 41, 44
Kati, Dimitra 52 Niemeier, Susanne 62
Kitagawa, Chisato 104 Nikiforidou, Kiki 52
Kitzinger, Celia 5, 90–91, 95–96, 98, 104, Nilsen, Alleen P. 43–44
129–131, 169 Nissen, Uwe K. 6, 14, 40–41, 66
Klein, Josef 66
Köhncke, Astrid 66 Ochs, Elinor 8, 10, 36–37, 79
Kondyli, Marianna 52 O’Donovan, Veronica 8, 38
Konishi, Toshi 68 Ogden, Richard 131
Koniuszaniec, Gabriela 43 Ostermann, Ana C. 8, 38–39
Köpcke, Klaus-Michael 9, 16, 55–56, 71
Kövesces, Zoltán 54 Papafragou, Anna 52
Kramarae, Cheris 44 Pastre, Geneviène 39
Kulick, Don 8, 38–39 Pavlidou, Theodossia-Soula 1, 3, 5, 21, 40–41,
43–46, 66–69, 73, 92, 98, 123, 128
Lakoff, George 9, 52–55, 71, 173 Panther, Klaus-Uwe 16
Lakoff, Robin 4, 32, 40, 173 Pauwels, Αnne 44, 174
Langacker, Ronald W. 54 Pawley, Andrew 42
Latour, Bruno 1, 171 Peirce, Charles S. 7
Lee, Penny 62 Penelope, Julia 4
Lehrer, Adrienne 104 Philippaki-Warburton, Irene 13
Lerner, Gene H. 95, 98, 101, 146–147 Phillips, Webb 6, 68–69
index of authors 197
Pomerantz, Anita 94 Stokoe, Elizabeth 5, 90

Potter, Jonathan 87 Swift, Kate 4, 43–44
Pütz, Martin 62 Stanley, Julia P. 41
Stivers, Tanya 77, 81, 84
Radden, Günter 54 Stockill, Clare 90
Raymond, Geoffrey 129, 147
Romaine, Suzanne 173 Tannen, Deborah 4
Rosch, Εleanor 53 Taylor, John R. 53
Ruusuvuori, Johanna 118 te Molder, Hedwig 87
Thorne, Barrie 4, 174
Saalbach, Henrik 68 Thüne, Eva-Maria 17, 42
Sacks, Harvey 10, 81, 85–86, 88–89, 93–94, Τopsakal, Maria 68
118, 129–130, 136, 146, 169 Treichler, Paula A. 44
Sapir, Edward 8, 59–60 Triantafyllidis, Manolis 19, 22, 41
Schafroth, Elmar 43 Tsangalidis, Anastasios 13
Schalk, Lennart 68 Tsokalidou, Roula 44
Schegloff, Emanuel A. 10, 73–74, 78, 81,
84–91, 93–94, 97, 98, 100, 104, 107, 111, van Dijk, Teun A. 84
113–114, 118, 129–131, 133, 136, 140, 148, 154 Verspoor, Marjolijn 62
Schmid, Sonja 6, 66
Schmidt, Lauren A. 6, 68–69 Walker, Traci 131
Schulz, Muriel R. 40–41 Weber, David John 64
Sera, Maria D. 6, 38, 68–69 Whorf, Benjamin L. 8, 59–61, 168, 170
Sidnell, Jack 62, 84, 86–87, 129–131, 143 Widerberg, Karin 3
Silverstein, Michael 10, 79–80 Wierzbicka, Anna 42, 68
Slobin, Dan Isaac 8–10, 52, 62–65, 71–73, 87, Wilkinson, Sue 94
166, 174 Wittig, Monique 9, 46, 48–49
Sorjonen, Marja-Leena 118 Wolkowitz, Carol 33, 47
Speer, Susan A. 5, 90
Spender, Dale 4, 32, 44, 173 Zubin, David A. 9, 16, 55–56, 71
Index of Subjects
Agreement 17–18 Orientation 84, 90
Assignment 15–17
Assumption 81. See also Performativity 33, 49–50, 171–174
Presupposition 79–80 performative 35
Presupposition 79–80
Categorization 52–55 Principles, assignment
Category-bound activity 89 morphological 15, 17
Conversation analysis 84–88 phonological 15, 17
Correction 131–132 semantic 15–17
Gender Recipient-design 94
feminine 1–2, 14, 16 Reference
generic use, masculine 4, 27, 42 collective self-reference 95
grammatical 1–2, 13–31 inanimate 15–17, 68–70
hierarchy 40–44 indefinite second person 104
identity 37–40 individual self-reference 95
lexical 2 person 16, 74–79
masculine 1–2, 14, 16 recipient 106
natural 2 third person 111
neuter 1–2, 14, 16 Relativity
Relevance 116 linguistic 59–61
sociocultural 32–36 structural 61–62
Resolution 28–30
Idealized cognitive models 54
Index 36–37, 77–81 Self-repair 129–131
indexicality 75, 79–80 Sex 2, 34–36
Inference 58, 81–83 Sexism 4, 40, 169–171
Interaction 72–74
Thinking for speaking 62–65, 166–167
Membership category 88–90
Metonymy 54–57
Next-turn proof procedure 86

Non-match 123
Noun class 13–15

Grammatical Gender in Interaction - Cultural and Cognitive Aspects PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Grammatical Gender in Interaction - Cultural and Cognitive Aspects PDF

Загружено:

Авторское право:

Доступные форматы

Brill’s Studies in Language,

Alexandra Y. Aikhenvald (Cairns Institute, James Cook University)

The titles published in this series are listed at brill.com/bslc

Library of Congress Cataloging-in-Publication Data

Copyright 2015 by Koninklijke Brill NV, Leiden, The Netherlands.

This book is printed on acid-free paper.

3 Grammatical Gender and the Intersection of Sociocultural

4 Grammatical Gender and Speakers’ Cognition 52

4.3 Indications of the Cognitive Role of Grammatical Gender:

5 Grammatical Gender, Culture and Cognition in Interaction 72

6 The Empirical Investigation of Grammatical Gender in

7 What Can We Conclude? 166

7.3 Reconsidering Gender Performativity 171

The study presented in this book is based on my PhD dissertation at the

In this book, I use broad transcriptions based on the International Phonetic

c voiceless palatal stop

I follow the transcription conventions used in the Corpus of Spoken Greek of

1 Temporal and Sequential Relationships

[ left brackets: point of overlap onset between two or more

] right brackets: point of overlap end between two or more

A pair of equals signs is used to indicate the following:

The single equals sign is used to indicate latching between two

2 Symbols and Combinations of Symbols for Representing Various

punctuation indication of intonation, more specifically,

h Hearable aspiration is shown with the Latin letter h. Its

((laughs)) Double parentheses and italics are used to mark meta-

1.1 Contextualizing the Problem

© koninklijke brill nv, leiden, ���5 | doi ��.��63/9789004283152_002

[vroçí] ‘rain’), masculine (ωκεανός [οceanós] ‘ocean’) or neuter (δέντρο [ðéndro]

1 However, gender-indefinite nouns in English can be pronominalized by male-specific or

1.2 On Grammatical Gender and Culture

How do grammars, mental representations of linguistic systems, connect

These questions have undoubtedly inspired my work.

1.3 On Grammatical Gender and Cognition

The assumption that language plays a role in speakers’ understanding of the

while thoughts or private events independent of semiotic/linguistic

Semiotic relativity serves as a background assumption in my research, as a sort

relation between grammatical gender and conceptualization is further sup-

1.4 Summary of the Argument to be Presented in the Book

the hierarchical gender dichotomy presupposes the role of grammatical gen-

This chapter examines the grammatical gender system in Standard Modern

2.2 Noun Classes or Genders

Grammatical gender is a device of noun categorization, which forms part

© koninklijke brill nv, leiden, ���5 | doi ��.��63/9789004283152_003

(1) Masculine Feminine

(2) Masculine Feminine Neuter

Noun class systems have two definitional properties, according to

a) Some constituent outside the noun itself, such as adjectives, numbers,

(3) Masculine Feminine

(4) Masculine Feminine

Contrary to person reference, gender assignment in inanimate reference is

(5) la ragazza alta

(6) der Platz an der Sonne

However, in nouns denoting human beings semantic agreement may also

2.3 Grammatical Gender in Modern Greek

In Modern Greek, the grammatical gender system is divided into three

(7) Masculine Feminine Neuter

table 1 Declension I-Masculine

nom o ουρανός οι ουρανοί

table 2 Declension II-Feminine

nom η θάλασσα οι θάλασσες

table 3 Declension iii-Neuter

© koninklijke brill nv, leiden, ��5 | doi ��.��63/9789004283152_002

1 However, gender-indefinite nouns in English can be pronominalized by male-specific or

© koninklijke brill nv, leiden, ��5 | doi ��.��63/9789004283152_003

7 Reference-tracking systems also include switch-reference systems, switch-function systems,

© koninklijke brill nv, leiden, ��5 | doi ��.��63/9789004283152_004

2 Women’s and Gender Studies is defined by Lykke (2004) as a post-disciplinary discipline

6 Identities should be understood here as discursive constructs emerging in interaction

9 Feminist postmodernism is an epistemological strand emerging from the critical