The origin of sexism in language

g&l (print) issn 1747–6321 g&l (online) issn 1747–633x Article The origin of sexism in language Ann Coady Abstract Although previous work on sexist linguistic structures has identified the causes of sexism in language as stemming from an androcentric world view, it has not described the social and semiotic processes involved in the historic production and reproduction of this kind of linguistic sexism. This article uses the three processes of iconisation, fractal recursivity, and erasure to bring together what appear to be disparate phenomena (such as the masculine generic, and even the very existence of the feminine grammatical gender) into a unifying theory. Iconisation results in the binary division of humanity into females and males; fractal recursivity explains how this division was projected onto language; and erasure demonstrates how certain discourses have been ignored, to the profit of others. A Queer critique of the two concepts of binarity and markedness (which arise as a result of iconisation) opens up exciting new ways to approach sexism in language, and to revitalise research in this area. keywords: sexist language; language ideologies; queer linguistics; grammatical gender Introduction In October 2014 during a parliamentary debate in the French lower chamber, right-wing representative Julien Aubert (UMP) addressed the left-wing representative, Sandrine Mazetier (PS1), as ‘Mme le président’ in the masculine. After stopping him, and asking him to refer to her as ‘Madame la présidente’, Aubert refused, claiming that he was simply following the standard rules of Affiliation Sheffield Hallam University, UK. email: anncoady@gmail.com g&l vol 12.3 2018 271–293 ©2018, equinox publishing https://doi.org/10.1558/genl.31445 272 Ann CoAdy French, and that ‘Madame la présidente’ was the wife of a president, not a president in her own right (Le Monde 2014). Aubert was reprimanded and fined a quarter of his monthly parliamentary allowance, and a media debate ensued (Fracchiolla 2014). Some job titles in French are difficult to feminise (e.g. docteur / médecin [doctor]), and others have negative connotations in the feminine (un entrâineur [trainer / coach] vs une entraîneuse [a woman employed in a night club whose job is to encourage clients to drink]. As for gender agreement rules, the masculine takes precedence when both masculine and feminine nouns share the same satellite elements. Even in a sentence like un hommemasc et cinq milliards de femmesfem sont mortsmasc [one man and five billion women died], where women clearly outnumber men, the past participle of the verb to die is traditionally in the masculine form morts not the feminine mortes. These examples serve to illustrate that sexist linguistic structures, such as the masculine unmarked generic form, are still an important problem in languages with grammatical gender. This article demonstrates how three linguistic processes used in the field of language ideology (LI): iconisation, fractal recursivity, and erasure (Irvine and Gal 2000) can be usefully combined to examine the origin of sexism in language. In this article ‘sexism in language’ refers to structural linguistic phenomena like the unmarked masculine, and gender agreement rules. The three concepts of iconisation, fractal recursivity, and erasure permit us to bring together previous work on sexism in language into one unifying theory, and provide ‘a fine-grained discourse analytical apparatus that allows us to tease out how social boundaries and inequalities are enacted through an ideological matrix where representations of language intersect with images of age, gender, ethnicity, race, sexuality, etc.’ (Milani 2010:121). Debates surrounding language and grammatical gender go at least as far back as ancient Greeks (Corbeill 2008:75; Baron 1986:28), and although some work has been done on the origin of sexist language and/ or the origin of grammatical gender (Luraghi 2009b, 2011; Michard 1996; Violi 1987; McConnell-Ginet 1984), they have rarely discussed the processes involved in these phenomena. Indeed, Milani notes that work needs to be carried out on ‘how such boundaries and intersections may become conventionalized and naturalized’ (Milani 2010:121), and Blommaert argues that the ‘historical production and reproduction of language ideologies, needs to be filled in’ (Blommaert 1999:1). This article aims to fill this gap, by peeling back the historical layer of ideologies, and tracing the social mechanisms that have resulted in sexist linguistic structures today. Without this historical knowledge, it is impossible to fully assess current research findings (Cameron 1995:x), and the impact of language planning. The origin of sexism in lAnguAge 273 Although most of the examples in this article will be taken from French, the following analysis of the emergence of sexist language will be pertinent to most European languages. Whereas in English the generic function of the masculine was the central problem for Second Wave feminist linguistic reform, in French the debate has, until very recently, concentrated on the feminisation of job titles (Elmiger 2008:111), using the visibility principle rather than neutralisation. This has been criticised by Queer linguists as a political and epistemological cul-de-sac which results in entrenching gender difference (Chetcuti and Greco 2012:11). In fact, Motschenbacher argues that ‘every time speakers or writers use binarily gendered forms, they reinstantiate the discursive formation of the heteronormative system’ (Motschenbacher 2014:250). But speakers of grammatically gendered languages have very little choice regarding binarily gendered forms, for example French only has two genders – feminine and masculine, there is no neuter. Even in languages such as German, which has a neuter gender, as opposed to an uter (or common) gender (Motschenbacher 2010:77), it is mostly used to refer to inanimate objects, and only very rarely for animate beings, and so is not a solution to neutralise gender. Research on sexist linguistic structures is often seen as ‘outdated and archaic’ (Mills 2008:9) as it seems to be incompatible with poststructuralist theories about the fluidity and performative nature of gender (Motschenbacher 2015:29); that is, whether words and grammar are understood to be sexist is highly context-dependent. As such, work on sexist linguistic structures has been marginalised in the English-dominated field of gender and language over the past two decades (Motschenbacher 2015:28). However, Queer linguistics opens up exciting new avenues for the study of sexism in language. Although most work on sexism in language has been done from a feminist perspective, this does not mean that it is incompatible with a Queer approach. Indeed, Mills (2008:6f ) notes that sexism needs to be reexamined in the light of Queer theory, and Jagose highlights ‘the difficulty, even the impossibility, of distinguishing decisively between feminist and queer critical traditions’ (Jagose 2009:172). Thus, the second aim of this article is to demonstrate how a Second Wave focus on sexism in language, can be fruitfully combined with Queer linguistics. Queer linguistics Traditionally, feminist linguistics has concentrated on highlighting linguistic inequalities between men and women (e.g. Second Wave-oriented work; Burr 2012; Houdebine 2003; Pauwels 1998; Spender 1980), and later sexist discourses (e.g. Third Wave work; Lazar 2014; Holmes 2006; Sunderland 2004). However, both Second and Third Wave feminism tend to take 274 Ann CoAdy gender binaries as given, whether biologically or socially constructed. By leaving the binary categories of ‘men’ and ‘women’ unquestioned, both of these waves leave the actual system of gender intact. Queer linguistics (QL) goes beyond feminist linguistic reform by questioning the very existence of gender categories. In this sense, QL clashes with some feminist linguistic reform, such as French féminisation, which promotes the visibility principle (policiersmasc et policièresfem) rather than neutralisation (police officers), thus reinforcing binary gender categories. It is perhaps easier to describe Queer as what it is not, rather than what it is. Queer is supposed to escape all attempts at definition (Motschenbacher 2010:6), and has been described as ‘a signifier without a signified’ (Saussure’s terminology), or a ‘floating’ or ‘empty’ signifier (Lévi-Strauss’s terminology), i.e. the word Queer (the signifier) is stable, but the concept it refers to (the signified) is not, as it is only defined in relation to current norms, and norms change. Because Queer challenges ‘whatever constitutes the normal, the legitimate, the generally accepted’ (Sicurella 2016:81), as those norms change over time, Queer relocates itself to retain its subversive force. It is, by definition, indeterminate and elastic, precisely the qualities which give it its political efficacy (Jagose 1996:1 cited in McConnell-Ginet 2002:138). Studies on language and gender carried out from a Queer perspective therefore challenge current gender and sexuality norms, which are produced in, and reinforced by, language, whether they are heteronormative (promoting heterosexuality as the norm) or heteronormative (promoting certain kinds of heterosexuality as the norm). In both of these cases the hetero element (from the Greek ἕτερος meaning ‘the other of two’, ‘other’ or ‘different’; OED undated) highlights the binary foundation of all forms of heteronormativity. Queer linguistics challenges how the language system promotes heteronormativity through ‘the linguistic construction of essentialist, binary gender categories’ (Motschenbacher 2014:250). Using QL we can critique binary linguistic categories, highlighting how they were constructed, based on social gender binaries, and examining how they function as a normative mechanism. Markedness, which is an essential concept in the sexist language debate, can also be unpacked to reveal how certain terms have become marked or unmarked. The concepts of binarity and markedness are of particular importance for the analysis of sexist linguistic structures, which I discuss under Iconisation below. Language ideology Language ideology (LI) as a field of study emerged from linguistic anthropology in the 1970s with the work of Silverstein (1979), but reflections on the relationship between language and ideology can also be found earlier The origin of sexism in lAnguAge 275 in the of work of Bakhtin and Voloshinov (Blommaert 2006). Although LI emerged from linguistic anthropology, work on attitudes to language have also been carried out in variationist sociolinguistics (Milroy and Milroy 2012) and applied linguistics (Cameron 1995). One of its main foci has been on attitudes in contexts of language contact (Jaffe 1999), and language standardisation (Johnson 2005), and as such has close ties with language reform. LI has also been used to analyse the ideological drive behind historical language change e.g. the disappearance of thou and thee in English, which were gradually abandoned because they had come to index an unpopular Quaker identity (Silverstein 1985:251). It is precisely the emphasis on language use as a politically invested ideological construct that brings LI and QL together. From an LI point of view, standardisation (including the normalisation of the masculine as the generic form) should be seen as the result of a discursive project, or ideological process (Woolard 1998:4). Not only do ideologies of language serve to rationalise language use, they can also ‘actively and concretely distort the linguistic structure it represents’ (Silverstein quoted in Woolard 1998:12). One example that Silverstein uses to illustrate this phenomenon is feminists’ ‘misanalysis’ of generic he: ‘the diagnosis of the purported structural ailment [that generic he is sexist] is really a process of unambiguous creation of – or infectious innoculation with – the pragmatic disease’ (Silverstein 1985:254). In other words, feminists have not understood the principles of structural gender categories i.e. that he is part of a formal structural hierarchy of language going from more inclusive to less inclusive e.g. masculine includes feminine but not vice versa, animate includes personal but not vice versa (Silverstein 1985:225–6). He claims that feminists have failed to differentiate between the masculine’s notional (inclusive) sense, and its exclusive (male) sense. I am sure that the metaphor of disease in this quote has not escaped the reader’s attention. This rather elitist comment can be understood in terms of what I have termed a ‘Tower of Babel’ ideology of language, in which language is apparently on a constantly downward slope (Deutscher 2006:ch. 3). Silverstein adds that generic he is a ‘structurally dictated indexical usage’ (Silverstein 1985:256), but he does not indicate how these constraints came about, i.e. the social and semiotic processes which resulted in masculine being at the top of this formal structural hierarchy, something which this article goes some way to explaining. Iconisation Iconisation is a dichotomising process whereby two groups of speakers are created according to linguistic features that they share, or are perceived to share. The linguistic feature becomes representative of one particu- 276 Ann CoAdy lar group. In other words, it becomes an icon of them. As well as being a dichotomising process, iconisation is also an essentialising process, in which individuals are treated as belonging to homogenous social groups and any intra group differences are minimised, whereas inter group differences are highlighted. In fact, the very existence of ‘a language’ is the result of iconisation. For instance, before the breakup of Yugoslavia in 1991, Serbo-Croatian was a single language with very minor differences. After the breakup, Serbo-Croatian fractured into four ‘different languages’: Serbian, Croatian, Bosnian and Montenegrin, along ethnic and religious lines, rather than linguistic, a division which tends to highlight differences between these four varieties, and make any similarities less visible. In fact, the ‘separate languages’ of Serbian, Croatian, Bosnian and Montenegrin are more similar to one another than British and American English, which are classed as two varieties of the same language (Thomas 2002:314). This process of iconisation in the Balkans is currently having direct consequences on non-sexist language reform: One reason for the rejection of feminist language reforms in Serbia is that neighbouring Croatia has accepted them, and Serbia has spent the better part of 25 years trying to create a separate national and political identify for itself (Rajilic 2016). Thus, iconisation can also describe perceived, rather than real differences. In order to analyse the emergence of sexism in language, I have reversed the process of iconisation, that is, rather than groups being partitioned because of (real or perceived) linguistic features, humans are partitioned on the basis of sex/gender. Humans themselves underwent a process of iconisation, and because of our binary conceptions of sex/gender, two groups were formed. Women and men were (and still are) essentialised as homogenous groups, thus ignoring any variation in biological sex and / or social gender. The fact that iconisation is an essentialising process is particularly interesting in relation to QL, whose central aim is, ‘de-essentialisation – a mechanism at the heart of Queer Linguistics’ (Motschenbacher and Stegu 2013:528). In other words, QL tries to reverse the process of iconisation in order to highlight its damaging social effects. One effect of essentialisation is the creation of binary categories. Binarity As a dichotomising and partitioning process, iconisation necessarily results in a binary. Gender binarity is seen as a form of normativity, which forces individuals to fall onto one side of the binary and marginalises those who do not. In fact, grammatical gender can be seen as the linguistic reflection and reinforcement of binary social gender. For Queer linguists the reason The origin of sexism in lAnguAge 277 that gender binarism exists, is to establish and stabilise a heteronormative system, in which men and women are supposed to be different from one another, in which opposites attract (Motschenbacher 2014:250), and inherent in which is a hierarchy between men and women. One example of how a binary gender system can work as a normative mechanism is when grammatical gender and referential gender clash. Concerning animate nouns, in most Indo-European languages, a man is usually referred to with a masculine noun and a woman with a feminine noun. There are a few exceptions like une vigie [a lookout], or une sentinelle [a sentry] in French, which are grammatically feminine, lexically neutral, but usually referentially male. Apart from these cases, when there is a clash between grammatical and referential gender, it is often to insult e.g. tapette, fiotte, pédale, tantouze and tarlouze [all meaning something like fag, pansy, or poofter] are all grammatically feminine nouns in French used to insult gay men. Studies in QL have found that insulting terms for gay men (Coutant 2014 for French) or ‘feminine men’ (Motschenbacher 2010:75–7 for German2) are often grammatically feminine. However, this trend is much weaker for women, for whom the terms tend to be grammatically feminine, not masculine (e.g. KampflesbeFEM [‘bulldyke’] in German. Coutant found no grammatically masculine terms for lesbian women in her corpus). This clearly shows the social hierarchy of man at the top reflected in grammatical gender with the masculine as more prestigious. Using the grammatically masculine form to insult a man involves no extra wounding potential because it is the most prestigious form, but using a grammatically feminine form to downgrade him to the status of woman is possible, as that is a step down on the grammatical hierarchy. Insults for lesbians tend to be grammatically feminine, because on the grammatical hierarchy, there is nothing lower than the feminine, apart from neuter, which has more of a dehumanising effect than an insult to somebody’s gender or sexuality (McConnell-Ginet 2014:23; Motschenbacher, personal communication, 27 July 2016). Markedness Through iconisation men became icons for humanity, and so were able to represent the whole population. This is also known as ‘prototypicality’, where one prototypical member of a group, comes to represent the rest of the group. Thus, men were placed at the top of the social hierarchy, which was then reflected onto language. Terms such as man became inclusive or unmarked terms, able to refer to the whole of humanity. Queer linguistics sees markedness as a tool for establishing normative ideologies (Motschenbacher 2010:94–6), or as a means of promoting 278 Ann CoAdy certain values (Klinkenberg 2006:21). In other words, the grammar system is a means of reinforcing (or challenging) dominant social values: The pattern in formal linguistics has been to interpret marked forms in relation to their unmarked counterparts much in the way that alternative expressions of gender and sexuality have traditionally been ideologically viewed in relation to their statistically more common heteronormative counterparts – precisely the ideology that queer theory seeks to challenge, not to uphold. (Barrett 2014:215) In French, the masculine is often used as the unmarked term. However, there has been much debate in recent years over this, especially when referring to a specific woman. In the Introduction, I referred to Julien Aubert, who was fined for addressing Sandrine Mazetier in the masculine, and the resulting media debate about whether or not the male politician was correct or not. The Académie française, the official French language authority (but see Viennot, Candea, Chevalier, Duverger, and Houdebine 2016 for a critique of the institution) came to his defence, claiming that: Si, en effet, le français connaît deux genres, appelés masculin et féminin, il serait plus juste de les nommer genre marqué et genre non marqué. Seul le genre masculin, non marqué, peut représenter aussi bien les éléments masculins que féminins. (Académie française 2014) [If, indeed, French has two genders, called masculine and feminine, it would be more accurate to call them the marked gender and the unmarked gender. Only the masculine, the unmarked gender, can represent masculine as well as feminine elements.] Nevertheless, the concept of markedness is controversial in linguistics, and there is no general agreement on which criteria are necessary to show markedness, whether some are more important than others, what to do if the criteria give conflicting results, or how they interrelate (Waugh and Lafford 2000:276). There are several different types of markedness, for instance, semantic, distributional, contextual, conceptual, and formal markedness, and the masculine is not always the unmarked term in a gendered pair. There are many examples where the feminine is unmarked, for example, the feminine vache [cow] is the semantically unmarked term, used to refer to cattle in general, not the masculine boeuf [bull]. Other feminine semantically unmarked terms include poule [hen] compared to poulet [cock/rooster], oie [goose] compared to jars [gander]: The main reason for this would seem to be that males of the species are normally kept in smaller numbers by farmers than females, and purely for breeding: the main stock is female, and this is treated […] as the unmarked norm. (Lyons 1977:308) The origin of sexism in lAnguAge 279 Although the above examples refer to animals, it serves to highlight that the unmarked form represents the socially more valued form, or simply the more frequently encountered form, whether it is masculine or feminine. In fact, Haspelmath (2006) argues that the term ‘markedness’ should be replaced with other, more precise terms. In the case of the masculine being the de facto unmarked form, he suggests using standard semantic concepts like hyponymy (e.g. man to refer to males only) and polysemy (e.g. man having two different meanings – an inclusive and an exclusive one) (Haspelmath 2006:28). This would put the emphasis on the function that the word fulfils rather than suggesting that that is has an innate unmarked value. Etymologically speaking, man in English, and homme in French, were originally generic terms to refer humans in general. Mann (or man) in Old English meant human, for example wifmann literally meant female (wif) human (mann). A man was referred to as a wermann,3 literally a male (wer) human (mann) before gradually narrowing down to refer only to men (Curzan 2003:62; Baron 1986:138). This process of semantic restriction can also be seen in French: homme comes from the Latin homo, which also meant human, as in the term homo sapiens (vir was used to refer to a man and mulier referred to a woman). In fact, this narrowing down from human to man seems to be a widespread phenomenon in many different languages (Doleschal 2015:1161). Haspelmath explains the processes involved in how some terms become unmarked and other marked: Consider, as an example, the word America. Originally this referred to the entire continent in the western hemisphere that Europeans had become aware of after 1492. But English speakers of course used America primarily for the part of the continent that was settled from England, and nowadays it has become restricted to the United States of America. It is no longer possible to cancel this enriched meaning in English (*I’m from America, more specifically from South America). The enriched meaning has become conventionalized. (Haspelmath 2006:51) Rather than for any linguistic reason, markedness and genericity are based on the relative importance and power of one group over another. It is no accident that the USA, the most powerful country on the American continent, has appropriated the term for itself. Had Canada been the more influential country, we would probably be calling it ‘America’ today. The more powerful a group, the more frequently we talk about them. The more frequent a term is, the more likely it is to be shortened, which is simply due to linguistic economy. The less powerful the group is, the less frequently they will be talked about, and the less likely it is that the term referring to them will be shortened. Markedness is essentially about distinguish- 280 Ann CoAdy ing what is seen as normal (unmarked), from what is abnormal (marked) (Barrett 2014:215). It seems as though man and homme became the unmarked terms because men were, quite simply, talked about more often than women, because they were the more powerful, thus socially valued group. For most people Man no longer refers to all humans, but only to male humans. Psycholinguistic studies tend to show that in the majority of cases, masculine nouns are understood to refer to men, rather than to both women and men (see Brauer and Landry 2008; Stahlberg, Sczesny and Braun 2001; Gastil 1990). Centuries ago, it did describe all humans, but since at least 1000 ce it has been used to refer exclusively to adult males (Curzan 2003:167). Those against non-sexist language reform often refer to the etymology of man and homme as evidence of their current generic value, but this is a rather simplistic idea, which Curzan describes as ‘etymological fallacies’, in other words: that words ‘mean’ – in some fundamental way – what they used to mean or originally meant, and all subsequent semantic changes are corruptions or temporary ‘misunderstandings’ of the ‘correct meaning’. Words fundamentally mean what speakers believe that words mean and what they use words to mean. (Curzan 2003:175) Silverstein (1979:193) also mentions that looking for a word’s ‘true’ or central meaning in its etymological origins is a common linguistic ideology, not unlike the Tower of Babel, in which speakers look back to an imaginary ‘Golden Age of perfection’ in language (Deutscher 2006:80). Arguing that because man referred to all humans almost 1000 years ago, it still does today, is about as logical as arguing that because girl used to mean ‘a child of any sex’, it still does today (Curzan 2003:133). Fractal recursivity Fractal recursivity is a term borrowed from geometry, which refers to two interrelated phenomena. A fractal is a pattern which (a) is the same across different scales (i.e. it looks the same whether we zoom in or out), and (b) is driven by recursion (repetition) of itself. Fractals can also be found in nature (e.g. sunflowers, Romanesco broccoli, crystals in snowflakes) and art (e.g. Jackson Pollock, the Sierpinski triangle). Within Irvine and Gal’s framework, fractal recursivity describes how the dichotomies created from iconisation are reflected onto some other level (e.g. gender, sexuality, ethnicity, social class …) and repeated. Thus, in its original format, fractal recursivity projects oppositions, which are created at a linguistic level, onto other semiotic tiers, like gender and sexuality. The origin of sexism in lAnguAge 281 One example of fractal recursivity is the effect that different pronunciations of [s] can have on gender and sexuality perception. Men tend to produce a lower frequency ([s-]) than women, who tend to produce a higher, ‘crispier’ ([s+]). However, this seems to be socially conditioned rather than biological (Hazenberg 2016:274). Obviously there is a considerable amount of variation among men and women – some men produce high frequency [s]’s, and some women produce low frequency [s]’s. What is interesting here is that the pronunciation of [s] is projected onto another semiotic tier, in this case male sexuality. A higher frequency [s] pronunciation becomes an iconic marker of gay men, and indexes a whole host of traits related to stereotypical male homosexuality, such as effeminate behaviour. In addition, this phenomenon is not restricted to English, and has been noted in other languages (see Welker 2016 for Spanish; and Pharao, Maegaard, Møller and Kristiansen 2014 for Danish). In a study of perceptual bias of the pronunciation of [s] in English (Munson and Zimmerman 2006), male participants were perceived as gay if they used a high-frequency [s], whatever their actual sexuality. Obviously not all gay men produce [s+], and not all straight men produce lower-frequency [s-], but iconisation tends to blur any inter-group similarities. Iconisation ‘describes how linguistic phenomena are portrayed as if they flowed ‘naturally’ from a social group’s biological or cultural essence’ (Milani 2010:120). In this example, a certain pronunciation of [s] is projected onto male sexuality, and portrayed as a result of gay men’s ‘naturally’ effeminate nature. This creates a ‘natural’ opposition between gay and straight men, with a clear social hierarchy. This higher-frequency pronunciation of [s] does not seem to be the result of gay men’s biological make-up,4 and the notion of ‘cultural essence’ is highly problematic. It seems more likely that the pronunciation of [s] by gay men is used as an identity marker under certain circumstances, when they want to make their gay identity apparent. As with iconisation, I have reversed the process of fractal recursivity for my analysis of sexist language. It is not the linguistic feature, which is projected onto gender, but gender that is projected onto the language. Probably the most powerful example of fractal recursivity is the origin of the feminine grammatical gender itself, which has never been definitively proven. Historical linguists generally concur that there were two noun classes in Indo-European – animate and inanimate (Elmiger 2008:51). However, the reasons as to how these two categories came about have been under discussion for over a century (Luraghi 2011:436). It seems that at a relatively recent point, the feminine originated as an offshoot of the animate category. As to why a third gender (the feminine) should emerge in the first 282 Ann CoAdy place, Luraghi claims that, ‘the only possible motivation for a new gender which expands on an animacy-based three-gender system is sex’ (Luraghi 2011:448). French feminist linguist, Claire Michard (1996:44) also supports this perspective, and argues that the feminine grammatical gender emerged because the female sex was already a socially marked sex. Marking women in this way allowed men to appropriate the notion of ‘unmarked human’, while relegating women to the status of ‘marked human’. This means that the feminine was, from its very origin, a restricted gender (because it could only refer to females and not humans in general). Luraghi concurs on this point. She suggests that even before the emergence of the new feminine gender in Indo-European, words relating to females were already being linguistically marked with suffixes: if one looks at Anatolian,5 where a feminine grammatical gender is not available, one finds a number of nouns that refer to human females and derive from masculine nouns with the addition of the suffix -(š)šara-, as in ḫaššuššaraš ‘queen’, from ḫaššuš ‘king’ or išḫaššaraš ‘lady’, from išḫaš ‘lord’ (Hoffner and Melchert 2008). (Luraghi 2009a:19) A second example of my modified version of fractal recursivity is social gender being projected onto inanimate nouns. As previously mentioned, historical linguists do not really know why grammatical gender for inanimate nouns emerged, and why, for example, a bridge should be masculine in French (un pont) but feminine in German (Die Brücke) (for how grammatical gender affects how speakers think about the objects concerned, see Boroditsky, Schmidt and Phillips 2003; Sera et al. 2002). Grammatical gender for inanimate nouns has often been described as semantically arbitrary, with no basis in human physiology or sexual behaviour. However, it is not entirely semantically arbitrary. Scholars have argued that all gender systems are at least partially semantic (Corbett 1991:8; Violi 1987:15). There are certain classes of nouns which can be categorised according to their semantic value; for example, names of trees, days of the week, months and seasons, cheeses, wines, metals and minerals are usually masculine in French; names of cars and academic subjects are usually feminine. This said, for the majority of inanimate nouns, grammatical gender has no semantic basis whatsoever, and is based on morphology and phonology (Corbett 1991:61). One important result of this leakage between grammatical and social gender (Romaine 1999:63–90; Violi 1987) is the current rule of the gender agreement in French. Codification of the language flourished in the sixteenth century, when hierarchies were established between nouns and their relative importance. Although there is no grammatical agreement between nouns and their qualifiers in English, grammarians recommended putting The origin of sexism in lAnguAge 283 nouns in order of importance, e.g. king and queen (not queen and king), father and mother (not mother and father): ‘The concept of worthiness is […] a reflection of a natural order that places man at the head of creation, with woman in a subordinate, subservient, and frequently invisible place’ (Baron 1986:98). The same concept of worthiness can be found in French grammar, as well as word order e.g. un hommemasc et cinq milliards de femmesfem sont mortsmasc [one man and five billion women died]. The one man will generally come before the five billion women, and the past participle morts is in the masculine. However sexist this may be, we can clearly see some kind of logic in it. What is less logical is the idea that bonnet [hat] is more worthy than écharpe [scarf ]. Once bonnet has been identified as a masculine noun and écharpe as feminine, the ‘logic’ becomes clearer. Thus, in the sentence le bonnetmasc et l’écharpefem sont vertsmasc [the hat and the scarf are green], verts is in the masculine. The masculine noun bonnet is considered more worthy than the feminine écharpe by virtue of its grammatical gender (attributed for morphological reasons), even though both are inanimate objects, with no obvious masculine or feminine qualities. Social gender, and its hierarchy, was reflected onto inanimate objects. Erasure As opposed to iconisation and fractal recursivity, which I slightly modified, I am able to use erasure in its original format. Erasure is the process in which ideology, in simplifying the sociolinguistic field, renders some persons or activities (or sociolinguistic phenomena) invisible. Facts are inconsistent with the ideological scheme either go unnoticed or get explained away. […] Because a linguistic ideology is a totalizing vision, elements that do not fit its interpretative structure – that cannot be seen to fit – must be either ignored or transformed. (Irvine and Gal 2000:38) There are numerous examples of erasure regarding feminist linguistics (Baudino 2001; Viennot 2014), a process whereby any evidence that contradicts the naturalness of one side of the argument is ignored, not recorded, not discussed, and then simply fades away into the shadows of forgotten history. Previously, I discussed how the processes of iconisation and fractal recursivity resulted in the masculine becoming the generic form. In this part I will demonstrate how counter discourses to the masculine generic were erased from the public arena. When Vulgar Latin6 transitioned into French, the neuter gender in Latin was absorbed by the masculine in French. This phenomenon is part of the reason that some see the masculine as more inclusive, and therefore able to fulfil a generic role: 284 Ann CoAdy L’une des contraintes propres à la langue française est qu’elle n’a que deux genres: pour désigner les qualités communes aux deux sexes, il a donc fallu qu’à l’un des deux genres soit conférée une valeur générique afin qu’il puisse neutraliser la différence entre les sexes. L’héritage latin a opté pour le masculin. (Académie française 2014) [One of the constraints particular to the French language is that it only has two genders: in order to designate the qualities that are common to both sexes, one of the two genders had to be attributed a generic value so that it was able to neutralise the difference between the sexes. Because of our Latin heritage, the masculine was chosen.] Khaznadar, on the other hand, vehemently disagrees claiming that, ‘[d]ire que le masculin français est “héritier du neutre latin” est une contrevérité’ [‘to say that the French masculine is the “heir of the neuter in Latin” is an untruth’] (Khaznadar 2007:33). In fact, the Académie française ignores several arguments that contradict their position. There are four main issues that need to be addressed with regard to the Latin neuter: (i) many neuter nouns became feminine; (ii) the etymology of neuter is ambiguous; (iii) most neuter nouns were inanimate; and (iv) neuter does not necessarily mean generic. Firstly, although the masculine did absorb most neuter nouns in Latin, over a third became feminine nouns in Old French (Polinksy and van Everbroeck 2003:376–8) e.g. mareneut sg [sea] → merfem sg, gaudianeut pl [joys, delights] → joiefem sg, and folianeut pl [leaves] → feuillefem sg (Solodow 2010:230). This can be explained by the fact that gaudiumneut sg and foliumneut sg were more widely used in their plural forms gaudianeut pl and folianeut pl in Vulgar Latin, which, because they ended in -a, were mistaken for the feminine singular, and so became feminine in French.7 Secondly, the etymology of neuter does not necessarily support the claim that it has a generic value. Neuter (ne- + -uter) literally means ‘not either’ (Kennedy 1906:14). It could therefore be argued that if neuter means neither masculine nor feminine, that it excludes rather than includes both of these noun classes, defies logic and is ‘littéralement un non-sens’ [literally nonsense] (Khaznadar 2006). This argument is supported by studies of the first Latin grammars, one of which, De lingua latina by M. Terentius Varro (116–27 bce), translates the Greek σκεύη [things] (Corbeill 2008:80) as neutrum [neuter] in Latin (Burr 2012:31). Other Latin works also confirm this perspective: in his Institutiones grammaticae, Priscianus (fifth century ce) wrote that the communis (common gender) referred to both males and females, as opposed to the neuter, which signified neither male nor female (Burr 2012:31). The origin of sexism in lAnguAge 285 Thirdly, the vast majority of neuter nouns in Latin, as well as IndoEuropean (Luraghi 2011:440), had inanimate referents (Khaznadar 2007:33), apart from a few exceptions such as vulgus [the common people] (Kennedy 1906:222) or scortum and prostibulum [prostitute] (Pitavy 2014:175). It seems very unlikely that the handful of animate neuter nouns which became masculine, transmitted their ‘unmarked’ quality to the thousands of existing masculine nouns, thus giving these masculine nouns a kind of double identity – marked when used with a male referent, and unmarked when employed in a non-specific context. It could also be argued that the absorption of the neuter by the masculine simply increased the size of the masculine noun class, rather than modifying the value or quality of the nouns already there. The neuter was as marked as any other gender (in that it had specific endings that coded it as neuter), and it was only ‘neutral’ in that it referred to inanimate entities. Finally, the underlying problem here is seems to be a conflation of the terms neuter and generic, which are not synonymous. Neuter refers to a specific noun class, which in Latin was composed almost entirely of inanimate nouns e.g. templum [temple], mare [sea], carmen [song / poem]. Generic, or hyponym, on the other hand, refers to the capacity of a noun to refer to a whole class or group of things e.g. fruit is a generic term referring to bananas, apples, oranges, kiwis etc. Neuter nouns do not therefore necessarily have a generic value. In fact, any noun is capable of fulfilling the role of generic depending on the context, for instance in the pair fruit-banana, fruit is the generic term. In the pair banana-ensete (ensete is a variety of banana), banana is the generic form. However, according to traditional grammar, the masculine has an inherent generic value when referring to animate nouns thanks in great part, to its absorption of two thirds of Latin neuter nouns, which were not necessarily generic, and which referred to inanimate objects for the great majority. Using the Latin heritage of French mobilises a discourse of tradition. Woollard notes: representations of the history of languages often function as Malinowskian charter myths,8 projecting from the present to an originary past a legitimation of contemporary power relations and interested positions. (Or, we might prefer to say, projecting from the past a legitimating selection of one from among contending centers of power in the present). (Woolard 2004:58) In other words, the story-tellers of history cherry-pick the elements which support their arguments, while erasing those that do not. Khaznadar (2006) also noticed that the choice of the Latin heritage argument by the Académie française was not anodyne: ‘Inscrire dans le débat les origines latines 286 Ann CoAdy du français impose le respect aux non-initiés, les impressionne peut-être’ [incorporating the Latin origins of French into the debate imposes respect from the non-initiated, perhaps intimidating them]. Which discourses get promoted, and which are erased is not up to just anyone. In order to promote a particular discourse, one needs to be in a position of linguistic authority, hence the power of language bodies like the Académie française. Early on, I briefly mentioned Silverstein,9 who claimed that generic he is a ‘structurally dictated indexical usage’ (Silverstein 1985:256). Implying that inclusive masculine is simply a fact of grammar fails to take into account the fact that languages do not evolve in a social and cultural vacuum (Curzan 2003:184). A language structure does not just build itself, speakers shape it over centuries, and with powerful speakers having more influence than powerless speakers. Cameron labels this tactic ‘mystification’: ‘to deny that authority could be at work (by saying, for instance, that such and such a usage is ‘just a fact about the grammar of x’) is a mystification’ (Cameron 1995:6). There is always somebody behind language change; the question is how visible they are. Discussion Prior work on sexism in language has identified the causes of sexism in language as being rooted in an androcentric view of the world (Cameron 1995:134; McConnell-Ginet 1984:124), but not the social mechanisms involved in how it is historically produced and reproduced. This article argues that the processes of iconisation, fractal recursivity, and erasure provide a unifying theory to explain the social and semiotic mechanisms involved in sexist grammar and semantics. Iconisation results in the partitioning of humans into two binary groups based on gender. Men became the unmarked icon of the whole of humanity. This partitioning, and resulting hierarchy, was then projected onto language through the process of fractal recursivity, and the masculine gender became the generic form. Finally through erasure, certain discourses were able to become dominant, while others were erased from the public arena. It is through these processes that current grammatical norms such as le bonnetmasc et l’écharpefem sont vertsmasc [the hat and the scarf are green] can be explained. They are norms, which certain people have been in a position to implement over the centuries. The generic status of the masculine is ‘an integral part of a doctrine which […] was consciously constructed over the centuries and [that] the natural order it proposes concurs with the idea that men are “worthier” than women’ (Burr 2012:30). An understanding of not just why sexism in language exists, but how it exists allows us to deconstruct arguments against feminist linguistic reforms more easily. The origin of sexism in lAnguAge 287 It allows us to argue that institutions such as the Académie française are not just sexist, but that their arguments are linguistically unsound (also see Viennot et al. 2016). In the Introduction, I said that QL opens up exciting new ways to approach sexism in language. From a poststructuralist perspective, grammar is simply the discursive sedimentation of the repetition of certain categories (Motschenbacher 2010:87). This allows us to fuse elements of Second Wave analyses of grammar as relatively stable, with a poststructuralist approach that sees language as performative, and thus malleable. The masculine generic is the discursive sedimentation of ideologies of gender (the superiority of males) and ideologies of language (meaning as fixed). If sexist ideologies have changed immensely in the West since the mid-twentieth century, there is still a high level of resistance to feminist linguistic interventions in France, especially with regard to the masculine generic. This article highlights the how the unmarked quality of the masculine is simply the result of semantic restriction, and shows how grammatical gender can act as a normative mechanism for gender and sexuality. Because Queer criticises binaries, it puts the visibility principle of feminist linguistic planning, used for most Romance languages, into question. In France, terms for women have been promoted as part of the féminisation project, but this has resulted in ‘categorical homeostasis’ (Silverstein 1985:252), meaning that the same binary system is still in place. Queer linguistics addresses the very existence of gender and its binary nature. In this way, rather than simply adding feminine forms to balance masculine ones, we can question why a binary grammatical system exists in the first place, which has important repercussions on the direction of future feminist language policy. Recent work in QL has identified various solutions to sexist linguistic structures. For instance, using one orthographic word with punctuation marks separating the word into either feminine or masculine (e.g. étudiantsmasc + étudiantesfem = étudiant-e-s [student-s]). They are at the same time both masculine and feminine, and neither masculine nor feminine. However, abbreviated splitting (étudiant-e-s), and full splitting (étudiantset étudiantesfem) still separate masculine and feminine, and so ‘can be masc read iconically as a sign of female-male incompatibility’ (Motschenbacher 2014:225). Slightly more radical is the blending of the feminine and masculine forms to create a neutral one: professionnelsmasc + professionnellesfem = professionèles [professionals] (Labrosse 2002:100), and ilmasc [he] + ellefem [she] = ille [s/he / singular they] (Abbou 2011:63) (also see Elmiger 2015 for a good overview of neutralisation techniques in French). Abbreviated splitting and full splitting are becoming more and more commonplace in French. However, blending is still a very marginal practice. 288 Ann CoAdy A Queer perspective on gendered linguistic forms needs to be taken into account before any further language reforms are carried out. Furthermore, in order for non-sexist language reforms to work, we need to know why people reject or accept change. Analysing the origin of the problem allows us to better evaluate the current linguistic situation, and people’s attitudes to it. In fact, many arguments that people mobilise today for or against change are based on language ideologies that have been constructed through the processes of iconisation, fractal recursivity, and erasure, and which are reinforced by language gatekeepers whose authority and ideological bias is often hidden. About the author Ann Coady is a PhD student at Sheffield Hallam University. Her thesis is a comparative study of feminist linguistic reforms in French and English, specifically the ideologies of language that emerge in debates about politically motivated language change. Her other research interests include corpus linguistics, CDA, historical linguistics, Queer linguistics and language planning. Her publications include ‘La Construction socio-discursive du masculine générique: discours et contre-discours‘ in Bailly et al. (eds), Pratiques et langages du genre et du sexe: déconstruire l’idéologie sexiste du binarisme (2016), and ‘Mademoiselle va-t-il perdurer “malgré les oukases”?’ in Cahiers de Linguistique (2014). Notes 1 2 3 4 5 6 The UMP (Union pour un Mouvement Populaire) is the major right-wing party in France. In 2015 it changed its name to Les Républicains. The PS (Parti Socialiste) is the major left-wing party in France. Although Motschenbacher asked his participants for terms denoting ‘masculine women’ and ‘feminine men’, not specifically lesbian women or gay men, many of the terms elicited were insults for gays and lesbians. Werewolf (literally man + wolf) is the only surviving remnant of the term wer [man i.e. male] in Modern English (Baron 1986:139). Munson posits that genetic factors may play some role in the more frequent ‘lisp’ pronunciation that he found in a study on boys with gender ‘dysphoria’ (Munson, Crocker, Pierrehumbert, Owen-Anderson and Zucker 2015), although he does not go into detail as to exactly what these genetic differences may be. Anatolian refers to a group of extinct Indo-European languages that were spoken in Asia Minor. The best known is Hittite, which had a noun-class system based on an animate/inanimate distinction, rather than a masculine-feminine distinction. It is thought that the masculine-feminine gender divide happened in late PIE, after Anatolian had spilt off from that branch (Beekes 2011). This suggests that grammatical gender originated as a reflection of social gender. I am not suggesting that language does not influence society, but in this case social gender seems to be the most logical explanation for the origin of grammatical gender. Vulgar Latin refers to the forms of Latin spoken by the common people, as opposed to written Classical Latin. The origin of sexism in lAnguAge 7 8 9 289 The -a ending of the neuter plural goes all the way back to Indo-European. Many collective nouns were neuter, and so took the ending -a, a suffix also shared by the feminine. For a long time, linguists assumed that there was a semantic link between collectives and the feminine, but recently this has been discounted. It seems as though the neuter plural and the feminine were two separate morphological developments (Luraghi 2009a). Anthropologist Bronisław Malinowski advocated that myths tended to advance the agendas of people in power. Incidentally, Silverstein was part of the group who wrote the infamous ‘pronoun envy’ letter to the Harvard Crimson in 1971 (Silverstein 1971). References Abbou, J. (2011) Double gender marking in French: a linguistic practice of antisexism. Current Issues in Language Planning 12(1): 55–75. https://doi.org/10.1080/1466420 8.2010.541387 Académie française (2014) La féminisation des noms de métiers, fonctions, grades ou titres – mise au point de l’Académie française. 10 October. Retrieved on 30 November 2014 from www.academie-francaise.fr/actualites/la-feminisation-des-nomsde-metiers-fonctions-grades-ou-titres-mise-au-point-de-lacademie. Baron, D. (1986) Grammar and Gender. New Haven, CT: Yale University Press. Barrett, R. (2014) The emergence of the unmarked: queer theory, language ideology, and formal linguistics. In L. Zimman, J. Davis and J. Raclaw (eds) Queer Excursions: Retheorizing Binaires in Language, Gender and Sexuality 195–224. Oxford: Oxford University Press. Baudino, C. (2001) Politique de la langue et différence sexuelle: la politisation du genre des noms de métier. Paris: L’Harmattan. Beekes, R. S. P. (2011) Comparative Indo-European Linguistics: An Introduction (2nd edition). Amsterdam: John Benjamins. https://doi.org/10.1075/z.172 Blommaert, J. (1999) The debate is open. In J. Blommaert (ed.) Language Ideological Debates 1–38. Berlin: Mouton de Gruyter. Blommaert, J. (2006) Language ideology. In K. Brown (ed.) Encyclopedia of Language and Linguistics (2nd edition) 510–22. Oxford: Elsevier. Retrieved from https://tinyurl. com/1410944-2006. Boroditsky, L., Schmidt, L. A. and Phillips, W. (2003) Sex, syntax, and semantics. In D. Gentner and S. Goldin-Meadow (eds) Language in Mind: Advances in the Study of Language and Thought 69–79. Cambridge, MA: MIT Press. Brauer, M. and Landry, M. (2008) Un ministre peut-il tomber enceinte ? L’impact du générique masculin sur les représentations mentales. L’Année Psychologique 108: 243–72. https://doi.org/10.4074/S0003503308002030 Burr, E. (2012) Planification linguistique et féminisation. In F. Baider and D. Elmiger (eds) Intersexion 29–40. Munich: Lincom. Cameron, D. (1995) Verbal Hygiene. London: Routledge. Chetcuti, N. and Greco, L. (2012) La Face cachée du genre: langage et pouvoir des normes. Paris: Presses Sorbonne Nouvelle. Corbeill, A. (2008) ‘Genus quid est?’ Roman scholars on grammatical gender and biological sex. Transactions of the American Philological Association 138(1): 75–105. https://doi.org/10.1353/apa.0.0007 290 Ann CoAdy Corbett, G. G. (1991) Gender. Cambridge: Cambridge University Press. https://doi. org/10.1017/CBO9781139166119 Coutant, A. (2014) Masculin/féminin en français: (dé)construction par le genre des sexes et des sexualités. Paper presented at the Journée d’études du séminaire interne du LASCO, Univerisité Paris Descartes, June. Retrieved from www.academia.edu/ 7454768/Presentation_des_travaux_en_cours_corpus_methodes_resultats_. Curzan, A. (2003) Gender Shifts in the History of English. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9780511486913 Deutscher, G. (2006) The Unfolding of Language. London: Arrow Books. Doleschal, U. (2015) Gender marking. In P. Müller O., I. Ohnheiser, F. Rainer and S. Olsen (eds) Handbook of Word Formation . Berlin: Mouton de Gruyter. Elmiger, D. (2008) La féminisation de la langue en français et en allemand: querelle entre spécialistes et réception par le grand public. Paris: Honoré Champion. Elmiger, D. (2015) Masculin, féminin : et le neutre? Le statut du genre neutre en français contemporain et les propositions de ’neutralisation‘ de la langue. Implications Philosophiques. Retrieved from www.implications-philosophiques.org/actualite/une/masculin-feminin-et-le-neutre/ Fracchiolla, B. (2014) ‘Madame le Président’, ‘Monsieur la députée’. 23 January. Retrieved from http://violenceverbale.hypotheses.org. Gastil, J. (1990) Generic pronouns and sexist language: the oxymoronic character of masculine generics. Sex Roles 23(11/12): 629–43. https://doi.org/10.1007/BF00289252 Haspelmath, M. (2006) Against markedness (and what to replace it with). Journal of Linguistics 42(1): 25–70. https://doi.org/10.1017/S0022226705003683 Hazenberg, E. (2016) Walking the straight and narrow: linguistic choice and gendered presentation. Gender and Language 10(2): 270–94. https://doi.org/10.1558/genl. v10i2.19812 Hoffner, H. and Melchert, C. (2008) A Grammar of the Hittite Language. Winona Lake, IN: Eisenbrauns. Holmes, J. (2006) Gendered Talk at Work: Constructing Gender Identity through Workplace Discourse. Malden, MA: Blackwell. https://doi.org/10.1002/9780470754863 Houdebine, A.-M. (2003) Trente ans de recherche sur la différence sexuelle, ou Le langage des femmes et la sexuation dans la langue, le discours, les images. Langage et Société 106: 33–61. https://doi.org/10.3917/ls.106.0033 Irvine, J. and Gal, S. (2000) Language ideology and linguistic differentiation. In P. Kroskrity V. (ed.) Regimes of Language: Ideologies, Polities and Identities 35–84. Santa Fe, NM: School of American Research Press. Jaffe, A. (1999) Ideologies in Action: Language Politics on Corsica. New York: Mouton de Gruyter. https://doi.org/10.1515/9783110801064 Jagose, A. (1996) Queer Theory: An Introduction. New York: New University Press. Jagose, A. (2009) Feminism’s Queer Theory. Feminism and Psychology 19(2): 157–74. Johnson, S. (2005) Spelling Trouble? Language, Ideology and the Reform of German Orthography. Clevedon: Multilingual Matters. Kennedy, B. H. (1906) The Revised Latin Primer (12th impression). London: Longmans, Green, & Co. Retrieved from https://archive.org/details/ revisedlatinprim00kenniala. Khaznadar, E. (2006) La Pertinence linguistique du rapport de la Commission Générale de Terminologie sur la féminisation. Paper presented at the GrALP, Université de Pau, France, May. Retrieved from www.edwige-khaznadar-parite-linguistique.fr/ affiche_articles.php?id_art=14. The origin of sexism in lAnguAge 291 Khaznadar, E. (2007) Le non-genre académique: doctrine de la domination masculine en France. Nouvelles Questions Féministes 26(3): 25–38. https://doi.org/10.3917/ nqf.263.0025 Klinkenberg, J.-M. (2006) Le Linguiste entre science et idéologie: le discours épilinguistique sur la féminisation comme trace d’un savoir dégradé. Bulletin VALS-ASLA (Association suisse de linguistique appliquée) 82(2): 11–32. Labrosse, C. (2002) Pour une langue française non sexiste. Montréal: Les Editions des Intouchables. Lazar, M. (2014) Recuperating feminism, reclaiming femininity: hybrid postfeminism I-dentity in consumer advertisements. Gender and Language 8(2): 205–24. https:// doi.org/10.1558/genl.v8i2.205 Le Monde. (2014) Un député sanctionné pour un ‘Mme LE président’. Le Monde (8 October). Retrieved from www.lemonde.fr/politique/video/2014/10/08/un-deputesanctionne-pour-un-madame-le-president_4502359_823448.html. Luraghi, S. (2009a) Indo-European nominal classification: from abstract to feminine. In S. Jamison W., C. Melchert and B. Vine (eds) Proceedings of the 20th Annual UCLA Indo-European Conference 115–31. Bremen: Hempen. Retrieved from http://attach. matita.net/silvialuraghi/file/Abstract%20and%20feminine.pdf. Luraghi, S. (2009b) The origin of the feminine gender in PIE: an old problem in a new perspective. In V. Bubenik, J. Hewson and S. Rose (eds) Grammatical Change in Indo-European Languages: Papers presented at the workshop on Indo-European Linguistics at the XVIIIth International Conference on Historical Linguistics, Montreal, 2007 3–13. Amsterdam: John Benjamins Publishing Company. Luraghi, S. (2011) The origin of the Proto-Indo-European gender system: typological considerations. Folia Linguistica 45(2): 435–63. https://doi.org/10.1515/flin.2011.016 Lyons, J. (1977) Semantics (2 vols). Cambridge: Cambridge University Press. McConnell-Ginet, S. (1984) The origins of sexist language in discourse. Annals of the New York Academy of Sciences 433: 123–35. https://doi.org/10.1111/j.1749-6632.1984. tb14764.x McConnell-Ginet, S. (2002) ‘Queering’ semantics: definitional struggles. In K. CampbellKibler, R. Podesva J., S. Roberts J. and A. Wong (eds) Language and Sexuality: Contesting Meaning in Theory and Practice 137–60. Stanford, CA: CSLI Publications. McConnell-Ginet, S. (2014) Gender and its relation to sex: the myth of ‘natural’ gender. In G. G. Corbett (ed.) The Expression of Gender 3–38. Berlin: De Gruyter. Michard, C. (1996) Genre et sexe en linguistique: les analyses du masculin générique. Mots, Les langages du politique 49: 29–47. Milani, T. M. (2010) What’s in a name? Language ideology and social differentiation in a Swedish print-mediated debate. Journal of Sociolinguistics 14(1): 116–42. https:// doi.org/10.1111/j.1467-9841.2009.00435.x Mills, S. (2008) Language and Sexism. Cambridge: Cambridge University Press. https:// doi.org/10.1017/CBO9780511755033 Milroy, J. and Milroy, L. (2012) Authority in Language: Investigating Language Prescription and Standardization (4th edition). London, New York: Routledge. https://doi. org/10.4324/9780203124666 Motschenbacher, H. (2010) Language, Gender and Sexual Identity: Poststructuralist perspectives. Amsterdam: John Benjamins Publishing Company. https://doi.org/10.1075/ impact.29 Motschenbacher, H. (2014) Grammatical gender as a challenge for language policy: the 292 Ann CoAdy (im)possibility of non-heteronormative language use in German versus English. Language Policy 13(3): 243–61. https://doi.org/10.1007/s10993-013-9300-0 Motschenbacher, H. (2015) Some new perspectives on gendered language structures. In M. Hellinger and H. Motschenbacher (eds) Gender across Languages (vol. 4) 27–48. Amsterdam: John Benjamins. Motschenbacher, H. and Stegu, M. (2013) Queer linguistic approaches to discourse. Discourse and Society 24(5): 519–35. https://doi.org/10.1177/0957926513486069 Munson, B., Crocker, L., Pierrehumbert, J., Owen-Anderson, A. and Zucker, K. (2015) Gender typicality in children’s speech: a comparison of boys with and without gender identity disorder. Acoustical Society of America 137(4): 1995–2003. https:// doi.org/10.1121/1.4916202 Munson, B. and Zimmerman, L. (2006) Perceptual bias and the myth of the gay lisp. Retrieved from http://goo.gl/Au6S6m. OED (undated) hetero-, comb. form. Oxford English Dictionary online. Retrieved from www.oed.com/view/Entry/86423?rskey=L79U6x&result=2&isAdvanced=false#eid. Pauwels, A. (1998) Women Changing Language. London: Longman. Pharao, N., Maegaard, M., Møller, J. S. and Kristiansen, T. (2014) Indexical meanings of [s+] among Copenhagen youth: social perception of a phonetic variant in different prosodic contexts. Language in Society 43: 1–31. https://doi.org/10.1017/S00474 04513000857 Pitavy, J.-C. (2014) Genre, pronom et personne: une information pertinente dans l’énonciation? Cahiers de Linguistique 40(1): 171–84. Polinksy, M. and van Everbroeck, E. (2003) Development of gender classifications: modeling the historical change from Latin to French. Language 79: 356–90. https:// doi.org/10.1353/lan.2003.0131 Rajilic, S. (2016) Challenges and limitations to non-sexist language reforms in Serbia. Paper presented at the IGALA9, City University, Hong Kong. Romaine, S. (1999) Communicating Gender. New York: Lawrence Erlbaum. Sera, M. D., Elieff, C., Forbes, J., Burch, M. C., Rodriguez, W. and Dubois, D. P. (2002) When language affects cognition and when it does not: an analysis of grammatical gender and classification. Journal of Experimental Psychology: General 131: 377–97. https://doi.org/10.1037/0096-3445.131.3.377 Sicurella, F. G. (2016) The approach that dares speak its name: queer and the problem of ‘big nouns’ in the language of academia. Gender and Language 10(1): 73–84. https://doi.org/10.1558/genl.v10i1.20895 Silverstein, M. (1971) Pronoun envy. The Harvard Crimson (16 November). Retrieved from www.thecrimson.com/article/1971/11/16/pronoun-envy-pto-the-editors-of. Silverstein, M. (1979) Language structure and linguistic ideology. In P. R. Clyne, W. F. Hanks and C. L. Hofbauer (eds) The Elements: A Parasession on Linguistic Units and Levels 193–247. Chicago, IL: Chicago Linguistic Society. Silverstein, M. (1985) Language and the culture of gender: at the intersection of structure, usage, and ideology. In E. Mertz and R. J. Parmentier (eds) Semiotic Mediation 219–59. Orlando, FL: Academic Press. https://doi.org/10.1016/B978-0-12-4912809.50016-9 Solodow, J. B. (2010) Latin Alive: The Survival of Latin in English and the Romance Languages. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO97805 11809903 Spender, D. (1980) Man Made Language. London: Routledge & Kegan Paul. The origin of sexism in lAnguAge 293 Stahlberg, D., Sczesny, S. and Braun, F. (2001) Name your favorite musician: effects of masculine generics and of their alternatives in German. Journal of Language and Social Psychology 20: 464–69. https://doi.org/10.1177/0261927X01020004004 Sunderland, J. (2004) Gendered Discourses. London: Palgrave Macmillan. https://doi. org/10.1057/9780230505582 Thomas, P.-L. (2002) Le serbo-croate (bosniaque, croate, monténégrin, serbe): de l’étude d’une langue à l’identité des langues. Revue des études slaves 74(2): 311–25. https://doi.org/10.3406/slave.2002.6801 Viennot, É. (2014) Non, le masculin ne l’emporte pas sur le féminin! Petite histoire des résistances de la langue française. Donnemarie-Dontilly: Éditions iXe. Viennot, É., Candea, M., Chevalier, Y., Duverger, S. and Houdebine, A.-M. (2016) L’Acdémie contre la langue française: le dossier ‘féminisation’. Donnemarie-Dontilly: Éditions iXe. Violi, P. (1987) Les origines du genre grammatical. Langages 85: 15–34. https://doi. org/10.3406/lgge.1987.1526 Waugh, L. R. and Lafford, B. A. (2000) Markedness. In G. Booij, C. Lehmann and J. Mugdan (eds) Morphologie / Morphology. 1. Halbband 272–80. Berlin: Walter de Gruyter. Welker, C. (2016) Indexicality, vocal gesture, and the negation of muxe identity in Juchitán, Mexico. Presented at the IGALA9, Hong Kong City University Retrieved from www.english.cityu.edu.hk/en/portal/igala/abstract/WelkerCraig.pdf Woolard, K. A. (1998) Language ideology as a field of inquiry. In B. Schieffelin, K. Woolard and P. Kroskrity (eds) Language Ideologies: Practice and Theory 3–49. Oxford: Oxford University Press. Woolard, K. A. (2004) Is the past a foreign country? Time, language origins, and the nation in early modern Spanish. Journal of Linguistic Anthropology 14: 57–80. https://doi.org/10.1525/jlin.2004.14.1.57