Pluri-Grammars for Pluri-Genders: Competing Gender Systems in the Nominal Morphology of Non-Binary French

Abstract

Although much has been written in recent years on the emergence of non-binary English linguistic innovations, comparatively little has been written on non-binary French forms, especially neo-morphemes marking non-binary gender on nouns. As French is a grammatical-gender language with a traditionally binary (masculine and feminine) system, many non-binary Francophones have circumvented the social connection between grammatical gender and human gender in innovating new, non-binary markers for animate nouns and their modifiers. This study uses a mixed methods approach, combining analysis of non-binary French grammars alongside interview data in order to highlight the divergent morphological approaches underlying non-binary marking systems. Three approaches to the formation of non-binary nouns are identified: A Compounding Approach, which combines masculine and feminine markers; a Systematic Approach, which phonologically conditions the use of non-binary allomorphs, with the markers themselves ranging from phonologically novel within French syllabic structure, to homophonous with masculine and/or feminine variants; and an Invariable Approach, which applies a single non-binary marker across all nouns. Ultimately, this study disentangles both morphological patterns in the formation of non-binary words and some of the motivations behind them in an emerging French subtype well-known to be heterogeneous.

1. Introduction

French is a grammatically-gendered language with a binary, masculine-feminine gender system. As with the other Romance languages, grammatical gender assignment is arbitrary with regard to inanimate nouns but retains a semantic function within animate nouns, especially those describing humans (Battye et al. 2001, p. 120; Loporcaro 2018, p. 50; Riegel et al. 2011, p. 329); this function remains whether or not gender is explicitly marked on a given noun (i.e., via a masculine or feminine morpheme). The connection between social gender and grammatical gender has frequently been a topic of debate within Francophone countries. Throughout the 16th and 17th centuries, male grammarians invented prescriptive rules dictating that the masculine should be the default grammatical gender used in describing mixed-gender groups (Labrosse 1996, pp. 27–29), leading to what Viennot (2014, p. 66) labels a “masculinisation” of the French language; prior to this point, gender agreement was triggered based on the gender of the nominal item closest to the modifier (proximity agreement; (Labrosse 1996, pp. 67–69)). However, as women increasingly ventured beyond the domestic sphere leading up to and during the French Revolution (Lodge 2004), linguists and lay-persons alike began to question the use of the generic masculine, instead proposing feminized neologisms that could specifically identify women professionals. These discussions evolved throughout the 18th (Domergue 1784; Dougnac 1982; Kaplan 2022b) and 19th centuries (Christofides 2019). By the mid-20th century, feminist linguistic reformers advocated for increased adoption of feminized titles with the specific goal of visibilizing women in language, in a movement known as féminisation [feminization] (Becquer et al. 1999; Houdebine 1987; Ozello and Marks 1984). Although initially largely ignored by the public (Close 2000; Fleischman 1997), by the late 1990s these feminized titles had gained widespread acceptance in both France (Gervais 2002) and Francophone Canada (Moron-Pueche et al. 2020).

However, not all French feminists were content with state-sanctioned feminization strategies, including feminine neologisms. For example, Abbou (2011) found that some anarchists writing in the 1990s through the early 2000s offered alternatives to these feminine forms, which they opposed on the basis of a broader anti-authoritarian stance. Some authors have gone even further in writing anti-sexist grammars that include neo-agreement markers. Labrosse (1996, p. 9) devises a system of alternative agreement strategies so that “la langue française puisse projeter une image autrement plus conforme aux valeurs et aux aspirations d’égalité de la communauté qui la parle [the French language can project an image otherwise better matching the values and the aspirations of equality of the community that speaks it]1” which operates by constructing a third “genre commun [common gender],” binarily defined as “femmes + hommes [women + men].” More recently, Lessard and Zaccour (2017) created a guide to tackle various strategies of feminization, including the creation of neologisms. While these grammatical systems are revolutionary in their disruption of the French gender system, they differ from the non-binary grammars described in this paper in key ways. First, they are united by goals of developing and illustrating anti-sexist language, which at its core seeks to specifically move away from grammatical rules (e.g., use of the generic masculine in describing mixed groups) viewed as favoring masculine gender marking, and by extension, men in the French language. While the result is a system (or, in the case of Lessard and Zaccour, a series of alternative systems) of inclusive French, these remain grammars predicated on binary notions of inclusivity that work to either emphasize the feminine (i.e., women), or de-emphasize the masculine (i.e., men). Relatedly, they are not grammars that originate within the non-binary Francophone community for the explicit purpose of illustrating non-binary French forms. Although both Labrosse and Lessard and Zaccour’s work has been influential in the construction of alternative French agreement patterns—and serves as a useful point of comparison to non-binary French grammars—their systems are not analyzed in-depth here for the reasons listed above.

In the 21st century, a linguistic movement similar to feminization has emerged aimed at visibilizing non-binary individuals in French. Although attempts at creating a gender-expansive grammar of French are not new (Greco 2019), they have gained significantly more attention in the past decade, with Francophone LGBT+ organizations (Bolter 2019), media (Swamy 2019), and government institutions acknowledging innovative non-binary forms (OQLF 2018; HCE 2016). Historically non-binary Francophones have used a variety of workarounds in order to avoid misgendering themselves or others in French, which operate at syntactic, lexical, and morphological levels.2 However, non-binary neomarkers represent the emergence of a new form of oral and written French that specifically includes non-binary individuals morphologically. For this reason, this paper will specifically focus on how non-binary gender operates at the morphological level in French.

Notwithstanding the fact that much has been written in recent years on the emergence of non-binary English linguistic innovations (e.g., Beemyn and Rankin 2011; Bornstein [1994] 1995; Corwin 2009; Factor and Rothblum 2008), particularly neutral pronouns (Baron 2020; Bergman 2017; Feinberg 1998), comparatively little has been written on non-binary French forms, especially neo-morphemes marking non-binary gender on nouns and adjectives; existing research either collects data exclusively on written forms (Shroy 2017), focuses on pronominal forms (Kosnick 2019), or describes a limited number of oral agreement markers (Dumais 2021; Knisely 2020). We know that language usage by non-binary Francophones varies considerably (Hord 2016; Knisely 2020; La Vie en Queer 2018b; Shroy 2017). However, the morphology of oral non-binary agreement patterns has yet to be comprehensibly studied. This study uses a combination of sociolinguistic interview data and corpus analysis of non-binary French grammars to both outline the major morphological innovations transpiring in non-binary French and to argue that these ongoing changes complicate the traditional French morphological gender system through their plethora of neo-markers. Further, it describes materials on French non-binary grammar that have previously only been available in French.

This paper begins with an overview of nominal and adjectival agreement patterns in French; as the two operate very similarly (Riegel et al. 2011, p. 599), the primary focus of this chapter will be how non-binary gender is marked on nouns. This necessary background is followed by an overview of the paper’s methodologies. Data is presented from a combination of a corpus of non-binary grammars and sociolinguistic interviews. This is followed by an analysis of the three major approaches to forming non-binary nouns as evidenced by the study data. Three approaches to agreement patterns in non-binary French are identified and discussed: A Compounding Approach, which combines masculine and feminine markers; an Invariable Approach, which utilizes one invariable non-binary gender marker; and a Systematic Approach, which includes the creation of several non-binary morphemes that attach to bases on the basis of their phonological properties. Three distinct Systematic Approaches to forming non-binary nouns are identified and discussed. This paper concludes with a cross-comparison of all three approaches, highlighting the fact that while non-binary French is diverse in its approaches to (un-)marking gender, the emergence of patterns within these marking systems allows for better cross-linguistic analysis, as well as comparison with historical language change.

Nominal Gender in Standard French

Within standard French, nouns are masculine or feminine. It should be noted that this paper focuses on animate nouns, for which gender surfaces in different ways than in the majority of the French language; the peculiarities of French gender and animate nouns will be discussed momentarily. The fact that more than 93% of nouns referring to humans differ in grammatical gender (Labrosse 1996, p. 43) underscores the challenges that non-binary Francophones face in using the language without misgendering themselves or others. In French, gender works at the lexical, morphological, and syntactic level (Abbou 2011; Michard 2002). At the lexical level, the entire word is semantically gendered (Riegel et al. 2011, p. 330), as is the case with kinship terms:

(1)	sœur ‘sister’
	frère ‘brother’
(2)	homme ‘man’
	femme ‘woman’

In some cases, semantically distinct masculine and feminine nouns may be identical in form, despite their different underlying genders:

(3)	[la] concierge ‘concierge [fem.]’
	[le] concierge ‘concierge [masc.]’

As is evidenced by (3), this results in some nouns appearing the same but being distinguished by their modifiers—in this case, feminine versus masculine determiners (adapted from (Riegel et al. 2011, p. 330)).

Gender may also be marked morphologically via the use of masculine or feminine affixes, although this is much more common on animate nouns than other types of nouns. We see this in the case of the masculine suffix –teur and the feminine suffix –teuse:

(4)	chanteuse ‘singer [fem]’
	chanteur ‘singer [masc]’

In the case of morphological gender-marking, it is much more common for the feminine to be formed by the addition of a feminine suffix (e.g., maitresse ‘mistress’ < maitre ‘master’) than for the reverse to occur (e.g., compagnon ‘comrade [masc.]’ < compagne ‘companion [fem.]’); partly due to this normative derivational direction, allomorphy may be gender-imbalanced as in the case of the masculine variant –[t]eur having four possible feminine counterparts: –euse, –eure, –eresse, and –trice (Riegel et al. 2011, pp. 330–31).

Moreover, because of phonological change in the language and the tendency to retain etymological spelling, gender surfaces in different ways in the oral code versus the written code. In written French, nouns with morphological gender marking tend to have four written forms: The masculine singular, the feminine singular, the masculine plural, and the feminine plural (adapted from (Riegel et al. 2011, p. 603)). The written–e suffix canonically marks the feminine on adjectives and animate nouns (although see above for the many exceptions)3, while the –s suffix canonically marks the plural (although see McLaughlin forthcoming: 2–3 for exceptions). An example of the prototypical four-way distinction of the written forms is given in

Table 1. Gender and Number Distinctions in Written and Oral French.

	Masculine	Feminine
Singular	/sypstity/ substitut ‘substitute’	/sypstityt/ substitut–e ‘substitute’
Plural	/sypstity/ substitut–s ‘substitute’	/sypstityt/ substitut–e–s ‘substitute’

below:

However, gender distinctions are different in the oral code. As illustrated by Table 1, the oral code typically only has two forms (Riegel et al. 2011, p. 603) that are distinguished in gender (but not number)4 by the pronunciation of the final consonant, in this case, [t]. In many cases, there is no oral distinction between masculine and feminine variants (McLaughlin forthcoming: 3–4; Riegel et al. 2011, p. 330).5 These include most nouns ending in vowels, as well as many ending in –l/le, –r/re, –c/que, –s/(s)se, –t/te, and –x/xe6, as illustrated in

Table 2. Masculine and Feminine Homophones.

Feminine			Masculine
IPA	Written	Gloss	IPA	Written	Gloss
/libeʁαl/	libérale	‘liberal’	/libeʁαl/	libéral	‘liberal’
/dɔktœʁ/	docteure	‘doctor’	/dɔktœʁ/	docteur	‘doctor’
/ɛspaɲɔl/	espagnole	‘Spanish’	/ɛspaɲɔl/	espagnol	‘Spanish’
/gʁɛk/	grecque	‘Greek’	/gʁɛk/	grec	‘Greek’
/aʁpɛt/	arpette	‘apprentice’	/aʁpɛt/	arpète	‘apprentice’
/ely/	élue	‘elect’	/ely/	élu	‘elect’
/ami/	amie	‘friend’	/ami/	ami	‘friend’

:

Modifiers, including adjectives, match the gender of a given noun—although some adjectives are epicene and may modify either masculine or feminine nouns. Epicene adjectives most commonly have one form ending in –e (e.g., simple, magique ‘magical,’ rouge ‘red’).

Syntactic gender-marking occurs when a given noun is not gendered per se, but has gender marked somewhere else in the sentence. A small number of animate French nouns are epicene, or common gender nouns, meaning that they can refer to either masculine or feminine referents (e.g., adulte ‘adult’). As such, epicene nouns are not neutral per se but rather flexible in their reference (Dawes 2003). However, some members of the non-binary community have interpreted epicene nouns as neutral, while still others have advocated for their usage as one possible means of referring to non-binary referents without innovating neologisms (Ashley 2019; Swamy 2019). Epicene nouns may be functionally-gendered; for example, as part of its proposed reforms designed to visibilize women in the French language, the Roudy Commission7 recommended feminizing epicene nouns by using them alongside feminine determiners (e.g., la professeur, ‘the professor [fem.]’8) (Fleischman 1997). Non-epicene adjectives may also be used to functionally gender epicene nouns (e.g., beau prof ‘handsome professor,’ versus belle prof ‘beautiful professor’).

Because of the differences in the number and kinds of gender distinctions in the written versus the oral code, there have been different kinds of gender-motivated reforms in the two codes. In written inclusive French (écriture inclusive) writers are able to make use of unpronouncable orthography in order to disrupt the grammatical binary, with the most common strategies using punctuation or capitalization to simultaneously include masculine and feminine morphology (Abbou 2011; Burnett and Pozniak 2020; Manesse and Siouffi 2019). This writing system results in forms such as those represented in (5a–c):

(5)	a. aviateur·ice ‘aviator·ix’
	b. aviateur.ice ‘aviator.ix’
	c. aviateur/ice ‘aviator/ix’

Although the orthography of inclusive writing is itself unpronounceable (Knisely 2020)—for example, one would not oralize the form docteur·e ‘doctor’ as ‘docteur-point-e’—there is a coalescence around how particular forms should be pronounced. In the case of the form docteur·e, the expected pronunciation would be a balanced form with the masculine and feminine variants joined by a conjunction, such as docteur et docteure ‘doctor (masc.) and doctor (fem.)’ (Gunther 2022, p. 51; Les Salopettes 2017). Notably, however, these forms are homophonous: Pronouncing both variants is a pragmatic move designed to highlight the inclusion of feminine referents, even though the masculine and feminine oralizations are phonologically indistinct.

Attempts to create gender-inclusive oral forms have had to contend with the peculiarities of gender distinctions in oral French. Earlier linguistic analyses characterized French as having a subtractive system (Bloomfield 1933, p. 217), whereby the masculine is derived by dropping the final consonant of a given form, such that:

(6)

C → Ø/__#

However, problems with the subtractive model have been pointed out by a number of linguists (see Manova 2019 for an overview). Labrosse (1996, p. 44) argues that, from a morphological standpoint, grammarians have traditionally argued for the masculine as the base for most nouns and adjectives in French, which she refutes, arguing instead for the interpretation that “aucune forme, féminine ou masculine, ne dérive de l’autre [no form, feminine or masculine, derives from another].” There are additional problems in characterizing French as possessing a predominantly morphological gender system, as have been highlighted by Riegel et al. (2011, pp. 604–5). The peculiarities of the French gender system—and especially the relative rarity of explicit nominal gender-marking—have led to the argument that in French, gender is marked at the level of the noun phrase for the majority of French nouns, especially inanimate ones. Gender-marking most frequently surfaces on modifiers (Riegel et al. 2011, pp. 274, 603), especially ones that are both orally and orthographically distinct (McLaughlin forthcoming, p. 3).

Michel (2016) argues that, because social representations associated with men and women are bundled into our perceptions of masculine and feminine morphology on animate nouns, we must create grammatical models that work to visibilize these social associations. Michel’s major contribution in this regard is the creation of a lexicographic model analyzing masculine and feminine variants of gendered words by comparing them based on their shared elements, rather than analyzing feminine morphology as affixed onto a masculine base:

(7)	Con $\begin{array}{c}–Ø \left(m.\right)\\ –ne \left(f.\right)\\ \end{array}$ ‘stupid’

Alpheratz (2018, pp. 79–81) adopts and expands upon Michel. In their grammar, bases are identified as the largest chunk that masculine and feminine variants have in common once all written “marqueur[s] de genre [gender markers]” (80) have been separated out. This leads to re-analyses such as the one present in (8):

(8)	C $\begin{array}{c}–on \left(m.\right)\\ –onne \left(f.\right)\\ \end{array}$

Notably, this analysis of French morphology is non-standard, as it presupposes that –on is a masculine marker, and –onne is a feminine marker. A more standard approach would view only <ne> on the feminine variant as morphological components, as in writing a doubled consonant and –e typically mark the feminine and correspond to a pronounced final consonant in the oral code (Riegel et al. 2011, p. 330). Most of the components that have been separated out in (8) (i.e., <on>) are not traditionally viewed as gender morphology at all, but rather parts of the bases of the masculine and feminine forms. The motivation behind Alpheratz’s radical reanalysis in (8) is to create a system where “masculin, féminin et neutre se trouvent dans la même classe, aucun ne dérivent de l’autre, et sans que le masculin ne soit plus intégré à la base [masculine, feminine, and neutral find themselves within the same class, none deriving from the other, and without the masculine being integrated into the base any longer]” (Alpheratz 2018, p. 80), an approach which they acknowledge “peut perturber les lexicographes [might trouble lexicographers]” (80). This reanalysis will become important for discussing Alpheratz’s Systematic Approach to non-binary French in Section 3.3.2.

2. Materials and Methods

This study cross-compares data from two sources: a corpus of non-binary French, which is supplemented by data on the usage of non-binary French gathered from sociolinguistic interviews. This study takes a mixed-methods approach: comparative analysis of corpus data is socially contextualized9 through analysis of sociolinguistic interviews, which elicited both non-binary forms and metalinguistic analysis of these forms.

Corpus Analysis: Only comprehensive grammars that describe methods for incorporating non-binary gender into oral and written French are included. In practice, this excludes word lists of non-binary forms, instructional works focusing on written-only strategies such as the graphies tronquées10 (which are punctuation-based), and survey analyses of usage that list the most popular forms but do not provide a consistent method for using them (i.e., when a particular morpheme is conditioned in the written code, or whether a given form is orally distinct from its masculine and feminine counterparts).11 Three grammars are analyzed here: the AN-EX system (de Villeneuve and Gheeraert 2018), the system developed by Alpheratz (2018), and a modification of Alpheratz’s system proposed by Ashley (2019).12 It is notable that among the four writers of the three systems analyzed here, only Alpheratz is a trained linguist—Ashley is a bioethicist, while de Villeneuve is a mathematician and Gheeraert is a chemist. This is not to dismiss the validity or the importance of their work, but rather to explain some of the terminological and technical gaps that become apparent when analyzing these authors’ works in-depth. Further, while Ashley’s grammar has been published as part of a larger article discussing strategies for referencing non-binary individuals in French, and Alpheratz’s grammar is in print, de Villeneuve and Gheeraert’s system has been presented but not peer-reviewed. While the varying aims and backgrounds of these authors leads to varying levels of explicit accounting for linguistic issues (e.g., phonological constraints, morphological paradigms), it is important to clarify that grammars produced by non-linguists fill an important gap. In many Francophone contexts conservative attitudes toward gender-motivated linguistic innovation have left the attention given to non-binary linguistic issues outside of the LGBT+ community almost non-existent. This is especially true in France, where such conservative attitudes predominate even among those who were pivotal in the feminization movement of the 20th century, such as Bernard Cerquilini (Kieffer and Joubin 2017; Mackenzie 2019).

Interviews: data from the corpus is supplemented with data on non-binary gender-marking strategies collected during one-on-one semi-structured sociolinguistic interviews with six self-identified non-binary Francophones living in Montréal in summer 2018. Five of these participants were local to Montréal, while one (Bardot) was a Parisian student studying abroad in Montréal. Participants were recruited via outreach to Montréal-based LGBTQ+ organizations, including flyer distribution and social media posts on organizations’ pages. All interviewees were between the ages of 18–24, White, and L1 Francophones. Although they were not specifically recruited for their bilingualism, all participants demonstrated differing degrees of English fluency. Participants were orally instructed to speak in whatever language (French or English) they felt most comfortable in; however, likely due to a combination of my status as an L1 English and L3 French speaker, as well as the common perception that English has more linguistic resources to discuss gender diversity (See Kaplan 2022a), interviews were largely conducted in English; excerpts in French, which predominantly occurred as either intra-sentential or token code-switches (Poplack 1980), are italicized. The interviews were designed for two purposes: Firstly, to collect data on the morphology of non-binary French, and secondly, to collect data on language attitudes toward non-binary French. Each interview was divided into two parts, with the first focused on participants’ self-reports of their preferred systems of non-binary French, as well as metalinguistic analyses underlying these preferences. The second part of the interview focused on participants’ usage patterns of non-binary French, including their self-reported usage in educational, workplace, and personal contexts, as well as their own external encounters with both written and oral non-binary French forms (e.g., in reference works, media, policy, etc.). In this latter part of the interview, many participants cited particular, written grammars of non-binary French as reference points for their own usage; how these works were cited is discussed in my analysis. Interview data contextualizes both the strategies that non-binary Francophones use in constructing non-binary nouns, and also provide a general idea of which strategies are used most frequently.

3. Results

This paper focuses specifically on non-binary nouns as these are not only sites of significant discrepancies in strategy, but also where the large variety of gendered endings in standard French provides unique challenges to innovators.13 Animate nouns with morphological gender are the primary focus of this paper; lexically-gendered nouns will be discussed briefly but are not the major focus here.

As noted previously, much of the work on non-binary agreement strategies has focused on exclusively written forms, such as those found in l’écriture inclusive, which are not pronounceable in and of themselves.14 Those circumnavigating the grammatical gender binary in their speech are forced to implement pronounceable variants; among non-binary speakers, this has led to the innovation of non-binary (e.g., Ashley 2019) or neutral (Alpheratz 2018, p. 41) neo-markers. Although some guides (La Vie en Queer 2018a, 2018b; Divergenres 2021) make a distinction between the function of inclusive (including all genders) versus neutral (making no reference to gender) markers, in the majority of non–binary grammars no functional distinction is made—the same marker may be used to refer to non–binary individuals, agender individuals, groups of mixed–gender individuals (also known as common gender), gender-unidentified persons, and in impersonal constructions.15 For the purposes of this paper, when describing morphology I use ‘neutral’ and ‘non-binary’ interchangeably.

Because many non-binary French speakers seek varieties that explicitly highlight their non-binary identities, many of the grammars of non-binary French take special care to highlight forms that are distinct from both masculine and feminine variants. In this way, their efforts parallel the reforms of 20th century feminists (e.g., Labrosse 1996) who have argued that the mere existence of non-masculine variants is not enough, and that minoritized subjects are best visibilized in French when they can be hailed using terms that will not be confused with masculine variants in either the written or oral codes. However, not every form proposed in every grammar is successful in being orally distinct from its existing binary counterparts, a critique that has also been levelled at the non-binary pronoun iel (Greco 2019; Knisely 2020, p. 862).

In the following three sections, I discuss three strategies for forming non-binary nouns as attested in the corpus and interview data, which I have labelled the Compounding Approach, the Invariable Approach, and the Systematic Approach.

3.1. The Compounding Approach

The Compounding Approach combines masculine and feminine suffixes in either order. It can only be used to form non-binary variants where the masculine and feminine variants are (1) not homophones, and (2) contain phonologically distinct masculine and feminine morphemes. In practice, this leaves few nouns that can be rendered non-binary under the Compounding method. Within the corpus assembled here, this comprises adjectives ending in –if and –ive and nouns with the masculine ending –eur, which can be compounded with two of its feminine allomorphs, –rice and –euse.16 These yield non-binary forms such as acteurice or actriceur, ‘actor,’ and joueureuse or joueuseur, ‘player.’

The Compounding Approach has been referred to by other names in resources on non-binary French. Ashley (2019, p. 4) cites it as a “hybrid” approach, Alpheratz (2018, p. 230) refers to it as “double-flexion [double-marking],” Elmiger (2015, p. 9) uses “contamination,” Burnett and Pozniak (2020) use the phrase “long forms,” while de Villeneuve and Gheeraert (2018, p. 45) refer to it as concatenation, a term which some interviewees also used. I use the term Compounded here as it is more specific than terms such as “hybrid” and “long forms.” It also avoids the multiple meanings that can be conveyed by contamination, which includes blends (Elmiger 2015, p. 9), as well as the multiple meanings of double-marking, which can also refer to syntactically-inclusive phrases that include both masculine and feminine referents joined by a conjunction (e.g., éducateurs et éducatrices, ‘educators (masc.) and educators (fem.)’, Alpheratz 2018, p. 224). In their (La Vie en Queer 2018b) poll of non-binary Francophones, La Vie en Queer found that compounded variants vary in popularity in comparison with other non-binary alternatives. For example, their results found créateurice (‘creator,’ n.) to be the most popular form (76.7%) of the set –eur (m.)/–rice (f.) (with the alternatives being non-compounded neomorphemes), while the compounded variant –ifive (n.) was only the second most popular variant (21.2%) for the set –if(m.)/–ive (f.). The reasons underlying this variation are unclear.

Notably, there is a divide in the interpretation of what makes masculine and feminine variants phonologically distinct enough to compound. For example, while de Villeneuve and Gheeraert (2018, pp. 44–45) accept the forms acteurice and actriceur, they recommend using the systematic non-binary suffix–xe in cases where –eureuse or –euseur would be possible, thus yielding forms such as “danseuxe” from danseur, danseuse ‘dancer’ (de Villeneuve and Gheeraert 2018, p. 44; See Section 3.3 for a full explanation of the uses of –xe).

While the mechanics of this approach make it theoretically plausible for any two phonologically distinct masculine and feminine morphemes to be compounded, in my corpus I only record combinations of –eur/rice and–eur/euse, but not –eur and –eresse. All attested combinations and possible combinations of –eur and its feminine counterparts are presented in (9a–d):

(9)	a. créateurice ‘creator’
	b. créatriceur ‘creator’
	c. joueureuse ‘player’
	d. joueuseur ’player
	e. ?demandeureresse ‘applicant’
	f. ?demanderesseur ‘applicant’

The lack of recorded uses of compounded ––eur/–eresse is likely the result of a combination of phonological factors and the influence of historical gender-involved reforms in French. Perhaps the fact that dyads of the type –eur/–eresse are no longer productive in French (Riegel et al. 2011, p. 330) explains the absence of their compounded variant in the corpus.

For further explanations on the idiosyncrasies of the Compounding Approach and its apparent exclusions, we must return to discussing historic gender-involved linguistic reforms in French, with which the Compounding Approach also has strong parallels. The Compounding Approach may have originated as the oralization of forms that have become increasingly common alongside raised awareness of inclusive writing. For example, the guide published by the Haut Conseil à l’Égalité entre les femmes et les hommes (2016) recommends writing the masculine word form followed by the feminine suffix, with punctuation in-between (in this case, a period). This yields forms such as travailleur.euse ‘worker,’ ambassadeur.rice ‘ambassador,’ and acheteur.euse ‘buyer’ (HCE 2016, p. 60), all of which look like the compounded non-binary forms described above. Influence from inclusive written forms would also explain the absence of forms compounding –eur/–eure, which are blocked in inclusive writing by the construction <eur·e>, as in auteur·e (Haddad 2016).

However, as I explained in the introduction, normative pronunciation of inclusive written forms would include both the masculine and the feminine variant separated by a conjugation (e.g., travailleur.euse is pronounced travailleur et travailleuse, ‘worker (masc.) and worker (fem.)’). It appears that the Compounding Approach may have emerged as an oral shortcut influenced by French inclusive writing norms. What problematizes this hypothesis, however, is that some forms that are double-marked in the inclusive writing system are not recorded as oralizable. For example, while the HCE (2016, p. 60) includes double-marked forms for the suffix set –er[je]/–ère[jɛʁ], such as premier.ère ‘first,’ I can find no evidence of an oral non-binary form *premierère or perhaps *premierière [pʁɛmjejɛʁ]. This is perhaps due to the lower frequency of the written form <–er.ère> as opposed to <–er.e>, which the HCE acknowledges in a footnote noting the acceptability of both premier.e and premier.ère (2016, p. 60). It is in my opinion, however, that it is more likely that there is some phonotactic constraint preventing the surfacing of some double-inflected forms which remains to be explored in a future study.

There is additionally one recorded case where a feminine derived from the masculine via a vowel change is compounded. Within the corpus, we only see this with the non-binary ending –ifive17:

(10)	actifive ‘active’

The precise nature of the Compounding Approach in French may well be unique. French nominal morphology noticeably differs from other Romance languages in that it does not contain canonical masculine and feminine suffixes in the oral code (c.f. masculine –o and feminine –a in Spanish), but rather prototypically forms the feminine from the masculine by either adding a final consonant or changing the final consonant of the masculine form (Riegel et al. 2011, p. 608). What is especially intriguing is the lack of compounded non-binary or neutral forms across languages beyond French. The most analogous examples from other languages are blends, rather than compounds. For example, within guides to non-binary Portuguese, Cassiano (2020) recommends the form príncise as a neutral alternative to príncipe ‘prince’ and princesa ‘princess,’ while Caê Almeida (2020, p. 17) recommends pãe as a substitute for mãe ‘mother’ and pai ‘father.’ In fact, blends of lexically-gendered words are common cross-linguistically. As an example of the phenomenon occurring outside Romance, the German non-binary wiki page for “Nichtbinäre Wörter [Non-binary Words]” (Nichtbinär-Wiki 2022) cites both Mapa and Pama as acceptable alternatives to Papa ‘papa’ and Mama ‘mama,’ alongside other blends such as Tankel, Onkente, Tonke or Onte in place of Tante ‘aunt’ and Onkel ‘uncle.’ There are analogous blended forms occurring in non-binary French as well, such as tancle (Ashley 2019; Divergenres 2021, p. 11; La Vie en Queer 2018a) or tonte (de Villeneuve and Gheeraert 2018, p. 60) from tante ‘aunt’ and oncle ‘uncle,’ and sère (Ashley 2019; Comme poussent les pissenlits 2015) or frœr ‘sibling’ from frère ‘brother’ and sœur ‘sister’ (Ashley 2019; Divergenres 2021, p. 11; La Vie en Queer 2018a; Comme poussent les pissenlits 2015). Cross-linguistically, blends occur almost unilaterally on kinship terms that are gendered at the semantic, rather than morphological, level.

It is in fact notable that compounded forms are absent from non-binary manuals across other languages. With its variety of gender morphemes on animate nouns, it would be possible to form compounded words in Portuguese, for example. This is most apparent in a word pair like ator ‘actor’ and atriz ‘actress,’ which is cognate with the French pair acteur, actrice. While the masculine and feminine Portuguese endings could combine to form *atoriz or *atrizor, forms like these do not surface in any of the non-binary Portuguese manuals I am familiar with. Alternative forms are suggested instead (e.g., atore ‘actor [neut.],’ Cassiano (2020), p. 45; ater ‘actor [neut]’ Bem-vinde (2021)). However, it is difficult to speculate on the absence of a phenomenon, thus making the lack of compounded forms in other languages beyond the scope of this paper.

The Compounding Approach in Use

The Compounding Approach is a common strategy among non-binary Francophones in Montréal. Noe, who works as a preschool teacher, uses a compounded variant to describe their own occupation:

Noe:

I was working as an educator, and we say éducatrice [e.dʒy.ka.tʁis] and éducateur [e.dʒy.ka.tœʁ], I was saying éducateurice [e.dʒy.ka.tœʁ.is].

Note that in Noe’s usage the first [ʁ] of –rice is dropped after the final [ʁ] of –teur; the realization of one or both rhotics in the compounded set varies among interviewees. Noe uses this agreement strategy in juxtaposition with other agreement strategies common among non-binary Francophones, such as switching between masculine and feminine agreement patterns (Ashley 2019):

Noe:

Um, [I use the] masculine, sometimes I use feminine too, but like, I don’t really want to do it, and I try to invent words sometimes, or use words that I already heard that do not exist in the dictionary already[...] Um, like, belle [bɛl], I’d say beaulle [bol], yes, like beau [bo] and belle [bɛl].

However, as Noe clarifies, switching is a dispreferred strategy for them, a finding echoed in Knisely’s (2020) survey data on non-binary Francophones. Rather, Noe prefers to use neologisms, such as beaulle, ‘attractive’ which is a blend of beau ‘handsome’ and belle ‘beautiful,’ a form which is morphologically more complex than the majority of words formed using this pattern, as the dyad represents a case of weak suppletion (Battye et al. 2001, p. 123; Mel’cuk 1994, p. 429), making both the masculine and the feminine variants impossible to morphologically decompose. It is especially interesting to note how Noe accounts for this novel form: In their own words, they have “already heard” it being used by other non-binary folks, but it does “not exist in the dictionary already.” Although in their reference to the fact that these forms “do not exist in the dictionary already” Noe actively acknowledges that they are using a form not (yet) incorporated into standard French, they latently highlight the emphasis on dictionary precedent and historical usage common to discourse on the French language (Walsh 2016), including in former French colonies such as Montréal (Walsh 2014). Noe emphasized this disconnect between their preference for neo-forms and their non-canonicity when they recounted their mother’s shocked reaction to Noe explaining how they would like to be referred in non-binary French: “’I looked in the dictionary, it doesn’t even exist!’”

One participant who was particularly enthusiastic about the Compounding Approach was Bardot, a visiting student from France. Throughout our interview, they referred to the Compounding Approach as “concatenation,” and gave examples of how one could use it:

Bardot:

For example, utilisateur [y.tɪ.li.za.tœʁ] and utilisatrice [y.tɪ.li.za.tʁɪs], so a user. What you could do is just concatenate the two endings in the sense that you say utilisateurice [y.tɪ.li.za.tœʁ.ʁis]. Like you first take the masculine form and you just add the feminine ending… So that’s the second way, um, that we go, the concatenation.

Bardot has spent a great deal of time studying non-binary French forms and was therefore more eager to discuss word construction than most other participants. Bardot’s metalinguistic analysis of the Compounding Approach is interesting for two reasons. First, they perceive the forms produced by this approach as morphological compounds of masculine and feminine suffixes that, taken together, can be used to refer to non-binary individuals. Secondly, they perceive the masculine form as underlying, since in their interpretation, the feminine form is added to the masculine base— an analysis highlighted by their articulation of the initial [ʁ] of –rice [ʁis] in the syllable following the final [ʁ] of -teur [tœʁ], indicating that these suffixes remain semantically transparent in the compounded form. The interpretation of the masculine form as the base in Bardot’s speech contradicts other instances of the Compounding Approach, where (1) the ordering of masculine and feminine morphemes is flexible, and (2) the root of an individual noun is analyzed as the base.

As I have highlighted in my analyses, this approach is limited in scope. Bardot is well aware of cases in which the Compounding Approach cannot be used due to phonological constraints:

Bardot:

It wouldn’t work for a lot of words. For example, content, contente, you just add an <e> at the end so you can’t really say—it doesn’t work. You can’t add the feminine ending that, that’s already how the word, uh, works.

Here, Bardot speaks to the fact that, in the oral code, many French nouns and adjectives have a consonantal suffix in the feminine, but no corresponding masculine suffix (McLaughlin forthcoming, p. 3). Bardot’s instinct that these words can’t just have a “feminine ending” added to them is a sound linguistic analysis: Absent distinct masculine and feminine suffixes, the Compounding Approach doesn’t work.

Bardot also addressed the similarities between the processes for creating non-binary nouns and adjectives using the Compounding Approach:

Bardot:

We should make the difference between grammatical words, pronouns, articles, and all that, versus the lexic[al] [words]...Um, and so these rules work for nouns and adjectives which are actually the only words that are gendered, but when it comes to grammatical words—pronouns and all that—you have to have like just a bunch of non-binary versions […] and you’re good to go with that. As in, [non-binary pronoun] iel for example.

JK:

OK. So like new words.

Bardot:

Yeah. These are the new words and we know who [sic] they are and we don’t have to add anything because there is a finite number of grammatical words.

According to Bardot’s perspective, broadly speaking, non-binary French grammatical systems address content words, such as nouns and adjectives, differently from function words such as pronouns and articles. When Bardot characterizes “nouns and adjectives” as “the only words that are gendered,” it is likely that they are referring to the fact that nouns and adjectives describing humans are more likely to have word-final features characteristic of either masculine or feminine gender (e.g., morphological gender markers, the pronunciation of a word-final consonant on many feminine nouns and adjectives, etc.). However, these words are not the only words that are gendered in French: Pronouns, participles, and articles may also have grammatical gender (Alpheratz 2018, p. 26). Nonetheless, Bardot’s intuition about nouns and adjectives as having specific “rules” points to their understanding of the various phonological and morphological processes that often distinguish masculine and feminine nouns and adjectives in the oral and—to an even greater extent—within the written codes.

Bardot’s intuition that “grammatical words, pronouns, articles” don’t operate along the same “rules” as “nouns and adjectives” also points to an interesting metalinguistic analysis of how gender operates in French. When they speak to the necessity of how “you have to have like just a bunch of non-binary versions” of pronouns and articles, they are getting at the well-known fact that function words are typically closed-class. Whereas new content words such as nouns and adjectives are being added to the French language all the time—via borrowings, conversion from other parts of speech, neologisms or any number of word-formation processes—it is generally much harder for new pronouns and articles to be adopted into a language. This comparative difficulty of pronoun innovation and adoption is what leads Bardot to conclude that “there is a finite number of grammatical words.”

3.2. The Invariable Approach

The Invariable Approach is the most straightforward method of creating non-binary variants of French nouns. As the name implies, a single non-binary suffix is added after the final pronounced vowel in the masculine form of a word; if the word ends in a consonant, this consonant is deleted before the non-binary suffix. In this way, the Invariable Approach analyzes the masculine form as the base. Thus, in an approach that takes –z as an invariable non-binary morpheme, we see forms such as (8) and (9):

(11)	acteur/aktœʁ/18 ‘actor’
	actrice/aktʁis/‘actress’
	acteuz/aktœz/‘actor [neutral]’
(12)	ami/ami/‘friend [masc]’
	amie/ami/‘friend [fem]’
	amiz/amiz/‘friend [neutral]’

Alpheratz (2017) notes that –z, –x, and –æ are “des morphèmes de genre neutre [neutral gender morphemes],” in that they commonly have the semantic quality of neutrality in non-binary grammars. Other invariable morphemes have included –m (La Vie en Queer 2017, 2018a; Divergenres 2021), –s [s] (Divergenres 2021; La Vie en Queer 2017, 2018a), –u (La Vie en Queer 2017) and –t (Knisely 2020; La Vie en Queer 2017), although La Vie en Queer notes that invariable –m has failed to become popular (La Vie en Queer 2017, 2018a). Similarly, many non-binary Francophones are unfamiliar with –t (Knisely 2020).

The Invariable Approach has its precedence in the work of Causse (1998), who, citing Monique Wittig as an influence, developed a radical “Alphalecte” in order to supersede not only the link between grammatical gender and human gender in the French language, but also challenge the power structures embedded in both concepts.19 This latter goal becomes clear when Causse describes Standard French as an “androlecte,” which encodes “l’institutionnalisation d’une subjectivité (d’un corps-pensée) sexuée au masculine [the institutionalization of a subjectivity (of a thinking body) sexed in the masculine]” (2). The alphalecte, by contrast, is “débarrassé des genres [cleared of genders]” (5). It marks “articles et adjectifs” invariably (7) with the suffix – α, an “entité typographique, un sigle, qui fait signe à toutes les personnes [typographic entity, an initial, that gestures toward all persons]” (9). As the reference to typography indicates, the alpha is an orthographic symbol; much like the graphies tronquées of inclusive writing, it cannot be pronounced. Thus, Causse’s proposal operates as a philosophical written precursor to the Invariable Approach, as opposed to a speakable one.

The Invariable Approach in Use

Invariable forms were not widely discussed by speakers. However, as the interview pool was small, I cannot extrapolate on how this reflects usage among non-binary Francophones. One speaker commented on their friend’s use of invariable –z:

Noe:

I have a friend, he makes the ending of words with ‘z,’ yeah...Uh, just ‘z’ instead of adding an ‘e’ or ‘x’ or something.

Noe noted that the –z is pronounced/z/; in French varieties that insert word-final schwa (Riegel et al. 2011, p. 176), we may expect the pronunciation to be [zə] allophonically. More data on when, where, how, and by whom invariable suffixes are used will be explored in a future study.

3.3. The Systematic Approach20

The grammars analyzed in this section here share several important commonalities. Firstly, the authors describe their systems as either already popular among non-binary Francophones (de Villeneuve and Gheeraert 2018), or a more formal extension of agreement patterns they have heard used communally (Alpheratz 2018; Ashley 2019). Further, none of the grammars analyzed here lay claim to their recommendations being the only way to create non-binary French forms. This is important, as it demonstrates that the authors recording these systems are aware of variation in non-binary French. In fact, Alpheratz (2018) is quite explicit in presenting their system as one option among many alternatives. Their legitimacy as community reference sources became apparent during my interviews—all three systems were referenced by at least one participant, with Alpheratz’s system coming up several times.

As alluded to in my earlier discussion of the different methods of analyzing nominal and adjectival bases, complications arise regarding how individual authors choose to analyze non-binary French. For example, some analyze forms such as confus (m.) and confuse (f.), ‘confused’ and substitut (m.) and substitute (f.) ‘substitute’ as being part of the same morphological pattern, with masculine unmarked (Ø) and feminine marked with the morpheme –e (although this only holds true in written French); however, as discussed in the introduction, other authors (notably Alpheratz (2018) analyze these forms as instances of two different patterns associated with underlyingly different morphological paradigms, with –us marking the masculine and –use marking the feminine in confus/confuse, while –ut signals the masculine and –ute the feminine in subsitut/substitute. Alpheratz justifies this analysis by appealing to the idea that “le genre masculin est tout aussi ‘marqué que le genre féminin [the masculine gender is just as marked as the feminine]” as in cases where either the masculine differs from the base, or grammatical gender surfaces lexically (2018, p. 81).

The following sections describe three major Systematic Approaches to forming non-binary nouns, beginning with the AN-EX system—which has the smallest set of non-binary allomorphs of the three, followed by Alpheratz’s (2018) system, and finally Ashley’s (2019) addenda to Alpheratz’s system.

3.3.1. The AN-EX System

The AN-EX System is described by de Villeneuve and Gheeraert (2018), who are both based in Paris. According to the authors, the system is influenced by approaches taken within non-binary communities, especially the blogs La Vie en Queer (2018b) and Divergenres. This system employs three non-binary endings on singular nouns and adjectives: –x, —xe, and—an. These endings are conditioned phonologically.

If the masculine and feminine variants are homophones, the non-binary variant is formed by adding –x to the masculine form, which is left unpronounced (“x muet” (42)):

(13)	député/depyte/‘deputy [masc]’
	députée/depyte/‘deputy [fem]’
	députéx/depyte/‘deputy [neutral]’

If the feminine variant ends in [s] or [z], which is orthographically realized as —se, —sse, or —ce, as in joueuse ‘player,’ duchesse ‘duchess’ or actrice ‘actress,’ the non-binary form is created by adding dropping the —se, —sse, or —ce ending, and replacing it with —xe [kse].

(14)	prince/pʁɛ̃s/‘prince’
	princesse/pʁɛ̃sɛs/‘princesse’
	princexe/pʁɛ̃sɛkse/‘monarch’s child’

If the masculine variant is more than two syllables and ends in <n>, which the authors interpret as underlyingly/n/, encompassing the forms —en, –in, –on, and –un, then the non-binary form replaces the –Vn ending with —an. Although the authors denote this as underlyingly/αn/(de Villeneuve and Gheeraert 2018, p. 44), they also note its homophony with an, ‘year,’ which is realized as [ᾶ].

(15)	Canadien/kanadjɛ̃/‘Canadian [masc]’
	Canadienne/kanadjɛn/‘Canadian [fem]’
	Canadian/kanadjα21/‘Canadian [neutral]’

The authors provide an exception for words ending in <n> that are only one or two syllables long, which they propose take the –xe [kse] ending:

(16)	chien ‘dog [masc]’
	chienne ‘dog [fem]’
	chienxe ‘dog [neutral]’

In cases where the written binary pattern does not conform to feminine = masculine + suffix (i.e., –e), the masculine and feminine suffixes should be combined. Their grammar cites only examples of forms –eur (m.) and –rice (f.), as in the dyad acteur, actrice ‘actor’ (17). However, as discussed regarding the Compounding Approach, there are a number of feminine allomorphs that do not resemble their masculine counterparts (e.g., –if (m.), –ive (f.)); see

Table 3. Neutral Singular endings22 in the AN-EX System (de Villeneuve and Gheeraert 2018).

Phonologically-conditioned Forms	Where the masculine and feminine are homophones	Orthography: The masculine base takes a silent –x [Ø] Oral: Homophonous with masculine and feminine forms	ami·e → amix député·e → députéx libéral·e → libéralx
	Where the feminine ends in [s] or [z]	Orthography The ending —se, —sse, or —ce is dropped, and replaced with –xe/kse/. Oral: Non-binary forms are distinct from both masculine and feminine forms.	danseur, danseuse→ danseuxe anglais·e → anglaixe
	Where the masculine ends in [n] or a nasal vowel	Orthography: Words were the masculine ends in –n (–en, –in, –on, –un) replace –Vn with –an/αn/ Exception: Words ending in <n> that are only two syllables long add–xe/kse/to the masculine Oral: Non-binary forms are distinct from both masculine and feminine forms. However, masculine forms already ending in –an are not specifically addressed and would presumably follow the rules for “all other cases.”	cousin·e → cousan musicien·ne → musician
	Where the feminine and the masculine have orally distinct suffixes	Orthography: Masculine and feminine suffixes are compounded. Exception: The set –eur (m.)/–euse (f.) follow the rules for nouns with feminine forms ending in [s] or [z]. Oral: Non-binary forms are distinct from both masculine and feminine forms	acteur, actrice → acteurice
	All other cases	Orthography: –xe/kse/is added to the end of the masculine form Oral: Non-binary forms are distinct from both masculine and feminine forms	avocat·e → avocatxe

for more examples.

(17)	Asia Kate Dillon	est	an	acteurice	dans	Billion
	Asia Kate Dillon	is	INDF.N	actor.MF	in	Billions
	‘Asia Kate Dillon is an actor in Billions.’

All other cases take the ending –xe [kse], which is added to the masculine form:

(18)	étudiant/etydjᾶ/‘student [masc]’
	étudiante/etydjᾶt/‘student [fem]’
	étudiantxe/etydjᾶkse/‘student [neutral]’

The full list of neutral endings within the the AN-EX system is included in Table 3.

3.3.2. Alpheratz’s (2018) System

Alpheratz is a linguist based in Paris. Their Alpheratz (2018) grammar is the most detailed of all the grammars in the corpus. In addition to providing recommendations on non-binary agreement strategies, they also include examples of alternative non-binary forms that they have collected from written usage. They explain the underlying logic of their systematic grammar as based in the paradigms of standard French: “La formation des unités de genre neutre procède par identification des paradigms existant en français standard entre masculin et féminin et par leur reproduction ou leur continuation logique au neutre [the formation of the units of the neutral gender proceeds via the identification of paradigms existing in standard French between the masculine and the feminine and through their reproduction or their logical continuation in the neutral]” (Alpheratz 2018, p. 277).

Most likely because Alpheratz cites as influences a variety of different strategies for forming non-binary variants, their system is morphologically complex. Some allomorphs do not have conditioning patterns and must be memorized. Although the system has many exceptions, there are several major patterns that are best grouped based on how they relate to the pronunciations of the masculine/feminine dyads in oral French, each of which is illustrated in

Table 4. Neutral Singular Nominal Endings in Alpheratz (2018).

Phonologically Conditioned Forms	Where the masculine and feminine are homophones	Orthography: After unnacented vowels, <r>, and <l>, add –x to the masculine; replace –é with –æ; all others drop a final consonant from the masculine before adding –x. Exceptions: –el changes to –æl; –ic changes to –ix. V [V] (–i, –o(t), –u) → Vx [V] –é [e] → –æ [e] Vc [Vk] (–ic, –ec, –uc) → Vx [Vk] Vr [Vʁ] (–eur/–eure) → Vrx[Vʁ] Vl [Vl] (–al, –il, –ol, –ul) → Vlx [Vl] –el [ɛl] → –æl [ɛl] Oral: Homophonous with masculine and feminine forms.	ami·e → amix docteur·e → docteurx admiral·e → admiralx civil·e → civilx Espagnol·e → Espagnolx flic → flix député·e → députæ
	Where the feminine is orally distinguished from the masculine by the pronunciation of a final consonant	Orthography: For forms where the feminine is distinguished from the masculine by the pronunciation of a final consonant, drop the final consonant from the masculine and add –x [Ø]. Exceptions: change <e> to <æ> before <r>, as in masculine forms ending in –ert, –ier; for the dyad –eux/–euse, use –euz.24 V(n)t/d/s (–a(n)t, –a(r, n)d, –ais, –ent, –ond, –oint), → V(C)x [V] –ert [ɛʁ]→ –ært [ɛʁ] –(i)er/[e]–(i)èr [ɛʁ] → –ær [e] –eux[ø]/–euse [øz]→ –euz [ø] Oral: Homophonous with the masculine form.	avocat·e → avocax chauffard·e → chauffarx Anglais·e → Anglaix Allemand·e → Allemanx étudiant·e → étudianx profond·e→ profonx expert·e → expært infirmier, infirmière → infirmiær hereux, hereuse → heureuz
	Where the feminine is orally distinguished from the masculine by the pronunciation of a final consonant and a vowel change	Orthography: Nouns ending in –an or –on take an –x; nouns ending in –ain or –en change to –an; nouns ending in –in change to –aine. Exception: nouns ending in –on adopt the same orthographic pattern as other nouns ending in Vn, but are orally distinct from both the masculine and the feminine in the neutral. Vn (–an, –en) [Vnasalized]→ Vnx [Vnasalized] –on [ɔ̃] → –o(n)x [o] –ain [ɛ̃], –en [ɛ̃] → –an [ᾶ] ~[αn] –in [ɛ̃] → –aine [ɛn] Oral: Nouns of this type ending in orthographic –x are pronounced identically to masculine forms; non-binary nouns ending in –an or –aine are orally distinct from both the masculine and the feminine.	musulman·e → musulmanx or musulmaine patron·e → patronx citoyen·ne → citoyan patron·e → patrox or patronx voisin·e → voisaine
	Where the feminine and the masculine have orally distinct suffixes	Orthography: These forms end in –aire. –eur (m.) [œʁ]/–ice(f.) [is]→ –aire [ɛʁ] –eur (m.) [œʁ]/–euse(f.) [øz] → –aire [ɛʁ] –eur (m.) [œʁ]/–eresse[ʁɛs]~ [əʁɛs] (f.) → –aire [ɛʁ] Oral: Orally distinct from both the masculine and the feminine.	acteur, actrice → actaire danseur, danseuse → dansaire
Other Forms		Orthography: –et (m.) [e]/–ette (f.) [ɛt] → æt [ɛt] Oral: Homophonous with the feminine.	coquet·tte → coqæt

at the end of this section.

Where the masculine and feminine are homophones, the non-binary forms undergo orthographic changes, but remain homophonous with the masculine and feminine forms. The spelling changes are most easily discussed in comparison with the masculine forms. For words where the masculine ends in any vowel except [e] when pronounced, the non-binary form takes a silent –x [Ø] grapheme. For words ending in –é [e], the –é is replaced by the grapheme –æ [e]. Words ending in –c [k] drop the final consonant, but replace it with an –x that is pronounced as [k], so that words such as flic/flique [flik] ‘cop’ are spelled flix [flik], but pronounced identically. Words ending in [ʁ] and [l] (with the exception of –el [ɛl]) add a silent –x after the final consonant; words ending in –el change the final spelling to – æl, but retain the pronunciation of [ɛl].

Where the feminine is orally distinct from the masculine by the pronunciation of a final consonant, the non-binary forms undergo orthographic changes, but remain homophonous with the masculine. Masculine forms ending in orthographic –t, –d, or –s drop the final consonant and replace it with a silent –x [Ø] grapheme. Masculine forms ending in –ert [ɛʁ] or –er change the final vowel to –æ [ɛ] before <r>, leaving the oralization of the ending unchanged. Finally, words of the type –eux (m.) [ø]/–euse (f.) [øz] change the final –x of the masculine to a silent –z [Ø], which results in an ending –euz [ø] pronounced identically to the masculine.

Where the feminine is orally distinguished from the masculine by the pronunciation of a final consonant and a vowel change, some non-binary forms are identical to the masculine, and others are distinct from both the masculine and the feminine. Most nouns ending in –an add a silent –x [Ø] and are pronounced identically to the masculine. However, nouns ending in –on, –in, –ain, –en, and some ending in –an23 undergo changes in their vowels or vowel quality that make them distinct from both masculine and feminine variants. Masculine nouns ending in –on [ɔ̃] may be spelled either –onx or –ox, with their pronunciations changing to [o]. Nouns ending in –ain [ɛ̃] or –en [ɛ̃] change their spelling to –an, which is pronounced as [ᾶ] or [ɛn]. Finally, nouns ending in –in [ɛ̃] change their endings to [ɛn].

Nouns with orally distinct masculine and feminine suffixes, as in the sets –eur (m.) [œʁ]/–euse (f.) [øz], –eur (m.) [œʁ]/–ice [is], and –eur (m.) [œʁ]/–eresse (f.) [əʁɛs]~ [ʁɛs] the ending changes to –aire [ɛʁ], which is orally distinct from both the masculine and feminine variants.

Additional forms do not fit any of the above patterns. While the majority of the exceptions involve lexically-gendered kinship nouns, one particular case is worth highlighting. Nouns ending in –et [e] (m.)/–ette [ɛt] (f.) take the non-binary ending –æt [ɛt]; this represents one of the rare cases in Alpheratz’s grammar where a non-binary form is pronounced identically to the feminine variant.

As is highlighted by Table 4, in the majority of cases, the neutral singular is formed by either adding a silent–x to the end of the masculine, or replacing the final consonant of the masculine with a silent—x. This means that most singular nouns (and adjectives) end in a silent –x, while most plurals end in a silent –z.25 As a consequence, the majority of non-binary forms are pronounced like the corresponding singular masculine variant. The frequency with which non-binary forms are orally indistinct from masculine forms is an important motivator for Ashley’s (2019) addenda to Alpheratz’s system.

However, there are other critiques worth levying at Alpheratz’s (2018) morphological system. Among the innovative forms Alpheratz proposes, there is not a one-to-one grapheme-to-phoneme ratio, thus making their system orthographically opaque. For example, the innovative grapheme <æ> can be either [e] (as in députæ [e]) or [ɛ] (as in expært [ɛʁ]) (Alpheratz 2018, p. 286), which they acknowledge “nous semble introduire une complexité dans notre langue [seems to us to introduce complexity into our language]” (288). Even more confusingly, <x> is silent in the majority of cases, but pronounced [k] in the case of flix [flik] ‘cop’ and grex [gʁɛk]. While it is true that standard French is already an orthographically opaque language (Sprenger-Charolles and Serniclaes 2003), the choice to introduce even more opacity within their system is a curious one. Although these spelling quirks won’t affect oralization in most of these cases as they largely have non-binary forms homophonous with their masculine and feminine counterparts, they are still likely to present challenges for those looking to adopt Alpheratz’s system.

There are also issues in the consistency with which forms are transformed into non-binary variants orally. For example, why are some non-binary variants where the feminine is distinct from the masculine via both a vocalic change and the pronunciation of a final consonant orally unique, where others are identical to the masculine? An illustration of this phenomenon can be found in (19a–b):

(19)	a.
	musulman /myzylmᾶ/ ‘muslim [masc.]’26
	mululmane /myzylmαn/ ‘muslim [fem.]’
	musulmanx /myzylmᾶ/ ‘muslim [neutral]’
	b.
	patron /patʁɔ̃/ ‘patron [masc.]’
	patronne /patʁɔn/ ‘patron [fem.]’
	patronx /patʁo/ ‘patron [neutral]’

Further, the exceptional status of the non-binary form –æt [ɛt] from the set –et (m.) [e]/–ette (f.) [ɛt] as orally indistinct from the feminine is puzzling, especially given the fact that –æ can be pronounced [e] elsewhere in Alpheratz’s grammar (so why not use –æ [e] for consistency?).

3.3.3. Ashley’s Addenda to Alpheratz

Ashley (2019), an author from Montréal, proposes moderate revisions to Alpheratz’s (2018) system. In Alpheratz’s system, any forms ending in the neutral singular marker –x and the neutral plural marker –z are pronounced the same way as their masculine counterparts. Ashley (2019), while not opposed to general homophony of non-binary variants with binary counterparts, criticizes what they view as implicit favoring of masculine-like pronunciations. Thus, in order to “éviter plus souvent l’indistinction phonétique entre le masculin et le neutre quand le féminin est phonétiquement distinct [more often avoid the phonetic indistinction between the masculine and the neutral while the feminine is phonetically distinct],” which Ashley concedes is often viewed as “invalidation et mégenrage par certains personnes non-binaires [invalidation and misgendering by certain non-binary people],” Ashley proposes marking non-binary forms with the endings –al, –at, –an, –ane, –aite, –aine, –aive, or –aire according to whichever of these variants will “préserver le plus possible les phonèmes utilisés dans les terminaisons masculines et féminines [preserve the highest possible number of phonemes utilized in the masculine and feminine forms],” which conforms with their reading of Alpheratz’s approach (Ashley 2019, p. 9). The approach of using traditionally epicene endings to form non-binary nouns has its predecessor in the AL system (Alpheratz 2017). For example, as in Alpheratz’s system, nouns ending with–in in the masculine take the neutral ending–aine:

(20)	cousin /kuzɛ̃/ ‘cousin [masc]’
	cousine /kuzin/ ‘cousin [fem]’
	cousaine /kuzɛn/ ‘cousin [neutral]’

Additionally, bases ending in oral vowels and nouns ending in <nt> or <nd> take –xe in the singular and –z in the plural (Ashley 2019, p. 9).27 While Ashley (2019, p. 9) writes that –xe is pronounced “’kse’” and –ze is “’ze,’” the use of quotation marks rather than brackets makes it unclear if they are using IPA; due to this lack of clarity, I will use E to signal that [e], [ɛ], and [ə]—each of which can be represented by <e> in French—are all possibilities here:

(21)	ami /ami/ ‘friend [masc]’
	amie /ami/ ‘friend [fem]’
	amix /amiksE/ ‘friend [neutral]’
(22)	étudiant /etydjᾶ/ ‘student [masc]’
	étudiante /etydjᾶt/ ‘student [fem]’
	étudianxe /etydjᾶksE/ ‘student [neutral]’

The fact that –xe [ksE] occurs in both the AN-EX and Ashley’s systems illustrates the ways in which non-binary morphological systems influence one another, and thus should not be analyzed in isolation but rather as examples of broader patterns occurring in non-binary French today.

Homophonous masculine and feminine variants that end in consonants preserve the unpronounced singular –x and plural –z non-binary markers from Alpheratz’s system, thus retaining homophony:

(23)	libéral SF: /libeʁαl/ CF:/libeʀαl/ ‘liberal [masc]’
	libérale SF: /libeʁαl/ CF:/libeʀαl/ ‘liberal [fem]’
	Liberalx SF: /libeʁαl/ CF:/libeʀαl/ ‘liberal [neut]’.

This proposal is part of a broader number of adjustments that Ashley suggests on the basis of building a non-binary grammar that favors neither masculine nor feminine orthography nor pronunciation.

Additionally, Ashley proposes two specific orthographic modifications:

• When the feminine variant of a word is differentiated from the masculine by accent-marking, the non-binary form shall retain the accent. This yields forms such as printanièr ‘spring [neut],’ which is orthographically distinct from both the masculine form printanier and the feminine printanière, but pronounced the same as the feminine variant.
• The <æ> marker shall be written as <ae> for greater accessibility among keyboard users.

To the first point, the mixed-accent marking results in blends of masculine and feminine suffixes—although these blends operate solely at the orthographic level. Ashley is not the first to recommend spelling reforms that use accents to form the neutral. Labrosse (1996) recommended neo-endings –èle for adjectives ending in –el (m.)/–elle (f.) (55) and –ète for the set –et/–ette (61), which resulted in forms such as the ones in (24a–b):

(24)	a. officièl ‘official [neutral]’
	b. nète ‘clean [neutral]’

Unlike the majority of grammars taking the Systematic Approach, Ashley provides one (adjectival) case where non-binary variants may be derived from the masculine via subtractive morphology. For the ending set –se (f.)/–x (m.), the neutral is formed by subtracting the final –x from the masculine:

(25)	jaloux /ʒalu/ ‘jealous [masc.]’
	Jalouse /ʒaluz/ ‘jealous [fem.]’
	Jalou /ʒalu/ ‘jealous [neutral]’

However, since the <x> was already silent, the resulting neutral form is homophonous with the masculine.

Because Ashley’s system has a higher number of neutral allomorphs than either the AN-EX system or Alpheratz’ system—and does not elaborate much on the choice of one ending versus another—it is difficult to create derivational rules for them. Instead, I have listed all possible nominal allomorphs described in Ashley’s system in

Table 5. Neutral Singular Nominal Allomorphs in Ashley (2019).

Type I: –x, –xe	–x [Ø]	libéral·e → libéralx
	–x [ksE]	ami·e → amix
	–xe [ksE]	étudiant·e → étudiantxe
Type II: Intralinguistic Epicene borrowings	–at	idiot·te → idiat
	–an	écrivain·e → écrivain
	–ane	No example given.
	–aite	Hittit·e → Hittaite
	–aine	cousin·e → cousaine
	–aive	créatif·ive → créataive
	–aire	auteur·ice → autaire
Type III: Masculine and feminine Blends	–ièt –ètte –èr	inquiet, inquiète → inquièt arpète, arpette → arpètte pasager, pasagère → pasagèr
Type IV: –æ	–æ	député·e → députæ

, with an example for each.

3.3.4. Cross-Comparison of Systematic Strategies

These differential analyses highlight individual authors’ priorities in drafting morphologies of non-binary French. While all three sources take into account morpho-phonological phenomena, as well as point out to varying degrees the ways they have seen the forms they cite utilized within non-binary Francophone communities, nonetheless each has taken an underlyingly different linguistic approach.

Alpheratz’s grammar is notable not only for its thoroughness and large scope, but also for its focus on etymology and language change. For example, Alpheratz argues for the use of –z as a non-binary plural marker on the basis of its historical use in French through the 18th century (Alpheratz 2017, 2018, p. 216), based on an earlier recommendation by Labrosse (1996, p. 63). Perhaps the most surprising manifestation of this focus is the inclusion of an etymological entry for each neo-morpheme they list, which tends to focus not on historical change in French morphology per se (though some entries do include this information), but rather on the history of the word they give as an example within their system.28

Their logical reasoning in innovating many neutral forms that are phonologically identical to the masculine lies in the principle of linguistic economy, and the preference to express meaning “avec le plus petit nombre d’unités [with the smallest number of units]” (Alpheratz 2018, p. 279). This strategy echoes the work of Labrosse (1996), whose own non-sexist grammar focused on creating homophones that could be used to refer to both men and women, predicated on the most common extant homophones in French (50).

Ashley’s (2019) priority is in creating greater phonological differentiation between neutral forms and their masculine and feminine counterparts. To this end, they use a combination of sounds uncommon in French (e.g., –xe [ksE]) alongside suffixes already found in French but more commonly marking epicene nouns. Philosophically, they are also occupied by the desire to privilege neither homophony with the masculine nor the feminine when homophony is inevitable.

Finally, de Villeneuve and Gheeraert (2018) advocate a system that is much more morphologically simple than either Ashley’s or Alpheratz’s. This system also places less emphasis on differentiation (e.g., epicene nouns and homophones sound the same, and are only distinguished in written French). At the same time, they characterize the morpho–phonology of standard French in much broader strokes, with the result being that the majority of the forms produced using their system have gender morphemes with consonant clusters less common in oral French. Alhough the authors acknowledge that these endings may be surprising, or perhaps “abruptes [abrupt]” for those familiar with the phonotactics of standard French, much like Alpheratz they justify this innovation on the basis of historical language change and the long history of phonological borrowings in French: “le français a par le passé incorporé des sons tout autant étrangers à notre paysage phonétiques [French has in the past incorporated (within its sound system) sounds that are foreign to our phonetic landscape] ” (42). This comment references the fact that French phonology has historically been supplemented via substrate influences form Germanic languages (among others) (Battye et al. 2001, p. 11), which contributed to such phenomena as the/gw/→/g/cluster in words such as guerre, ‘war’. The employment of the/ks/cluster here implies a similar borrowing phenomenon, this time either directly from English, where <x> [ks] is canonically associated with gender-inclusivity, or via other Romance varieties where/ks/has already been borrowed from English in this regard. For example, the high frequency with which the cluster [kse] is oralized in this system is reminiscent of how some hispanophones oralize the neutral singular –x marker as “[eks], [ɛks], or [oks]” (Papadopoulos forthcoming), which may also partially be a result of influence from the use of X [ɛks] as a gender-neutral marker in English orthography. This cross-Romance similarity highlights the ways in which the written form <x> and its oralizations as [ks] or [Vks]/[ksV] are becoming increasingly associated with gender inclusivity.

These systems do overlap. Notably, in all but two exceptional cases, all three Systematic Approaches to forming non-binary French nouns preserve homophony in non-binary nouns when the masculine and feminine variants are homophones, as illustrated in

Table 6. Homophones in Non-binary Grammars.

M.sg. Marker	IPA	F.sg. Marker	IPA	Alpheratz (2018)	IPA	AN-EX	IPA	Ashley (2019)	IPA	Examples
–al	[αl]	–ale	[αl]	–alx –auz (pl)	[αl] [o]	–alx	[αl]	–alx, –auz (pl)	[αl] [oz]	liberalx ‘liberal’
–eur	[œʁ]	–eure	[œʁ]	–eurx –eurz (pl)	[œʁ] [œʁ]	–eurx	[œʁ]	–eurx	[œʁ]	docteurx ‚‘doctor‘
–el	[ɛl]	–elle	[ɛl]	–elx –elz (pl)	—	–elx	[ɛl]	–ael	[ɛl]	professionelx ‘professional‘ professionæl professionae
–ol	[ɔl]	–ole	[ɔl]	–olx –olz (pl)	[ɔl] [ɔl]	–olx	[ɔl]	n.a.	—	Espagnolx ‘Spanish’
–oul	[ul]	–oule	[ul]	–oulx –oulz (pl)	[ul] [ul]	–oulx	[ul]	n.a.	—	saoulx ‘drunk’
–u	[y]	–ue	[y]	–ux –uz (pl)	[y] [y]	–ux	[y]	–ux	[ukse]	élux ‘elect’
–i	[i]	-ie	[i]	-ix –iz (pl)	[i] [i]	–ix	[i]	–ix –iz (pl)	[ikse] [iz]	amix ‘friend’
–é	[e]	–ée	[e]	–æ	e	–éx	[e]	–æ or –ae	[e]	députéx ‘deputy‘ députæ députae
–ète	[ɛt]	–ette	[ɛt]	n.a.	–	–ètex	[ɛt]	–ètte	[ɛt]	arpètex ‘apprentice‘ arpètte
–ec	[ɛk]	–ecque	[ɛk]	n.a.		–ecx	[ɛk]	–æac or –aec	[e]	Grecx ‘Greek‘ Græc

Forms shaded in gray have pronunciations distinct from the masculine and the feminine.

:

However, there are a number of significant differences. These differences are best embodied through a side-by-side comparison of the various strategies for innovating non-binary counterparts to nouns traditionally analyzed under the subtractive model, as presented in

Table 7. Three Systematic Strategies for Forming Non-binary Nouns.

M.sg. Marker	IPA	F.sg. Marker	IPA	Alpheratz (2018)	IPA	AN-EX	IPA	Ashley (2019)	IPA	Examples
–ais	[ɛ]	–aise	[ɛz]	–aix (sg) –aiz (pl)	[ɛ] [ɛ]	–aixe (sg) –aixes (pl)	[ɛkse]	–aix (sg) –aiz (pl)	[ɛ] [ɛ]	anglaix ‘English’ anglaixe anglaix
–ain	[ɛ̃]	–aine	[ɛn]	–an (sg) –ans (pl)	[ᾶ] [ᾶ]	–an (sg) –ans (pl)	[ᾶ]	–an (sg) –ans (pl)	[ᾶ]	écrivan ‘writer’
–an	[ᾶ]	–ane	[αn]	—aine or –anx (sg)	[ɛn]	–an (sg) –ans (pl)	[ᾶ]	–aine (sg) –aines (pl)	[ɛn] [ɛn]	musulman ‘Muslim’ musulmanx musulmaine
				–aines or –anz (pl)	[ᾶ]
–ant, –and	[ᾶ]	–ante, –ande	[αt]	–anx (sg) –anz (pl)	[ᾶ] [ᾶ]	–antxe, –andxe (sg) –antxes, –andxes (pl)	[ᾶkse]	–anxe (sg) –anze (pl)	[ᾶksE] [ᾶzE]	allemanx ‘German’ allemantxe allemanxe
–ard	[aʁ]	–arde	[aʁd]	–arx (sg) –arz (pl)	[aʁ] [aʁ]	–ardxe (sg) –ardxes (pl)	[aʁkse]	–aire –aires (pl)	[ɛʁ] [ɛʁ]	bagnarx ‘convict’ bargnardxe bagnaire
–at	[a]	–ate	[at]	–ax (sg) –az (pl)	[a] [a]	–atxe (sg) –atxes (pl)	[akse]	n.a.	—	avocatax ‘lawyer’ avocatxe
–ent	[ᾶ]	–ente	[ᾶt]	–enx (sg) –enz (pl)	[ᾶ] [ᾶ]	–entxe (sg) –entxes (pl)	[ᾶkse]	n.a.	—	agenx ‘agent’ agentxe
–ert	[ɛʁ]	–erte	[ɛʁt]	–ært (sg) –ærts (pl)	[ɛʁ] [ɛʁ]	–ertxe (sg) –ertxes (pl)	[ɛʁkse]	n.a.	—	expært ‘expert’ expertxe
–in	[ɛ̃]	–ine	[in]	–aine (sg) –aines (pl)	[ɛn] [ɛn]	–an (sg) –ans (pl)	[ᾶ]	–aine (sg) –aines (pl)	[ɛn] [ɛn]	voisaine ‘neighbor’ voisan
–it	[i]	–ite	[it]	–ix (sg) –iz (pl)	[i] [i]	–ixe (sg) –ixes (pl)	[ikse]	–aite	[ɛt]	hittix ‘Hittite’ hittixe hittaite
–ond	[ɔ̃]	–onde	[ɔ̃d]	–onx (sg) –onz (pl)	[ɔ̃] [ɔ̃]	–ondxe (sg) –ondxes (pls)	[ɔ̃kse]	n.a.	—	vagabonx ‘vagabond’ vagabondxe
–on	[ɔ̃]	–one	[ɔn]	–o(n)x (sg) –o(n)z (pl)	[o] [o]	–an (sg) –ans (pl)	[ᾶ]	n.a.	—	fripox ‘rascal’ friponx fripan
–ois	[wa]	–oise	[waz]	–oix (sg) –oiz (pl)	[wa] [wa]	–oixe (sg) –oixes (pl)	[wakse]	n.a.	—	chinoix ‘Chinese’ chinoixe
–ot	[o]	–ote	[ɔt]	–ox (sg) –oz (pl)	[o] [o]	–otxe (sg) –otxes (pl)	[okse]	–at	[at]	idiotx ‘idiot’ idiotxe idiat
–ut	[y]	–ute	[yt]	–ux (sg) –uz (pl)	[y] [y]	–utxe (sg) –utxes (pl)	[ykse]	–ux	[uksE]	substitux ‘substitute’ substitutxe

Forms shaded in gray have pronunciations distinct from the masculine and the feminine.

:

Like the Compounding Approach, these Systematic Approaches have clear parallels to earlier movements for inclusive French. Most strikingly, the Systematic Approach bears resemblance to the way feminization has operated in various French contexts in recent history. Feminization was increasingly systematized with the release of manuals in the late 20th century (e.g., Biron 1991; Becquer et al. 1999), which often provided general rules for deriving feminine forms. For example, the guide from the Office de la langue française in Québec lists the rules for deriving title and function nouns in a series of subsections with examples (Biron 1991, pp. 10–13). This is exemplified by the following excerpt for nouns ending in –er and –ier:

“L’Office propose de généraliser la formation régulière du féminin en -ère et -ière sur le modèle de bergère et ouvrière. D’où: une banquière, une bouchère, une brigadière, une charpentière…, etc. [The Office proposes to generalize the regular formation of the feminine in –ère and –ière according to the model of bergère (shepherdess) and ouvrière (worker [fem.]). Whence: a banker [fem.], a butcher [fem.], a carpenter [fem.]…etc.”.

(Biron 1991, p. 11)

However, unlike the Compounding Approach, the Systematic Approaches in French closely mirror non-binary marking patterns in other languages, particularly in the formation of non-binary nouns cross-Romance. For example, there are a variety of Systematic Approaches to non-binary Portuguese, including the Sistema Elu, which uses the singular non-binary gender-markers –e, –que, –gue, –re, –ce, –ãe, –oe, –one, –ese (or –esu), –esse, –ie, –ise, –íze, –eie, –use, -íne, -ue (Caê Almeida 2020; Cassiano 2020).

3.3.5. Systematic Approaches in Use

During our conversation, Bardot extensively referenced their use of the AN-EX system:

Bardot:

Um, a third common thing to do is, uh I’m just going to give an example, so mignon [mɪɲɔ̃], mignonne [mɪɲɔn], it’s like cute, and you can say mignan [mɪɲã], the ending is <an> instead of <on>.

Similarly, they describe their use of neutral singular –xe and neutral plural –xes:

Bardot:

The good-to-go-rule, like [when] I don’t want to […] think about it too much, […] I’m just going to make up a non-binary ending, is to add <xe> at the end of the word that you pronounce [ksə], as in, for example, um heureuxe [œʁ.øks]. So heureux [œʁ.ø], heureuse [œʁ.øz], which is like happy [...] heureuxe [œʁ.øks]. Uh, contentxe [kɔ̃tᾶks] would uh also be ‘to be happy’ [laughs], uh in non-binary. Um les anglaixes [ᾶglɛks], like ‘the English,’ in [the] non-binary version. Yeah, it’s quite common. Um, it’s quite common to do that…. so a lot of people do that.

[...]

JK:

And for like, content, contente, how do you deal with that?

Bardot:

 Contexe [kɔ̃ tᾶks]. And—and, just to be honest, we know that it sounds very weird. Um…And maybe it’s not gonna stay. Maybe people will say that it’s too weird and it’s never gonna work out. So maybe…some different use will just appear and become used.29

Bardot’s characterization of the AN-EX system as one they use when they “don’t want to think about it too much” reveals the relative simplicity of this approach. With fewer morphemes and conditioning environments to memorize, it represents several key advantages for speakers like Bardot. However, as Bardot acknowledges themselves, one of the key shortcomings of this system is its use of clusters not typically found word-finally in French, which they concede, “sounds very weird.” They also acknowledge that non-binary French is currently a site of significant instability, as, in their words, “And maybe it’s not gonna stay. Maybe people will say that it’s too weird and it’s never gonna work out. So maybe…some different use will just appear and become used.” In this way, they note that non-binary French is continuously evolving. Year-to-year surveys of non-binary Francophones support this characterization, as La Vie en Queer found a marked increase in the number of neo-agreement markers respondents included in their (La Vie en Queer 2018b) results, versus their (La Vie en Queer 2017) results. Perhaps there will eventually be coalescence around agreement markers in the way that there is increasing coalescence around the non-binary pronoun iel (La Vie en Queer 2018b), although it is too early to predict which approach(es) will win out.

Bardot’s phonological realizations of the set heureux (m.)/heureuse (f.)/heureuxe (n.) ‘happy’ are notable because of their divergence from expected surface forms. While de Villeneuve and Gheeraert (2018) describe –xe as [kse], Bardot uses the pronunciation [ksə] when they say “add <xe> at the end of the word that you pronounce [ksə].” Additionally, when they give examples of how to use this suffix, they drop the schwa from their pronunciation in every token from our interview, as in “heureuxe [œʁ.yks],” and “contexe [kɔ̃tᾶks],” which is consistent with their lack of schwa in the feminine form, where it could occur after [z] but does not in Bardot’s speech. The apparent lenition of word-final [e] to [ə]~[Ø] is consistent with historic trends in French phonology, which frequently exhibits weakening word-finally (McLaughlin forthcoming, p. 3). Bardot’s realization of –an is also different from the variant recorded by de Villeneuve and Gheeraert, as in their realization of “mignan,” which surfaces as [mɪɲã] rather than [mignan]—although nasal assimilation is a common phonological phenomenon in French (Walker 1984).

Chandler also discussed their use of a system that appears very similar to the AN-EX system. They discussed how to use non-binary forms in written and oral French:

Chandler:

Uh, yeah, so there’s, like, let’s say a word, where the feminine form just has an additional <e>. Then [when writing] you’ll just put dot <e> at the end. If it’s plural, it’s dot <e> dot <s>. Um, if it ends, if the masculine form ends with an <x>, then we can just add an <e> at the end. Like um… Let’s say, curieux [kyʁ.iø] it would be] curieuxe [kyʁ.iø.ksə]. And the feminine form is curieuse [kyʁ.iøz]. So that’s how we can deal with it.

Further, both Chandler and Bardot’s pronunciation of the morpheme –xe as [ksə] reveals movement away from the pronunciation indicated by de Villeneuve and Gheeraert (2018), which they consistently record as [kse]. This is especially notable due to the fact that both Bardot and Chandler were using careful speech, which is more likely to produce forms closer to their underlying phonological forms. Nor can we consider the deviation from [kse] to be strictly regional; if this were so, we would expect Bardot, who lives outside Paris, to pattern their speech more closely with de Villeneuve and Gheeraert’s recommendation. Instead, Bardot’s realization of <xe> more closely match’s Chandler’s in that they both have one token with schwa, and neither uses the form [kse], although Chandler is from Quebec. If the usage by these two participants represents a trend, it is perhaps an exciting indicator of emerging variation in the phonology of non-binary French—or at least, a movement away from prescribed forms.

4. Discussion

4.1. Themes across All Three Strategies—Compounding, Systematic, and Invariable

In analyzing each of these three strategies holistically, their debt to earlier inclusive French strategies becomes clear. The co-presence of distinct masculine and feminine morphemes in inclusive writing norms likely influenced their co-occurrence in the Compounding Approach. Likewise, the Systematic Approach, with its rules for deriving non-binary forms via sets of suffixes, has much in common with the standardization of feminization.

There are a number of specific resonances with additional inclusive French strategies. Already epicene forms such as –al and–aire come up repeatedly, as does –an which is similar to the epicene ending –ane. Appeals to neologism via epicene suffixation are found in much older inclusive grammars, including that of Labrosse (1996). This is perhaps unsurprising; Alpheratz (2018) highlights already-epicene nouns and places even greater emphasis on their utility within non-binary French in their earlier work (Alpheratz 2017).

The letters especially <z> and <x> (Alpheratz 2018, p. 284) are common features of non-binary French morphologies. However, the pronunciation (or lack thereof) of these forms varies across systems. For example, <x> is pronounced as [ks] in one invariable strategy, while <xe> is pronounced [kse] (surfacing as [ks]~[ksə] in interview data) within certain conditioning environments in the AN-EX system. It is worth noting that <x> and <z> are common features of both other Romance varieties, including Spanish, as well as in English neopronouns (e.g., the common sets ze/hir/hirs and xe/xem/xers; see Hord 2016). This commonality is not random. Within the English context, it’s likely that they became popular in neo-pronouns precisely because of their rare usage in English, thus marking these forms as orthographically distinct. As much early work on non-binary language is based on the English context, <z> and <x> may be becoming signifiers of non-binary gender more broadly (Papadopoulos forthcoming); the fact that many non-binary scholars explicitly cite both English (Alpheratz 2018, p. 283; de Villeneuve and Gheeraert 2018) and Spanish (Alpheratz 2018, pp. 52, 283; see also Swamy’s (2019) interpretation of non-binary media) gender-inclusive constructions further supports the theory that –x is a morphological borrowing across non-binary languages that semantically encodes gender expansiveness.30,31 Further, within and beyond the Spanish context, the <x> marker has taken on liberatory and anti-colonial connotations: As a legacy of Xicanx feminism, indigenous orthography in Latin America, and as a visual representation of the violence of colonialism (Scharrón-del Río and Aja 2020). As Scharrón-del Río and Aja note (2020) The oralized <x>-marker is also likely linked to the broader phenomenon of an X marker being used to denote non-binary gender on legal documents, which is becoming increasingly common in western countries (Ashley 2021).

Although I have discussed the Compounding Approach, the Invariable Approach, and the Systematic Approach separately, the strategies are not entirely distinct. In addition to the fact that non-binary individuals often mix–and–match approaches—both in different situational contexts, and within the same conversation—the strategies themselves often overlap; for example, the AN-EX system uses some compounded forms (see Example 17 in Section 3.3.1).

In looking at the different strategies in their totality, we can subdivide them on their morphological and phonological patterns. Excepting blends, which—at the nominal level—are almost exclusive to kinship terms, we find three affixational patterns. The first involves Compounding masculine and feminine affixes. The second involves affixing neutral suffixes to bases, which may be either extensions of endings commonly found on epicene nouns, or neo-morphemes. These neomorphemes may in turn be reintroduced historic suffixes, blends of masculine and feminine suffixes, borrowings from other languages, or single-letter morphemes with less clear origins. The final pattern forms the neutral by subtracting the final written consonant of the masculine form. These patterns are summarized in

Table 8. Patterns of Deriving Non-binary Nouns.

Derivational Pattern	Notes	Codes in Which It is Distinct from the Masculine and Feminine	Associated Approaches	Examples
1. Compounding masculine and feminine affixes	Highly Restricted; only possible on forms with a non-null masculine suffix	Written & Oral	Compounding approach; AN-EX System	Acteur, actrice → acteurice (n. sg.)
2. Neutral suffixes attaching to bases  a. Extension of epicene endings		Written & Oral	Ashley (2019) Alpheratz (2018)	Acteur, actrice → actaire (n. sg.)
b. Neo-morphemes
i. Historic affixes (intralinguistic borrowing)		Written; Sometimes oral	Alpheratz (2018)	Silent z (pl): Acteurs, actrices → acteuz (n. pl.)
ii. Affixes from other languages (extralinguistic borrowing)		Written; sometimes oral	Alpheratz (2018) AN-EX Invariable approach	Acteur, actrice → acteux (n. sg.)
iii. Blending masculine and feminine affixes	Highly Restricted; only possible where feminine suffix has a written accent	Primarily Written	Ashley (2019)	inquiet, inquiète → inquièt (n. sg.)
iv. Origins unclear	Potentially limitless variations; less common and more idiosyncratic than other affixational patterns	Written & Oral	Invariable approach	Pronounced –s: Acteur, actrice → acteus (n. sg.) Other examples: –m, –t, –u…
3. Subtractive morphology	Very rare	Primarily Written	Ashley (2019)	jaloux, jalouse → jalou (n. sg.)

below:

We can also sub-divide non-binary nominal forms into those that are phonologically innovative and those that are phonologically conservative (no one system is wholly one or the other). We might also highlight phonologically super-innovative forms, which contain features that are otherwise exceedingly rare in standard French (e.g., utilizing the consonant cluster [ks] syllable-finally, as with some variants of the –x marker).

Phonologically Conservative: Non-binary forms are homophonous with their masculine and feminine counterparts. Endings associated with homophony include–x muet and–z muet.

Phonologically Innovative: Non-binary endings are distinct from binary endings in the same paradigm, but are otherwise unremarkable in their phonology when compared to standard French, although they may be orthographically different or extended cross-paradigmatically. Examples include –eurice [compounded], –an [ᾶ] for –ain[ɛ̃]/–aine[ɛn]/ where it otherwise occurs with –ant/–ante, –and/–ande etc.

A subtype of this strategy includes borrowing endings characteristic of epicene forms, such as –al and –aire.

Phonologically Super-Innovative: Endings or suffixes incorporating sounds otherwise rare syllable-finally in French. These include: –xe/kse/~[ksə]~[ks], –x [ks], –z [ze].

A summary of the three approaches to forming non-binary nouns—Compounding, Invariable, and Systematic—can be found in

Table A1. Three approaches to forming non-binary nouns.

Masculine and Feminine Singular Forms	Gloss	Compounded Non-Binary Singular Forms	Systematic Non-Binary Singular Forms	Invariable Non-Binary Singular Forms
demandeur [œʁ] (m.), demanderesse [ʁɛs](f.)32	‘applicant’	—	demandeuxe [œkse] (AN-EX)	demandeux [œks] demandeuz [œz]
			demandaire [ɛʁ] (Alpheratz) n.a. (Ashley)
acteur [œʁ] (m.), actrice [ʁis](f.)	‘actor’ ‘actress’	acteurice [œʁis]~[œʁʁis] actriceur[ʁisœʁ]	acteurice[œʁis] (AN-EX)	acteux [œks] acteuz [œz]
			actaire [ɛʁ] (Alpheratz) actaire [ɛʁ]~[ɛʀ](Ashley)
joueur [œʁ] (m.), joueuse [øz] (f.)!	‘player’	joueureuse[œʁøz]	joueuxe [œkse](AN-EX)	joueux [œks] joueuz [œz]
		joueuseur[øzœʁ]	jouaire [ɛʁ](Alpheratz) jouaire [ɛʁ] (Ashley)
professeur [œʁ] (m.), professeure [œʁ] (f.)	‘professor’	—	professeurx [œʁ] (AN-EX)	professeux [œks] professeuz [œz]
			professeurx [œʁ] (Alpheratz) professeurx [œʁ]~[ø] (Ashley)
professionnel [ɛl](m.), professionnelle [ɛl](f.)	‘professional’	—	professionelx [ɛl](AN-EX)	professionnex [ɛks] professionnez [ɛz]
			professionæl [ɛl](Alpheratz) professionael [ɛl](Ashley)
infirmier [je] (m.), infirmière [jɛʁ] (f.)!	‘nurse’	—	infirmiexe [jekse] (AN-EX)	infirmiex [jeks] infirmiez [jez]
			infirmiær [jɛʁ] (Alpheratz) infirmièr [jɛ](Ashley)
canadien [jɛ̃](m.), canadienne [jɛn] (f.)	‘Canadian’	—	canadian [jᾶ ](AN-EX)	canadiex [jɛks] canadiez [jɛz]
			canadian [jᾶ ](Alpheratz) canadian [jᾶ ](Ashley)
copain [ɛ̃] (m.), copine [in] (f.)	‘friend’	—	copaine [ɛn] (AN-EX) copaine [ɛn] (Alpheratz) copan [ᾶ] (Ashley)	copaix [ɛks] copaiz [ɛz]
citoyen [ɛ̃] (m.), citoyenne [ɛn] (f.)	‘citizen’	—	citoyan [ᾶ] (AN-EX)	citoyex [ɛks] citoyez [ɛz]
			citoyan [ᾶ] (Alpheratz) citoyan [ᾶ] (Ashley)
député [e] (m.), députée [e] (f.)	‘deputy’	—	députéx [e] (AN-EX)	députéx [eks] députéz [ez]
			députæ [e] (Alpheratz) députae[e] (Ashley)
coquet [e] (m.), coquette [ɛt] (f.)	‘coquette’	—	coquetxe [eksE] (AN-EX)	coquex [eks] coquez [ez]
			coquæt [ɛ](Alpheratz) n.a. (Ashley)
arpète [ɛt] (m.), arpette [ɛt] (f.)33	‘apprentice’	—	arpètex [ɛt] (AN-EX) n.a. (Alpheratz) arpètte[ɛt] (Ashley)	arpèx [ɛks] arpèx [ɛz]
duc [yk](m.), duchesse [ɛs] (f.);	‘duke’ ‘duchesse’	—	duchexe [ɛkse] (AN-EX) dux [yk] or [yks](Alpheratz) n.a. (Ashley)	dux [yks] duz [yz]
pasager [e](m.), pasagère [ɛʁ](f.)	‘passenger’	—	pasagerxe[ekse] (AN-EX) pasagær [ɛʁ] (Alpheratz) pasagèr [ɛ](Ashley)	pasagex [eks] pasagez [ez]
grec [ɛk] (m.), grecque [ɛk] (f.)	‘Greek’	—	grecx [ɛk] (AN-EX) n.a. (Alpheratz) graec or græc [ek] (Ashley)	grex [ɛks] grez [ɛz]

in the Appendix A.

4.2. Closing Thoughts

The competing strategies for forming nominal non-binary French variants are striking examples of the classic tension between comprehensiveness and simplicity in language. While it is true that this tension is also present in the irregular masculine/feminine dyads, it is important to remember that these pairings were formed and reanalyzed over centuries; in fact, the feminine variants for many animate nouns used in modern standard French emerged only quite recently, with the concentrated efforts to visibilize women in language throughout the second half of the 20th century. Even these have not been entirely standardized, with variation occurring in some forms (e.g., Parisian French autrice, Quebec French auteure, ‘author’ (f.)). By comparison, non-binary grammars, and non-binary French more broadly, have emerged with innovative forms for the entire morphological gender system in just a few short years.

While many authors have been clear about the high degree of variation in non-binary agreement strategies (e.g., Knisely 2020; Shroy 2017), not unlike the irregularity in neutral English pronouns (Baron 2020) before recent coalescence around singular they, this paper has revealed commonalities and patterns in the morphology and morphophonology of non-binary French. While it remains to be seen which approaches become most common in the long-term, identifying these commonalities even among the diversity of approaches today underscores the point that “non-binary French” embodies a series of coherent linguistic patterns that are neither random nor un-influenced by the morphology and phonology of Standard French–even if the influence results in the creation of forms intentionally divergent from Standard French patterns.