Inglês (pdf)
Artigo em XML
Referências do artigo
Tradução automática
Enviar este artigo por email
Citado por SciELO 
Similares em
SciELO 
Permalink
1. Introduction. Inclusive language and third gender in Spanish
There are generally only 2 genders in Spanish, with the exception of some demonstrative pronouns (esto, eso, aquello) that refer not to nouns but to events or bigger units previously mentioned, and with the exception of nominalized relative clauses with “lo” (lo que dijo, that which s/he said), a construction that does not refer to nouns either but more global units, consistent with the fact that nouns always have a grammatical gender. Another important characteristic is that in most cases gender in Spanish is grammatical, and therefore arbitrary and non-semantic. Thus, Spanish has been typologically classified as a dual-gender language with grammatical gender and some semantic gender (Corbett 2005; Clegg 2011). This has created two sites for issues: on the one hand, what gender to use to represent mixed groups; on the other hand, how to represent those that do not identify as either gender (for nouns that have human referents)1. Inclusive language policies attempt to resolve these issues, which has meant using the same term for both kinds of issues. This paper will discuss only possible solutions for the second issue: how to refer to people that do not identify with either gender, rather than to non-sexist policies that offer more female visibility. There are two logical possibilities for a linguistic intervention here: to eliminate gender altogether (similar to most nouns in English, for instance), or to add at least one more gender2.
The complete elimination of gender would imply a very dramatic change in the system, with several ripple effects. Gender has several roles in Spanish. One of them is reference tracking and the consequent resolution of ambiguity, for instance with a resumptive pronoun that indicates gender, number and sometimes case (Silva Corvalán 2011). While not present in every case and sometimes solvable by context, the current system does rely on gender for reference tracking. There is also crosslinguistic evidence that shows how gendered languages that use gender grammatically rely on gender for syntactic category disambiguation (Rogers & Gries 2022). Furthermore, gender in Spanish is a productive morphological process to indicate differences in size (huerto/huerta small yard/big yard, río/ría river/estuary), tree vs. fruit/seed (manzano/manzana apple tree/apple, castaño/castaña chestnut tree/chestnut), and even gender minorities visibilities (female visibility) as with the famous miembra neologism vs. just the officially accepted miembro (member), or more common ones like presidenta/presidente. Finally, and although this could change in time, currently gender is so inextricably linked to each noun that a mismatch of gender and noun produces more processing effects that other perhaps more semantically relevant mismatches, such as those in number (Sagarra & Herschensohn 2013), and it has been shown to be acquired very early on and used in processing very early on (Dussias et al. 2013; Beatty-Martínez & Dussias 2019). For all these abundant reasons, this is not a proposal that has gained any traction.
The alternative proposal, however, to add one or more genders, has been increasingly embraced by at least certain social groups (Acosta Matos 2016; Papadopoulos 2022). Indeed, due to visibility and identity reasons, several proposals have been put forward to add a third gender in Spanish that would represent the option of “no gender”, similar to a neutral gender, although it has also been posited that such third gender would serve to represent, instead, groups with several genders. The body of literature preoccupied with how to implement this has been growing in the last few years, particularly with respect to Spanish as L2 teaching (Parra & Serafini 2021; García-Holgado et al. 2021; Conde 2022), social work (Bonilla Montesano 2019), translation (López-Medel 2021), and periodical publications (Guerrero Salazar 2022). Moreover, eight universities in Argentina have officially adopted some form of inclusive language between 2017 and 2020, although Buenos Aires banned its adoption from public schools in recent years. However, this impulse has also meant a multiplicity of proposals on how to proceed in order to mark this third gender.
In this paper, the different proposals for a third gender will be presented, as well as a set of criteria to evaluate them. Before presenting third gender options, though, it is worth mentioning the official position of the Real Academia Española (RAE).
The current general norm indicates that a specific marking is not needed, as there is already a non-marked marking for gender: the current masculine endings. However, as feminist approaches have pointed out, the mere fact that one of the genders is the default one is in itself discriminatory. As a result, all options proposed so far have been explicitly rejected by the RAE (de Miguel 2022; Bueno Ochoa 2022). To this, some academics add the dubious aesthetic argument of how much prettier the language is in its current state vs. possible future alterations (Pano Alamán 2022, 38; Grijelmo 2019, 137, citing Lázaro Carreter).
1.1. Norms as top-down approaches to language change
Concerning gender approaches, whether aligning with RAE guidelines or non-sexist recommendations in style manuals for various sectors, including press, corporate hiring, public affairs, and educational institutions, particularly higher education, we are talking of top-down language change strategies. Whether advocating language preservation by imposing norms on evolving aspects or promoting changes to align with recent social shifts, both aims are top-down proposals, coming from established institutions or specific social groups (as seen in Acosta Matos' work (2016) on anarchist publications). Regardless of the source, both strategies are normative and represent top-down efforts to either modify a language or to maintain it (de Miguel 2022).
In contrast, diachronic linguistics, the study of linguistic change, predominantly focuses on forces shaping languages over time. Despite attention to top-down changes from norms, the discipline mainly studies bottom-up changes arising from emerging and unconscious processes. This preference stems from limited knowledge about historical norms, restricted government influence on language homogeneity in the past, and historically constrained media and education access. Not only that, the importance of such endeavors was also a lot less salient than they may be nowadays. And while historical data is only written, it is frequently much more heterogeneous that it may be nowadays. Due to all this, diachronic linguists have been captivated by emergent, sometimes drastic, changes, such as shifts from inflectional to isolating languages or the loss of a case system, as typical bottom-up changes.
The success of a top-down change is based on exerting enough pressure, by being able to enforce, somehow, their norms. There are several ways in which this can be done: through grammars, through different style manuals and regulations in institutions, through entering the educational curriculum, through the shaming of opposite options or the infusion of prestige3 to the defended ones, or even adoption of this norm by the relevant writing or dictating software (such as word processors or even, in a less transparent way, by AI generated text). In this respect, and as de Miguel (2022) shows, norms, as arbitrary forms of regulation, have little limitations in what they could, in principle, ask of users.
The case of bottom-up emergent changes is a bit more complex. Although extremely radical changes can, and indeed do happen, their direction is not arbitrary. In fact, they seem to follow some principles, even if, at the current state of knowledge, such principles are being discovered and sometimes they are only theoretical. For example, it has been proposed that most morphological and syntactic changes go from a more marked to a less marked option in that language (Bergs & Stein 2001), or that systems try to avoid duplicates, for instance at the morphological or lexical levels (e.g., in Spanish, rather than having two words for the same entity, goose, the Latin one oca and the Germanic ganso, each specialized on a different meaning: domestic vs. wild respectively [Campbell 2021]). Similarly, it has also been claimed that phonological change tends to irregularization (by mostly following economic principles in production) while morphosyntactic changes tend to regularization of patterns (by reanalysis, leveling, etc.), the Sturtevant’s Paradox (Collinge 1965).
Is there a way for top-bottom norms to align with more natural bottom-up tendencies, that is, to follow common trends in linguistic change? And are there characteristics that a proposed norm should follow to adapt more easily within the current form of the language? If there is, this way would probably facilitate the change with a lesser need for top-down imposition or at least a gentler application of it. In the next section, I will propose a few characteristics that I believe should make it easier to adopt a form, in a similar way to Auxland (2020), although this work only evaluates one option (<-e> ending for Portuguese).
1.2. Potential characteristics of an ideal option for a third gender ending from the perspective of language change
One of the features of the Spanish language in its current form is the phonetic character of its writing: when we see a word in Spanish we know how to pronounce it (with the exception perhaps of some newer non-adapted word loans), and when we hear it we can, to a great degree, guess successfully how to write it, although there are exceptions (v/b, z/s, mute h, etc.). So one characteristic that would make it easier to adopt would be to be phonetically transparent and also easy to pronounce according to current Spanish phonotactics. Furthermore, the phonetic realization is extremely relevant as most spontaneous and natural linguistic change occurs at the oral level first, only later trespassing the writing frontiers, always better defended by older norms. Some of the proposals for a third gender ignore this and make pronunciation either very difficult or impossible.
As mentioned above, morphological change tends to avoid repetition and ambiguity while preferring a more transparent relationship between form and meaning (Trask & Millar 2015). Because of this, a form that reflects already the paradigm for genders, if possible at all, would be easier to acquire, interpret, and maintain. If the same ending is used for different, sometimes opposed, meanings (like different genders for instance), its acquisition becomes more difficult. Also, if there is any ending that is already doing the particular needed job in the language, extending its use should be easier than adding a new one and more consistent with the current system4.
In addition, a non-particularly socially marked ending, one with no negative connotations, should also be preferred by most speakers.
Finally, and related to a transparent form-meaning pairing, the ideal form would try to capitalize on already changing forms or forms that already exist in order for the new proposal to feel more natural and less new.
The next section will present the current proposals and discuss whether they fulfill these desired characteristics.
2. Current proposals for a third gender ending in Spanish
While the amount of academic and mediatic discussion dedicated to non-sexist uses of the Spanish language is vast, academic articles or books are very much lacking in the case of inclusive third gender for Spanish and has been done mainly through recent theses and dissertations not just in the linguistic field, but in the sociology, history, or even art fields of study. A bibliography of these works was very recently published by Cabello Pino (2020). Most of the works reviewed in this publication are dedicated to their use in the media or to observational and impressionistic studies, with only one corpus-informed quantitative study (with a very small sample) and only one psycholinguistic study on processing effects. Through the sum of those studies as well as a few afterwards - most of them focusing on speakers of different groups’ preferences (Reales Gil 2020 on differences among countries; Shenk 2023 on speakers’ attitudes; Zarwanitzer 2019 on processing difficulties; and Magagna 2021 on teachers’ attitudes) - 5 different options can be distilled: -@, <-*>, <-x>, <-e>, and <-i>. Of those, the options <-i> and <-*> are the less often commented, while, especially in recent years, <-e> and <-x> have been studied and discussed more frequently.
In the following subsections we discuss each one, evaluating pros and cons according to the set of preferred characteristics mentioned previously.
2.1. -@ ending
One of the earliest attempts to establish a gender-neutral ending in Spanish was the adoption of the symbol "@"-popularized with the widespread use of the internet and email in the late 1990s. Initially designed to encompass both masculine and feminine genders, its visual representation combining <-o> and <-a> was deemed ideal (Nissen 2002). Due to its origin predating the last two decades and already existing in the previous century, Cabello Pino (2020, 5) excluded it from his bibliographical review. Nevertheless, it is frequently examined in studies on attitudes and preferences toward inclusive gender language, especially those focused on a third gender ( Barrera Alvarado & Ortiz Rodríguez 2014; Acosta Matos 2016; Rodríguez Herrera 2019; Magagna 2021; Shenk 2023). Bengoechea (2015: 14-15) notes a shift from its initial non-sexist use to symbolizing intersexuality-a blend of chromosomal characteristics from both genders.
The popularity of the ending has probably declined, in part due to the rising of other options such as <-x> and <-*>, at least according to some researchers (Reales Gil 2020). However, Shenk (2023) found that, among invented endings <-e>, <-x>, and -@ for the purpose of a third gender, it was actually this symbol the one reported to be most frequently used in writing (22% of participants used it frequently or sometimes), and the second one most acknowledged as a form of inclusive language overall, only after the use of both masculine and feminine nouns. Most intriguingly, this was the second preferred option in Shenk’s study even for speaking, after <-e>, although neither one was very popular (8.3% and 13.9% respectively of frequent or occasional use). When asked how this symbol was used orally, one of the participants explained that s/he uses it by saying “arroba”. Similarly, Magagna (2021) found this to be the second better valued option for inclusive language by native teachers of Spanish after the <-e>, although its evaluation depended on what type of noun it was applied to. In contrast, it was the second least valued option by non-native teachers of Spanish. Finally, in a corpus study of inclusive Spanish in YouTube videos, Slemp et al. (2021) look at the frequency of some of the endings (-@, -x, -e) in written form as well as in oral form in these videos, across countries and across a period of time of ten years, from 2009 to 2019. They found -@ to be the first one to appear and the dominant one till 2015, but obviously only in writing.
The main obstacle to its adoption is the fact that it is unpronounceable; at most, the full symbol name can be pronounced without integrating it into the word pronunciation, having to pronounce either feminine or masculine and add later “arroba”. Besides not following a principle of economy, it also cannot be said to be a morphological ending if pronounced in this manner.
2.2. <-*> ending
One of the least visited options, only a few studies mentioned it: Bengoechea (2015), Reales Gil (2020), and the intersex activist from Argentina Mauro Cabral (2009), who prefers it over other options. Besides not being very popular or studied, it presents the same problem as above, it can’t be pronounced.
2.3. <-x> ending
One of the most extensively studied and discussed endings is the -x, a seasoned addition introduced around the 2000s (Borrell & Echeverría 2022). Gaining prominence post its use in “Latinx,” incorporated into Merriam-Webster in 2018, it sparked discussions mainly on ideological rather than linguistic fitness. Some authors (de Onís 2017, Murillo 2021) criticize its introduction as a linguistic colonialism from English to Spanish, with differing opinions on its ideological validity (Torres 2018, representing the Editorial Board of Latino Studies; Smith & Franco 2018; del Río González 2021). Thus, it is an option plagued with controversy.
In terms of self-declared use and preference, Shenk (2023) found it the least accepted and used among respondents, with 88.1% stating never using it orally and 80.7% in writing. Magagna's study (2021) showed native teachers' lower acceptance, while non-native teachers found it more acceptable than -@ but less than <-e>. A Pew Research Institute study (Noe-Bustamante et al. 2020) revealed low awareness and usage, particularly among Latin people in the US, with only 3% using it and 25% knowing about it, with a tendency for those who use it to be female, university educated, and younger5. In the YouTube videos study (Slemp et al. 2021), it appears in 2013 in their corpus and it is the dominant one since 2016, peaking in 2018, but only in written form, of course.
So while the <-x> ending has been quite popular in English, the same cannot be said for Spanish, as it is less recognized and accepted among Spanish speakers, even those in the US, and its true meaning and intentions have been controversial. In addition to this, it is either extremely hard to pronounce and against Spanish phonotactics or impossible, as it consists of two consonants and it would be typically added to a root ending already in a consonant. Thus, with this ending, a word like “chic-” (‘kid’) would phonetically be [‘tʃi.kks], and its plural [‘tʃi.kkss]. Spanish does not allow a syllable without a vowel or more than a two consonants’ cluster in one syllable. In fact, even in English, although for different reasons, there are proposals to go back to the term “Latin” (Salinas 2020) or even the French “Latiné” (Villanueva Alarcón et al. 2022).
2.4. <-e> ending
This option, argued to be initially proposed by García Meseguer (1976), is the first easily pronounceable. García Meseguer (1976) noted how words already ending in <-e> in Spanish are more inclusive than classically <-a/-o> gendered ones. For instance, words like “inteligente” or “grande” can be applied to any gender, although nouns such as “docente” or “estudiante”, when preceded by a determiner, still force the user to choose between one of two genders: el or la estudiante. Hence, a new ending in <-e> affecting all nouns and their determiners is proposed. The idea was first picked up in Argentina (Schmidt 2019), although not without opponents. For some feminists, it would undo the visibility of women in professions’ terms, such as “presidente”, as it would propose that no “presidenta” is used. And if it includes “presidenta” vs. “presidente” as options, it means the neutral and masculine would be conflated in one ending.
This option does fulfill some of the requirements for an ideal ending: it can be pronounced according to Spanish phonotactics and combines with roots easily. Furthermore, it is actually an ending that exists and conveys in some words the idea that both genders are being represented, so it goes along the lines of extending a currently less productive (more lexicalized) use. However, as we have seen, this posits other problems, such as less female visibility.
In terms of alignment with current paradigms in Spanish, this ending poses some problems. In particular, there are some Spanish pronouns that are a true neutral: the demonstratives esto, esto, aquello. In the demonstrative paradigm then, the ending <-e> signifies masculine, while the ending <-o> signifies neutral, and the ending -a signifies feminine. If the ending <-e> is adopted, we would have to either change the current demonstrative paradigm to make esto masculine and este neutral, or accept that the demonstrative paradigm is different from the article, noun, and adjective ending paradigms. Indeed, even the ending <-a> and <-o> for nouns are not consistently masculine and feminine, mainly because of different origins (Latin vs. Greek for instance), and only about 75% of nouns have a predictable gender according to their ending (Clegg 2011).
The other paradigm that would be disrupted is that of the clitic personal pronouns: le/les for dative case and lo/los and la/las masculine and feminine respectively for direct objects. Making le/les a signifier for direct object non-binary gender means that there is no distinction between direct and indirect object for this gender (although there would be for the other two genders). Thus, one option is to lose this case distinction for the third gender. Another option is to follow some minority dialects (Central and North Central Spain) and eliminate the case distinction entirely, substituting it by a gender distinction instead. Now, once again, this new dialectal paradigm uses <-e> for masculine, <-o> for neutral and <-a> for feminine (Echenique Elizondo 1981), as the demonstratives, and perhaps even by analogy with that established paradigm. This is right now not the most common dialectal option, and coming mainly from Madrid and the center of the Peninsula, it could be seen as a metropolis imposition both by Peninsular periphery and by American dialects. On the one hand, while for most linguists it would probably be sad to lose the last vestige of case in Spanish, it is a direction Romance languages have been veering towards. On the other hand, linguistic change that first leads to ambiguity may be dispreferred, although still certainly possible with later readjustments.
It must be noted that this proposal has gained traction among highly educated native speakers, especially in Argentina, where eight universities have endorsed it already. Despite its acceptance, the capital of Argentina has prohibited its use in schools, proving to be controversial also in Argentina. Among the options explored, the <-e> ending, particularly for LGBTQ+ inclusion, emerges as the most popular. Magagna (2021) noted widespread acceptance by both native and non-native teachers, in contrast to Shenk's (2023) findings, where native speakers preferred <-e> in speech but -@ in writing. Analyzing YouTube video data (Slemp et al. 2021), sporadic appearances have occurred since 2013, with increased frequency, albeit still trailing <-x>, in 2018 and 2019 in written form. Notably, videos using inclusive endings both in writing and orally, exclusively employed <-e> orally if the written form was -e, but less frequently than duplication (both binary genders) if it was <-x>. Moreover, <-e> was never used orally with -@, preferring a duplicated form again, suggesting this later symbol may specialize in non-sexist expression rather than third-gender expression.
2.5. <-i> ending
The final option is a new ending in <-i>. This option has not been mentioned in any of the works revised for this study, but it has been mentioned in conferences and conversations by several of the researchers, particularly those that investigated the topic after 2020, and it is one of the three options mentioned in the Gender in Language Project (genderinlanguage.com) started at UC Berkeley (Papadopoulos et al. 2022). It also appears briefly mentioned in the Nonbinary Wiki, in their entry for Spanish, and the site Pronombr.es.
Like the previous ending, this is one that can be not only pronounced but also easily added to the root of nouns, adjectives, and other words in the same way that <-o>, <-a> or <-e> are. So in morphological and phonotactic terms, this is as good a candidate as the previous one.
Unlike the <-e> ending, however, it does not figure yet in any paradigm, and because of this, it would be easier for it to have a one to one correspondence between form and meaning. Given that there are not, officially, many words ending in <-e> in Spanish, and the few ones are borrowings and have a stressed <-í> (i.e. colibrí) , for the most part, <-i> is “available” to get a new meaning. Because of this, it avoids the disruption of paradigms (like <-e> for the clitic pronoun paradigm) or contradictions between paradigms (like <-e> as masculine in clitic pronouns in leist, loist and laist systems, or in the demonstrative paradigm), and it is compatible with making women visible, by allowing <-e> to be masculine and therefore the introduction of <-a> for feminine in those professions named with a present participle, such as presidente or comandante. In those cases, we could have a male presidente, a female presidenta, and a non-binary presidenti.
One caveat is that it is not a form that would be extended in its use, but rather a new one. However, in oral informal uses of Spanish in some dialects, this ending is a familiar one, coming from a shortening of the diminutive: guapita or guapito > guapi. It is mostly used as a vocative: ¡Hola guapi! (hello handsome/beautiful) or ¿Cómo estás chiqui? (how are you, little one?). So the issue of what article to use has not arisen, but it shows how this is a form that is phonetically natural and already adopted to eliminate gender adscription. When adding an article, “li”, it would not create ambiguity with other current articles or pronouns, as the form “le” does.
Although it has not been as frequently proposed, it is starting to gain traction in some circles. For instance, in TikTok, a channel by an Argentinean influencer (Juana Sosa) proposed this option in 2021 and named it “lenguaje iconic”, reasoning that the ending is a lot “cuter” than the ending <-e>, which “no suena bien” (it does not sound good). The post had 10.9 million viewings, 2.2 million likes, and 37,300 comments, most of them positive, with only a few using the <-e> ending and a majority replicating the proposal in the comments, and several commenting they already use it.
3. Comparing the proposals to find the most optimal option
In order to compare and review all options together, Table 1 below shows pros of each option and issues for each option. Bold font and capital letters indicates a positive characteristic, while regular font and non-capital indicates a negative characteristic.
Table 1 Pros and cons of each third gender ending proposed for Spanish
| Ending | Ideal characteristics | Issues with perception | |||||
|---|---|---|---|---|---|---|---|
| Pronounceable | Extends current morpheme use | Does not disrupt paradigm | Does not disrupt function marking | US/Anglo imposition | Less female visibility | Less known** | |
| -@ | No | No | YES | YES | NO | NO | NO |
| -* | No | No | YES | YES | NO | NO | Yes |
| -x | No | No | YES | YES | Yes | NO | NO |
| -e | YES | YES | No | No | NO | Yes | NO |
| -i | YES | YES | YES | YES | NO | NO | Yes |
**According to its appearance in studies, since in terms of usage none of these options is often used by native speakers (for instance, one of the most often proposed and studied endings, <-x>, has been shown to be used by only 3% of Latin speakers in the US, being also dispreferred by speakers in the US according to Shenk 2023 and Magagna 2021).
If we just look at the number of positives and negatives, the options would be ranked as follows: -i > -@ > -x, -e, -*. However, it can be claimed that not all negatives (and positives) hold the same importance. For instance, not being pronounceable is probably the biggest barrier to the ending becoming an easy to adopt and more natural option, and hence this can be said to be a sine qua non condition. Thus, we should eliminate the 3 first options in the table, leaving us with two possibilities: <-e> and <-i>. Among the two, other than being less studied and less often proposed or investigated, <-i> is the option with fewer negatives and the only one that fulfills the ideal characteristics set at the beginning of this paper.
4. Conclusions and challenges ahead
As presented in the beginning of the paper, the proposal of a third gender in Spanish is a top-down approach defended currently by a minority, rather than a spontaneous change in the language motivated by some of the usual driving forces of linguistic change, such as economy, analogy, disambiguation, etc. That means that it may need an extra effort to become a reality, in the form of being embraced by normative forces, such as the education system and creators of norms in Spanish (mainly the academy), as well as software developers, media, etc.
In addition to its top-down direction, this is also the case because of the multiple roles of gender in languages like Spanish, which were mentioned above. Moreover, gender has a vast footprint in acquisition and processing. Gender has been shown to be acquired extremely early on in first language, together with the word, and consequently to have a big influence in processing, even more than more semantic features, like number. There are few studies that have focused on third gender processing (Zarwanitzer 2019), but they have shown third gender to slow down processing. This is probably an effect of the very narrow link between gender and word, and of frequency (third gender being not frequent at all currently), and may subside as a third gender becomes more accepted, used, and eventually automatized, but although surmountable, it is an effect that needs to be taken into account in languages with grammatical gender like Spanish.
In sum, a third gender is a better solution than no gender in order to be inclusive of all gender identities in flexive grammatical gender languages. However, the road to it is not an easy one, being a top-down type of linguistic change. It will need to be aided by norm, but it could also be aided by choosing an option that is best suited for the particular language system. This requires careful consideration of other elements and paradigms in the language, and it requires resisting the temptation to go with the first fashionable option. This paper is a first attempt to provide such analysis, which will need to be completed by studies on perception (both oral and written), and on processing and evaluating options that can be pronounced and therefore be integrated fully into the language system.
