A CAUTIONARY COMMENT ON THE MORPHOLOGICAL STATUS OF INCLUSIVE – E IN S PANISH 1

A BSTRACT . Inclusive means of expression in Spanish include the replacement of FEMININE and MASCULINE gender markers for the form –e . In recent literature, this element has been argued to be a third gender morpheme within the language. In this brief article, I contend that this hypothesis is premature, and that current evidence is actually compatible with a number of analytical alternatives. Given the complexity of the issue, I recommend taking a more cautious stance, relying on a combination of complementary approaches to achieve descriptive coverage of the phenomenon.


Introduction
Spanish is a language with two grammatical genders, FEMININE and MASCULINE. When these categories are in overt complementary distribution within a nominal inflection, they are systematically interpreted as categorizing entities as females or males, respectively. Thus, for instance, a nominal root like niñ-'child' may surface as niña 'girl' or niño 'boy' depending on whether the feminine inflectional marker -a or the masculine inflectional marker -o is attached to the stem. Additionally, the form -o also functions as a generic marker, allowing to refer to groups consisting of both female and male individuals, e.g., (1b). No matter if the noun is FEMININE or MASCULINE, determiners, adjectives and past participles establish concord in gender with it.
(1) a. las niñas tímidas the.FEM child.FEM shy.FEM 'the shy girls' 1 I am thankful to Andrea Bohrn, Fernando Carranza, Andrés Saab, Pablo Scolpino, Romina Trebisacce, Matías Verdecchia and two anonymous reviewers for discussion on the complex problem of offering a grammatical analysis for inclusive language. Special thanks go to my students at Universidad Austral de Chile for providing the acceptability judgements reported throughout the article. Usual disclaimers apply. b. los niños tímidos the.MASC child.MASC shy.MASC 'the shy boys/children' In recent years, "inclusive", "gender-neutral" and "non-binary" forms of expression have gained considerable traction among Spanish speakers. These consist on a number of strategies aimed to fight linguistic biases related to sex, sexual orientation and gender, e.g., the invisibilization of female individuals in utterances such as (1b) or the strict binarism of the gender paradigm. 2 Some of these strategies exploit lexical and grammatical resources that are already available in the language and conform the linguistic norm. These include the use of words that lack gender overt markers such as gente 'people' or persona 'person', and the explicit mention of different genders through coordinating structures, e.g., (2). Following and adapting the terminology in López (2019López ( , 2020, I will encompass these strategies under the label of indirect inclusive language. 3 (2) las niñas y los niños the.FEM.PL child.FEM.PL and the.MASC.PL child.MASC.PL 'the girls and the boys' Patterns such as (2) are in stark contrast with what can be called, also following López's terminology, direct inclusive language. Strategies in this line involve the intervention of standard grammatical forms through linguistic or paralinguistic means, with the explicit objective of demonstrating that inclusion has priority over language norms (López 2019); this is achieved, for instance, through the use of x in written words such as niñx 'child'. In particular, I will focus on the phenomenon exemplified in (3), in which the morphological slot reserved for the gender markers -a and -o in (1) is occupied instead by the form -e. I will employ the term Direct Inclusive Spanish (DIS) to designate the varieties that (i) productively employ this sort of marking and (ii) differ in this way from General Spanish, which optionally exhibits indirect inclusive patterns such as (2). As can be seen in (3), the typical interpretation of nouns exhibiting inclusive -e is parallel to that corresponding to the generic use of the form -o in (1b), i.e., it encompasses individuals of any gender.
(3) les niñes tímides the.INC child.INC shy.INC 'the shy children' In the incipient linguistic discussion about DIS, it is seemingly customary to assume that patterns such as (3) illustrate the emergence of a new gender morpheme. 4 That is, the form -e is taken to be the exponent of a meaningful unit of morphosyntactic combination introducing a gender value that is distinct from FEMININE and MASCULINE. For instance, Menegotto (2020) distinguishes two Spanish dialects based on the number of genders they exhibit: in short, she argues that General Spanish has two genders, while DIS makes a tripartite distinction between FEMININE,MASCULINE and INCLUSIVE. 5 Similar ideas can be found throughout the literature on the phenomenon, as the element exemplified in (3) is commonly referred to as "the -e morpheme" (e.g., Romero & Funes 2018, Martínez 2019, Pérez & Moragas 2020, Tosi 2019, Zunino & Dvoskin 2022. In this brief article, I contend that this line of grammatical analysis is premature considering the data that is available at the moment. That is, while there is nothing particularly wrong with the hypothesis that -e is a gender morpheme, the basic DIS patterns are also compatible with a number of competing alternative accounts, i.e., the fact that -a and -o are replaced with -e in (3) can be explained in other ways. Moreover, there seems to be no analytical proposal that by itself can fully account for the distribution of the form -e in real speech, as the functioning of this element exhibits considerable variation among speakers. Therefore, I suggest that a more cautionary approach is preferable for now: given the complexity of the data, it is better to rely on a mix of analytical tools that complement each other in order to achieve descriptive coverage of the phenomenon.
The structure of the paper is as follows. Section 2 describes a trait that, in my opinion, complicates our comprehension of DIS grammar: the fact that DIS speakers are "bilingual". In section 3, I discuss a number of alternative and equally plausible accounts for DIS. Section 4 contains the conclusions.

An obstacle for the study of DIS
I believe there is a basic problem obscuring DIS phenomena and preventing us from testing hypotheses regarding its grammatical functioning. The problem lies on the fact that all DIS speakers are, in a sense, "bilingual": they are native speakers of (General) Spanish and L2 speakers of DIS. 6 This means that, in principle, they have two distinct set of rules concerning gender inflections, one that generates General Spanish nouns, with FEMININE and MASCULINE inflections, and one that generates DIS nouns, with inflections in -e. The question is whether the DIS grammar includes or not the rules of General Spanish. That is, when a DIS speaker produces a sentence like (1b), we do not really know whether they are using their knowledge of General Spanish or their knowledge of DIS.
For concreteness, consider the example in (4), in which the markers -a, -o and -e appear together in the same structure. One could assume that this pattern proves that DIS has three gender markers. However, once we admit that DIS speakers are bilingual, other analytical options arise. For instance, it could 5 Since there are no native speakers of DIS, Menegotto presents her account as a "prospective analysis" of an Ilanguage based on data from a current E-language; see (Chomsky 1986) for the definition of these terms. 6 There are many further issues regarding this situation of "bilingualism" that I will not address here. In particular, the sociolinguistic factors triggering the switch between both grammars are far beyond the scope of this paper. This is an issue of much importance since, as pointed out by a reviewer, there seems to be no community of speakers that spontaneously use -e forms in every context. be the case that the speaker generates the first two nominals by applying the rules of (General) Spanish, and then switches to DIS to generate les niñes 'the children'. This possibility is sketched in (5).

General Spanish DIS
As we will see below, a potential way of understanding DIS is as a system that generates human nouns in -a and -e, but not in -o, cf. (17). Such hypothesis can also accommodate the pattern in (4) by appealing to code-switching, i.e., perhaps DIS speakers use General Spanish specific rules to generate los niños 'the boys', and DIS specific rules to generate las niñas y les niñes 'the girls and the (non-binary) children'.

General Spanish DIS
These examples show that there is no real way to tell what is the reach of the DIS grammar from examples such as (4). As a consequence, we cannot disregard many potential accounts of DIS, i.e., these theories of how the DIS grammar works become unfalsifiable. Since this problem arises quite systematically when considering analyses for DIS, I will give it a name.
(7) THE TWO GRAMMARS PROBLEM If General Spanish evidences a certain grammatical rule R, we cannot be certain that DIS grammar also has R. It could be that DIS speakers are just code-switching with General Spanish.
This implies that the only patterns that truly inform us about the DIS grammar are those that are unmistakably generated through it, i.e., expressions containing -e and derived phenomena, e.g., inclusive pronouns such as elles 'they'.
In the following section, we will see how the problem in (7) prevents us to empirically discern from a number of alternative analyses of DIS.

Analytical alternatives
In what follows, I discuss five ways of accounting for patterns such as (3). I will go from the most commonsensical and superficial approaches to the ones that imply an extensive rearrangement of the Spanish grammatical system. In all cases, the presentation is meant to be schematic, as it only aims to offer a flavor of the types of analysis that are conceivable in principle. The list of analytical alternatives sketched here does not pretend to be exhaustive. Moreover, I will restrict the discussion to the distribution of nouns exhibiting inclusive -e, and I will leave aside further phenomena related to agreement and to inclusive pronouns, e.g., elles 'they.INC'. In particular, I will not delve into the mechanisms dealing with gender concord. I will simply assume that it affects elements in the morphosyntactic context of a certain noun.
Almost at a pre-theoretical level, the most immediate explanation for patterns such as (3) involves assuming that DIS speakers consciously replace the sounds a and o in (1) for an e. This intuition can be informally captured as in (8). As can be seen, this is a rather simplistic mechanism that makes no reference to morphosyntactic primitives, i.e., (8) is not strictly speaking a grammatical rule. Unsurprisingly, this rule does not work: it wrongly predicts that vowels a and o can be replaced no matter the grammatical context in which they appear. Consider the following example. The word casa 'house' is inherently FEMININE in Spanish, i.e., its final -a is not a gender marker. Thus, for instance, it does not inflect for MASCULINE, e.g., (9b).  Even if the rule in (8) does not reflect the linguistic competence of (fluent) DIS speakers, it does have some use at describing certain phenomena. For instance, detractors of DIS satirize inclusive speech by overusing (8), e.g., they produce utterances like (10) for an intended burlesque effect. DIS speakers are very aware of these missuses and have developed norms explicitating the grammatical contexts in which the form -e can be properly employed. These norms have appeared in different media and have been compiled as part of guidelines on inclusive communication, e.g., Gómez (2016), Mascías (2018). In a nutshell, they state that for -e to appear on a certain noun (and its agreeing elements), (i) the noun must refer to a human entity, and (ii) the noun must inflect for gender. It is far from obvious whether these norms are entirely prescriptive and, in a sense, artificial, or whether they reflect a grammatical intuition about the distribution of gender markers. The fact remains that they allow to capture the speakers' judgements about the unacceptability of (10). These would not exist as norms if there were no people employing the surface rule in (8).
Errors in DIS speech are another domain in which the rule in (8) could be useful at a descriptive level. Some non-fluent DIS users produce utterances that are unanimously rejected by fluent DIS speakers. Consider the examples in (11), which are both taken from real use: (11a) is an oral example, while (11b) was written on a sign. The noun adjetivo 'adjective' is inherently masculine in Spanish, e.g., *adjetiva 'adjective.FEM'. However, a speaker spontaneously produced (11a) while attempting to employ DIS; notice that this example exhibits the same deviation as (10), and thus it should be equally unacceptable. As for (11b), it displays the inclusive form -e over the morphologically invariable adverb pronto 'soon', which is not supposed to agree in gender with the subject of the sentence, i.e., there is no conceivable grammatical reason for that -e to appear there. The nature of these mistakes suggests that some non-fluent DIS users do rely on surface rules. 7 Finally, surface rules like (8) are responsible for several neologisms found in DIS. In these cases, the inclusive version of a word is not due to a productive rule, but rather due to a conscious intervention of a lexical item. To illustrate the idea, consider first the following examples. In Spanish, cuerpo 'body' is an inherently masculine noun that does not inflect for gender; as a form of demonstration, the feminist collective has coined the term cuerpa 'body.FEM' to refer exclusively to the female body. A similar alteration is attested with utero 'uterus' and its corresponding neologism utera 'uterus.FEM'. This is not a productive mechanism, as it does not apply to nouns referring to other parts of the anatomy, e.g., dedo 'finger' is also a masculine noun, but there is no *deda 'finger.FEM'. In sum, cuerpa and utera obtain through the deliberate manipulation of the form of two words, and not through morphosyntactic means.
This line of analysis applies straightforwardly to certain DIS expressions. Consider the case of cuerpes 'bodies. INC' in (12). This neologism may refer to female, male and non-binary bodies, just as the noun cuerpos 'bodies' does in General Spanish; the difference between them is that cuerpos formally looks like a MASCULINE noun and also behaves like one regarding agreement. Arguably, cuerpes obtains through a purposeful change of the noun cuerpos by applying the surface rule in (8). This process is not productive, as it does not apply, for instance, to the noun úteros 'uteruses' to form úteres 'uteruses.INC', e.g., (13). The interim conclusion is that we cannot totally dispense with surface rules if we want to account for the whole corpus of inclusive expressions in DIS and beyond. Surface rules allow us to capture recurring errors in DIS speech, and also provide a rationale for the creation process of inclusive neologisms. Thus, a grammatical analysis of DIS needs to be complemented with a rule such as (8) in order to cover these phenomena.
Let's go back to the contrast between niñe 'child.INC' in (3) and case 'house.INC' in (10). As mentioned, the difference in acceptability between these examples relies on whether the nouns inflect for gender, i.e., niñ-'child' does, but casa 'house' does not. This is a grammatical distinction that calls for an explanation in grammatical terms. As said, there are different ways to do so. The first alternative that comes to mind builds on the functioning of so-called neuter gender in Spanish. Consider the copular sentences in (14). In this context, the adjective should agree with the subject of the sentence, i.e., with the infinitival subject hablar 'to talk'. Infinitives do not have gender features in Spanish. In this case, the adjective must surface in its MASCULINE form for the sentence to be acceptable. Patterns such as this have led to the traditional idea that MASCULINE is the gender value by default in Spanish.
*Hablar es divertida. to.talk is.3SG fun.FEM At least since Harris (1991), the standard way of accounting for this involves assuming that gender in Spanish is an asymmetric system. Roughly speaking, this means that FEMININE is the only true gender value; this is typically represented as a privative feature [+f]. Nominal inflections lacking a [+f] feature must be spelled-out as -o, e.g., (1b) and (14a). The informal rules in (15) capture the gist of this system for General Spanish.
[+f] → -a b. [ The "no gender" approach to -o can be adapted to account for DIS patterns. That is, -e could be taken to be one of the exponents for the lack of gender specification, i.e., -e also spells out [-f]. The intuition is captured in (16).

(16) INCLUSIVE AS NEUTER
The -e form appears when the inflection of a noun referring to a human entity lacks a [+f] feature. This is to say that the rule in charge of introducing -e is a more specific version of (15b), i.e., one that only applies in the context of human-referring nouns. When there is no noun referring to human entities, or when there is no noun at all, -o is introduced instead. The relevant rules are sketched in (17); these must be considered hypothetical principles that could be taken to pertain to the grammatical competence of DIS speakers.
[+f] → -a b. [ These rules allow to capture contrasts such as the one in (18). As can be seen in this example, the inclusive form -e only appears in adjectives agreeing with a human noun. There is a potential drawback in this analysis: it basically predicts that DIS lacks a featural specification for MASCULINE forms. That is, since nouns referring to human males are supposed to be specified as [-f], they should always be spelled-out with -e in accordance with (17c). In principle, this should indicate that the rules in (17) are wrong, as DIS speakers can produce and understand expressions such as (1b). Indeed, we should be able to arrive to the conclusion that the proposal does not work, but we are not: there is still the possibility that the DIS grammar functions as described in (17), and that DIS speakers resort to their (General) Spanish knowledge to form masculine nouns. This is basically the scenario already depicted in (6). Thus, despite its obvious shortcomings, we are able to maintain (17) as a working hypothesis.
Let's move to a different type of account for the pattern in (3). The second hypothesis to consider involves assuming that DIS applies some mechanism that affects the exponence of FEMININE and MASCULINE. This idea is captured in (19).

(19) INCLUSIVE AS SYNCRETISM
The -e form is the result of eliminating a distinction in exponence between FEMININE and MASCULINE in the nominal inflection.
According to (19), -e is a form that replaces the exponents -a and -o in nominal inflections. Under this approach, while General Spanish makes an overt binary distinction between MASCULINE and FEMININE through two different gender markers, DIS neutralizes this opposition and collapses their exponence into a single element.
This line of analysis equiparates inclusive -e to several forms of syncretism found across languages. For instance, consider the following examples taken from Arkadiev (2009: 107). In Russian, nominative adjectives inflect for masculine, feminine and neuter, and have different markers in singular, e.g., (20). However, this distinction disappears with plural adjectives, as the three genders receive exactly the same exponence, e.g., (21). The pattern in (21) can be captured at a descriptive level through a rule like (22). Basically, it states that no matter a gender feature is valued as FEMININE, MASCULINE or NEUTER, it will be spelled-out as -ie in the context of a plural adjective.

(22) {[FEM]/[MASC]/[NEUT]} → -ie / A[PL]
A similar rule can describe the core distribution of inclusive -e. Consider (23). This informal rule expresses that no matter a gender inflection is specified as MASCULINE or FEMININE, it will be spelled out as -e in the context of nouns that refer to human entities.

(23) {[+f]/[-f]} → -e / N[+HUMAN]
The intended effect of this rule at the surface level is similar to that of (8), as it introducese in the phonological representation instead of the exponents -a and -o; the advantage of this alternative is that the replacement is constrained to cases in which -a and -o are inflectional gender markers.
As a potential disadvantage, the rule in (23) should apply in all cases, making -e the only gender exponent in the variety. Since DIS speakers can also employ -a and -o, this should be enough to disregard a rule like (23). However, the "two grammars problem" attacks again: perhaps the DIS grammar only has one exponent for gender, and when DIS speakers produce overtly feminine or masculine forms they employ General Spanish rules. Thus, this issue is not enough to falsify (23).
As for the interpretation of -e, it can also be captured under this approach. The meaning of examples such as (3) follows straightforwardly from an underlying [-f] feature that is also responsible for the generic interpretation of (1b). Notice that under this analysis, -e should also be able to refer to either male or female entities. This prediction is partially borne out: while DIS speakers report a strong preference for interpreting (3) as gender-unspecific, they can also accept the example as referring to a group of female children or male children. The preference for the generic interpretation can be explained in standard Gricean terms: since the speaker has available the more specific forms -a and -o to convey FEMININE and MASCULINE, the use of -e for any of these interpretations goes against the Maxim of Manner (Grice 1975).
Moreover, there are grammatical patterns supporting an approach to DIS in terms of syncretism. Consider first the General Spanish sentences in (24). As can be seen, General Spanish speakers do not tolerate mismatches in gender in these contexts.
*Jorge es desordenado y Javier es obsesivo, pero igual Jorge is.3SG untidy.MASC and Javier is.3SG obsessive.MASC but same están enamoradas. are.3PL in.love.FEM 'Jorge is untidy and Javier is obsessive, but they are in love anyway.' b.
Jorge es desordenado y Javier es obsesivo, pero igual Jorge is.3SG untidy.MASC and Javier is.3SG obsessive.MASC but same están enamorades. are.3PL in.love.INC 'Jorge is untidy and Javier is obsessive, but they are in love anyway.' b.
Jorge es desordenade y Javier es obsesive, pero igual Jorge is.3SG untidy.INC and Javier is.3SG obsessive.INC but same están enamorades. are.3PL in.love.INC 'Jorge is untidy and Javier is obsessive, but they are in love anyway.' There are further aspects that could be incorporated into the schematic rule in (23) to make it more descriptively adequate. For instance, Kalinowski (2020) reports that around 94% of the uses of -e found in his corpus involve plural nouns. Within the line of analysis sketched in (23), this tendency strongly suggests a connection between number and gender exponence parallel to the one represented in (22). That is, perhaps the context for -e insertion should be constrained to plural nouns, which are overwhelmingly more productive, while occurrences of this form in singular contexts should be treated as neologisms.
The analyses depicted in (16) and (19) share an important similarity: they are both attempts to capture the differences between General Spanish and DIS in terms of distinct rules of exponence for gender markers. That is, these hypotheses do not posit stark distinctions between the grammars of these varieties; instead, the contrasts arise at the externalization level. Since both approaches posit "small differences" between General Spanish and DIS, they are, in a sense, microparametric in the terminology of Kayne (2005). Indeed, both proposals could be understood as involving a slight variation on how phonological rules handle grammatical features such as [+f] and [-f], which would be common to General Spanish and DIS.
In contrast, other ways of approaching DIS are, employing the term in Baker (2008), macroparametric in nature. That is, they posit that the distinction between General Spanish and DIS is one of typological magnitude. While such a claim may sound excessive in principle, this is exactly what follows from premises such as (26).
(26) THREE GENDERS DIS has a system with three different genders.
According to (26), General Spanish and DIS differ in how many grammatical genders each have. Thus, General Spanish is a two-gendered language, such as Hebrew, Tagalog and most of the Indo-European family. On the other hand, DIS has three genders and, therefore, must be grouped together with Russian, Icelandic and Tamil, among others; see Corbett (2013) for further details on this typology. Since a single privative feature [±f] cannot capture the functioning of three distinct genders, DIS must have a substantively distinct inflectional system for nouns. 10 For simplicity, I will assume that the relevant geometry for gender under this approach involves a symmetric representation with three values, as sketched in (27). (27) This is essentially the approach adopted by Romero & Funes (2018), Martínez (2019) and Menegotto (2020), among others. These authors contend that -e is the exponent for a gender morpheme that is grammatically distinct from FEMININE and MASCULINE: the -e morpheme is taken to denote (i) groups of human individuals with different genders or (ii) non-binary referents (or individuals whose gender is unknown); thus, its interpretation is proposed to be distinct from that of -a and -o, which are assumed to refer to female and male entities exclusively. As mentioned, this line of analysis is supported by patterns such as (4), in which nouns with the three genders in (27) are coordinated and form a grammatical expression. Moreover, each nominal inflection introduces distinct interpretations for the root niñ-'child'.
There are reasons to doubt an analysis of DIS as having three gender values. First, the proposal seemingly relies on the observation that DIS speakers produce nominal inflections with -a, -o and -e. However, this does not control for the "two grammars problem" depicted in (7): it is possible that expressions such as (4) involve code-switching. If this is the case, then a DIS speaker could employ rules of General Spanish to produce FEMININE and MASCULINE nouns and switch to DIS to produce nouns with inclusive inflections, e.g., (5). Thus, examples such as (4) cannot prove by themselves that the DIS grammar has three genders.
Second, the tripartite system in (27) predicts a clear-cut distribution of the exponents -a, -o and -e depending on the gender of the nominal referent. This is not always borne out in the data. Consider the patterns in (28), both of them produced by a publicly known DIS speaker from Argentina. In General Spanish, the noun padre is ambiguous between 'father ' and 'parent'; according to (26), in DIS it should trigger MASCULINE concord on the first interpretation and INCLUSIVE on the second. However, (28a) has the noun padres arguably referring to male entities and triggering the inclusive form of the determiner les. Similarly, the noun docente 'teacher' in (28b) triggers inclusive concord on the determiner even when context suggests that it refers to female entities.  (27) treats -e as a gender morpheme within the same paradigm as feminine -a and masculine -o. However, there are cases in which inclusive -e appears with nouns that do not inflect in gender. These are semantically neuter human-referring nouns such as persona 'person', sujeto 'subject' or gente 'people'. Consider the following example. As the pair in (29) shows, the noun personas 'persons' behaves in exactly the same way as the noun casas 'houses' in (9): it has inherent FEMININE gender and is unable to host gender markers.
*los personos the.MASC persons However, the inclusive form of the noun is highly productive, e.g., (30). Interestingly, DIS guidelines such as Gómez (2016) explicitly ban attaching -e to nouns that do not exhibit the contrast between FEMININE and MASCULINE. In fact, many DIS speakers know that examples such as (30) are marked according to the DIS norm, but notice that "people used them anyway". Examples such as (30) are problematic for the hypothesis that -e is a third gender morpheme, but also for the lines of analysis depicted previously in (16) and (19). That is, these cases show that the inclusive -e phenomenon is not totally restricted to nouns exhibiting inflectional gender.
A further troublesome pattern for the hypotheses in (16), (19) and (26) involves the DIS forms of nouns such as mujer 'woman' and hombre 'man'. These nouns do not inflect in gender either but, unlike persona 'person' or gente 'people', they refer to female or male entities exclusively, so they should be either FEMININE or MASCULINE. Some examples taken from the internet are offered in (31). These also go against DIS norms; some of my informants consider them unacceptable. As the glosses illustrate, there is seemingly no interpretative effect intended in employing inclusive -e with these nouns, i.e., they still refer to either females or males alone. A similar point can be made regarding cases like (30), in which -e does not make the denotation of the noun any more gender-neutral than it already is. Thus, these data suggest that, at least in certain contexts and for some speakers, inclusive -e is just a form to be used with any human-referring noun, no matter what the (non-grammatical) gender of its referent is. One way of capturing this observation is through the hypothesis in (32).

(32) ONLY ONE GENDER FOR HUMAN ENTITIES
The DIS grammar has only one gender for all [+HUMAN] nouns.
This possibility represents an even greater departure from the General Spanish gender system than the one sketched in (26). According to (32), the DIS grammar distinguishes between FEMININE and MASCULINE only for non-human nouns. All nouns denoting human entities are inherently INCLUSIVE. That is, nouns like niñe 'child.INC', persone 'person.INC' do not have FEMININE or MASCULINE counterparts within DIS; the same applies to mujer 'woman' and hombre 'man': their only gender is INCLUSIVE. 16 All instances of human-denoting FEMININE and MASCULINE nouns produced by DIS speakers would be due to code-switching with General Spanish. The representation in (33) sketches the organization of the DIS gender system according to this hypothesis.
Alternatively, patterns like (30) and (31) could be dealt with under the assumption that these forms obtain through surface rules such as (8), just like cuerpes 'bodies. INC' in (12); this solution is compatible with any of the hypotheses sketched in (16), (19) and (26). However, there is no obvious way to make an evidence-based decision between these approaches. In principle, all analytical options are equally plausible.
The variation attested between the examples used throughout this discussion deserves a separate comment. The acceptability judgments were collected directly from DIS speakers; there was virtually no variation regarding their reports. However, some of them reject the patterns that I took from websites, e.g., (31). My informants acknowledge the existence of these forms, but attribute them to other communities of DIS speakers; some even correlate these differences with political stances on the role of inclusive language in society. In any case, future research on DIS will have to account for what seems to be dialectal variation within DIS.

Concluding remarks
I have argued that there are different ways in which inclusive -e can be analyzed. As far as I can tell, there is no empirical reason to prefer one alternative over the others: potential codeswitching in DIS speech prevents reaching any definitive conclusions. In particular, I see no evidence unambiguously supporting the claim that DIS has three gender values. The available data does not even seem to support the idea that DIS phenomena must receive a single and unified treatment.
In light of the intricate nature of the DIS data, I believe it is prudent to adopt a cautious approach. Rather than immediately pursuing a single line of analysis to explain the grammatical behavior of DIS, a combination of complementary hypotheses could be used to attain descriptive coverage of the phenomena. This would enable a more holistic and nuanced understanding of DIS, allowing for the identification of patterns and relationships that may have been overlooked. This approach might provide a more robust and reliable foundation upon which to make informed theoretical claims.
A perhaps too common opinion among non-linguists is that inclusive -e is "unnatural" for the grammatical structure of Spanish (whatever that means). On the contrary, this brief article shows that there are many potential ways in which this element could be incorporated in the grammar of the language.