Gender is a complex grammatical category as it conflates elements that are purely linguistic in nature as well as social and cognitive aspects of language use and processing. From a linguistic perspective, in grammatical-gender languages like Italian and French, gender is a morphosyntactic feature that classifies nouns as masculine or feminine within a binary grammatical gender system (Corbett, 1991). In Italian, nouns referring to people can be gender transparent, namely, they can express a consistent mapping between their linguistic form and their referent’s gender (maestro_[MASC], ‘male teacher’; maestra_[FEM], ‘female teacher’) based on an attested regularity between a word ending and its grammatical gender (Padovani and Cacciari, 2003). Other nouns can be gender opaque, namely their ending is not informative with respect to their referent’s gender since they can both refer to a female and a male referent (cantante_{[MASC, FEM]}, ‘male/female singer’). Previous psycholinguistic research in gender-marked languages has shown that grammatical gender is mastered early in language acquisition (Mornati et al., 2023) and is immediately integrated during language processing to establish coordinated agreement across different elements in the sentence (Bambini and Canal, 2021). Indeed, gender traits are activated during sentence processing to ensure agreement at various levels, such as within noun phrases (e.g., article-noun or adjective-noun agreement) and across verb-subject pairs (Chomsky, 1995; Levelt et al., 1999). When a word is processed, gender information, if grammatically relevant, is activated immediately at the morphosyntactic level during the representation of the lemma in the mental lexicon (Schriefers et al., 1999). At the syntactic level, grammatical gender is then used for establishing agreement relations during sentence processing. In the case of agreement mismatches, several psycholinguistics studies show early effects of disruption during incremental processing in gendered languages (Molinaro et al., 2011), indicating that gender information is used in the first stages of parsing. Moreover, late measures of processing suggest that reanalysis is costlier after grammatical gender mismatches than number mismatches (Horacio and Carreiras, 2005). The different behaviors observed in the case of mismatch in gender or number features provide a key to understanding the interplay of lexical and conceptual features during sentence processing.

Nouns that refer to people might also be distinguished on the basis of semantic gender, that is the conceptual representation of the noun as referring to a man or a woman. Being morphosyntactic and semantic layers distinct, there is no necessary and straightforward relation between the morphological form of a noun referring to human referents and their gender (Corbett, 2012). Previous experimental studies on gendered languages have shown that morphosyntactic gender on nouns is a reliable cue for resolving agreement dependencies in sentences and may override semantic gender agreement. For example, a study by Cacciari et al. (1997) showed that grammatical gender in epicene words in Italian (like la vittima, ‘the victim’, which is grammatically feminine but can refer to male or female individuals) took precedence over the semantic agreement in anaphoric dependencies: slower reading times of the masculine pronoun lui (‘he’, compared to lei, ‘she’) were recorded following a grammatically feminine epicene noun, despite the fact that, semantically, this noun can refer to a man. Languages that do not exhibit explicit grammatical gender traits on nouns present a complex picture regarding gender processing as well. In these cases, gender processing can occur at the semantic or pragmatic level, even in the absence of linguistic gender cues. For instance, in an ERP study conducted on English, Canal et al. (2015) showed that mismatches between the expectations about the referent’s gender (i.e., gender stereotypes) and the linguistic input (e.g., references to counter-stereotypical characters like female engineers) require additional cognitive effort, observable through neural responses like the P600 or N400 effects. These results highlight that the initial representation of the referent’s gender reflects both the information conveyed by the noun’s grammatical gender and the gender typically associated with the referent’s role. The effects of gender stereotypical associations on the processing of role nouns have been investigated in different languages, with evidence pointing to the automatic activation of stereotypes in both word and sentence reading (Banaji and Hardin, 1996; Cacciari and Padovani, 2007; Casado et al., 2023; Osterhout et al., 1997; Pesciarelli et al., 2019; Siyanova-Chanturia et al., 2015). In Italian, gender stereotypes can surface on opaque role nouns particularly (Cacciari and Padovani, 2007). For instance, the gender opaque nouns badante (‘caregiver for the elderly’) and falegname (‘carpenter’) can be both used to refer to male and to female referents, but the former is generally associated with female referents, while the latter is typically associated with male referents. On the other hand, some nouns, such as cantante (‘singer’), are not generally associated with any specific gender. Furthermore, most bigender nouns end in -a; despite their ending, which is the typical grammatically feminine word-ending, some bigender nouns could show stereotypical associations with male referents (e.g., camionista, ‘truck driver’).

Recently, research on gender has focused on forms described as ‘inclusive’, i.e., intended to overcome gender binarism in language. The so-called Gender Inclusive Language (GIL) has been promoted over the years in several countries by activists and social movements, like some feminist circles and LGBTQIA+ communities. A public debate on GIL has emerged, and different linguistic strategies have been proposed, depending on how each language conveys gender. In natural-gender languages, like English or Swedish, which mainly express gender through agreement on personal pronouns, there has been an emphasis on using pronouns that may refer to a person of any gender (e.g., English singular they), including pronouns that have been actively created for the purpose of language inclusivity, such as English ze and Swedish hen (e.g., Arnold et al., 2021; Lindqvist et al., 2019; Sanford and Filik, 2007; Vergoossen et al., 2020). In gendered languages, such as Italian, French, and German, role nouns can express a relation between their form (i.e., their word ending) and their referent’s gender. As discussed above, the Italian words maestro and maestra refer to a male and female teacher, respectively, by virtue of their endings (-o, -a), which mark masculine and feminine gender. Similarly for the pair Lehrer and Lehrerin in German. In such languages, gender-inclusive forms have been proposed as alternatives to generic masculine role nouns and double forms (see Körner et al., 2022 for review). In German, for example, building on the fact that the feminine plural ending is often -innen, the female form is used, but the lowercase i is replaced by a capital I (e.g., LehrerInnen instead of Lehrer_[MASC] and Lehrerinnen_[FEM]) or it is preceded by a star symbol (e.g., Lehrer*innen). In French, the most common form of GIL uses a median point inserted after the word stem, followed by the feminine ending -n and, for plurals, by an additional median point and -s (e.g., citizens is spelled citoyen・ne・s, instead of citoyens_[MASC] and citoyennes_[FEM]; Burnett and Pozniak, 2021).

Since in Italian morphological endings generally provide reliable gender cues, one of the proposals of GIL recommends substituting the word ending of nouns referring to people with the schwa (‘ǝ’) symbol (e.g., avvocatǝ, ‘lawyer’, instead of the masculine and feminine forms avvocato and avvocata; Gheno, 2021).¹ Although other interventions have been proposed for Italian (such as using the asterisk ‘*’ as a neuter ending of words, which is widespread in written communications in substitution of generic masculine forms, e.g., car* tutt* instead of cari tutti, ‘dear all’),² one of the merits of the schwa proposal is its applicability to spoken language as well. This proposal has gained increasing attention in Italy in the last few years, with several attempts to advocate for or against it, with great media coverage in newspapers and social media. Linguists and experts of language, including representatives of the Accademia della Crusca (which is a recognized national institution for matters related to the Italian language), intervened more than once in the debate on the introduction of GIL forms, with official communications against their introduction. Several opinion contributions have also been published by linguists and opinion makers to express concerns on the feasibility and necessity of this proposal (e.g., Giusti, 2022; Arcangeli, 2022; De Benedetti, 2022; Robustelli, 2021).³ The main argument of the opponents is not focused on preserving the language from novelty or change, which is a natural part of any language’s evolution. Instead, they emphasize economic principles, such as the need to strengthen existing linguistic tools that promote inclusive language. For instance, they advocate using masculine generic forms or gender-neutral terms when referring to mixed groups of people.⁴ Most importantly, they argue that introducing a non-standard grapheme and phoneme would unnecessarily complicate the morphological, phonological, and orthographic structure of the language. In this respect, it is important to note that Italian has a rich agreement system in which a change in the gender of the nouns does not remain at the level of the lexicon, but it spreads potentially over all the elements in the sentence (articles, adjectives, and verbs), via gender agreement. This, in turn, might heavily impact the readability of texts, with potential increased difficulties for people with reading impairments or more fragile populations, such as the elderly.

Similar concerns about the introduction of GIL forms have been raised in other European countries by equivalent recognized institutions, such as the Academiè Francaise in France, with political consequences as well (such as the ban of the ‘écriture inclusive’ in schools by France’s education minister at the time, in July 2021). More on the sociolinguistic side, the interplay between linguistic change, linguistic prescriptivism, and ideology is a long-debated and important topic (cf. Silverstein, 1985; cf. also Burnett and Bonami, 2019 for a recent analysis focused on France).

In the last decade, several studies have been carried out with speakers of different languages to investigate the efficacy of neutralized forms, providing mixed results. Körner et al. (2022), for example, presented German participants with the first part of a sentence containing a role noun, either in the generic masculine, i.e., masculine form with an underspecified meaning, or in the neutralized star form, and asked them to rate the acceptability of a continuation sentence that disambiguated the gender of the role noun. Sentences following generic masculine forms were more often and more quickly judged to be compatible with male referents compared to female referents, whereas the opposite was true for neutralized forms, which seemed to carry a feminine bias. Using a similar paradigm with French speakers, Tibblin et al. (2023) similarly found that contracted double forms with a similar surface form to feminine nouns, such as les voisin·es_[MASC,FEM] (‘the neighbors’) carried a feminine bias. However, the opposite was true for forms such as le voisinage (‘the neighborhood’), although in both cases the bias was smaller than the one observed for generic masculine forms. In Zacharski and Ferstl (2023), German speakers judged whether pictures with male, female, or non-binary connotations were appropriate representations of role nouns presented in the masculine, feminine, or neutralized form. In this case, participants quickly accepted pictures of all genders in association with the neutralized form, suggesting that the intervention worked as intended.

To the best of our knowledge, no study has investigated the efficacy of schwa in neutralizing the referent’s gender information conveyed by the grammatical gender on Italian role nouns. Italian provides an interesting testing ground to assess the effects of GIL interventions. First, Italian is characterized by an overall transparent orthography (including regular and transparent gender associations) and a rich morphosyntactic agreement system, as discussed above. Thus, any change introduced at the phono-orthographic level impacts higher levels of linguistic processing. This does not happen in natural-gender languages, like English, in which the only proposed change pertains to the use of the plural pronoun (they) as singular to refer to non-binary individuals or to be used in cases in which neutral (ungendered) forms of pronouns are warranted. Second, schwa replaces standard suffixes with a non-letter symbol in writing (and a non-standard linguistic phoneme in the oral modality) with a specific linguistic function whose effects need to be empirically assessed. This proposal differs from GIL proposals in other gender-marked languages like French or German, in which a non-letter symbol is used to emphasize the juxtaposition of feminine and masculine linguistic forms, which nevertheless remain in their canonical form.

Beyond these general concerns, there are at least two other potential issues to face with respect to the introduction of ǝ. First, its visual similarity with the feminine ending -a (or the feminine plural -e). In the written modality, this may induce a female bias, namely, the preferred activation of a female representation when referents of any gender are grammatically acceptable, exacerbating a phenomenon already observed in other studies on GIL (e.g., Körner et al., 2022; Tibblin et al., 2023). Second, the schwa phoneme does not belong to the standard Italian phonemic inventory except in specific southern dialects (Bertinetto and Loporcaro, 2005). Its introduction might thus potentially have two side effects that need to be assessed experimentally. On the one hand, speakers of standard Italian may not perceive this sound as being significantly different from other Italian vowels and, consequently, assimilate it to one or more native vowels, as often happens in second language acquisition (PAM-L2 model, Best and Tyler, 2007). On the other hand, even when the schwa phoneme is correctly identified, the neutralized words might be associated with grammatical forms already present in regional dialects, such as Napoletano (spoken in Naples and the surrounding area in the Campania region) and Barese (spoken in Bari and the surrounding area in the Apulia region), which feature vowel reduction to [ə] (Repetti, 2000). As a consequence, the speakers of dialects that include [ə] in their vowel systems may recognize it but fail to associate it with gender neutrality, showing a male bias instead. Due to a common (passive) knowledge of the dialectal repertoire across Italy, this association could also occur among speakers of other dialects, leading to unpredictable effects on people’s interpretation. Given that one of the advantages of the schwa is the fact that it is also “pronounceable,” and that its use has recently started to extend to oral communication, it is important to investigate if the schwa sound is actually recognized as a distinct phoneme by speakers of standard Italian and if it is perceived as gender-neutral.

This article presents the findings of three studies investigating gender associations of Italian role nouns presented in different forms: gendered, bigender, or neutralized. The results of the three studies are then compared to draw conclusions about the effectiveness of the schwa in neutralizing gender on nouns.

In three independent studies, we investigated the gender of the referent associated with role nouns and the efficacy of the symbol ə in neutralizing the noun’s gender association. Given the novelty of this GIL proposal, our research aims at setting a first baseline to assess the efficacy of this intervention in basic, low-level terms: is a noun ending in schwa perceived as ‘neutral’ with respect to gender? Albeit limited in its scope, we believe that such preliminary assessment is necessary to progress in the debate about the feasibility or opportunity of such intervention.

By adopting a modified version of the task used by Misersky et al. (2014), participants were shown role nouns and asked to judge how much each role noun was more likely to refer to a man or a woman using a visual analog scale. No numerical intervals appeared on the scale; only the endpoints were labeled as Man on the left and Woman on the right ends of the scale. In all studies, the stimuli were in Italian, and the participants were native Italian speakers. The studies were administered online via a laptop or personal computer. Studies 1 and 2 were conducted using the Qualtrics web system, while Study 3 was administered through the Labvanced testing platform.

The methodological procedure was the same across studies: after reading the study information and providing consent to participate, participants filled in some demographic information (such as their self-reported gender and age) and received task instructions.

They were then presented with a list of role nouns (one at a time) in their singular form and with no preceding article, either in written form (Studies 1 and 2) or audio format (Study 3). Each stimulus was evaluated by participants on the visual analog scale. At the beginning of each trial, the slider handle appeared in the middle of the scale, and participants had to click on the slider to provide their rating and proceed to the next trial. The question “How likely is it that this noun refers to a man vs. a woman?” appeared immediately above each role noun, as shown in Figure 1.

Before conducting the studies, the research protocol was evaluated and approved by the local commission of the Psychology Department for minimal risk studies.

Study 1 served as the baseline: it included both gender-marked nouns either in their masculine or feminine grammatical endings (maestro_[MASC]/maestra_[FEM], ‘school teacher’), which convey explicit information about the gender of the referent, and common gender nouns, namely, bigender nouns that do not change their ending depending on their referent’s gender and can both indicate a female and a male referent (cantante_{[MASC, FEM]}, ‘singer’). This setup was designed to explore gender associations of role nouns by disentangling (i) the role of grammatical gender, specifically conveyed by gender suffixes on gender-marked role nouns, and (ii) the gender stereotypical associations that might affect the perceived gender category (man or woman) associated with bigender role nouns, in which the form of the noun per se does not convey any information about the referent’s gender. We expect that participants will rely on the grammatical gender cues available on gender-marked role nouns, thus using the left end of the scale (man) for nouns in the masculine form, and the right end of the scale (woman) for nouns in the feminine form, respectively. In the absence of linguistic gender cues, as in bigender nouns, we expect participants to rely on possible stereotypical associations, similarly to what has been observed in languages like English.

Study 2 we used the same bigender role nouns as In Study 1. However, for the gender-marked nouns, the symbol ə replaced the final vowel (i.e., it was used as a suffix), hypothetically neutralizing the gender information conveyed by the form of such nouns. Besides removing gender information conveyed by grammatical form, ə could also act as an explicit neutrality marker thus also reducing stereotypical associations. The rationale behind this manipulation was to understand the gender associations between a role noun and its referent by comparing the canonical forms of the noun used in Study 1 (i.e., masculine and feminine nouns ending in -o and -a, respectively) to the neutralized forms in Study 2 (i.e., the form ending in -ə). To be effective as a form of GIL, the nouns ending in -ə should be judged as neutral on the gender scale, with ratings around the middle point, compared to grammatically gender-marked nouns, which, by contrast, are expected to be polarized toward one of the opposite ends of the scale.

Study 3’s stimuli consisted of a subset of bigender role nouns from Studies 1 and 2, as well as gender-marked nouns with neutralized endings (i.e., ə as in Study 2), but presented in an auditory format. The purpose of Study 3 was to examine whether participants could perceive the schwa sound as genuinely gender-neutral. If this is the case, then people should associate the nouns ending in schwa with gender neutrality, similarly to what is expected for its written counterpart. As said above, since the schwa phoneme does not exist in the Italian phonemic inventory except in some southern Italian dialects, it’s important to assess its efficacy in being perceived as gender-neutral, irrespective of inter-speaker variability.

In this section, we will compare the results of the three studies by running linear mixed-effect models on a dataset that included all participants. The questions that can be answered by these analyses are: (a) whether the ratings on bigender nouns are consistent across studies; (b) whether the ratings on masculine and feminine nouns are significantly neutralized by the adoption of schwa, by comparing gender-marked vs. gender-neutralized forms of the same nouns across studies; and (c) if there is a difference in the effectiveness of neutralization between the use of schwa in the written form and its use in the oral form, by comparing ratings on the gender-neutralized forms in Study 2 and 3.

We first operated a selection of the role nouns that were shared across the studies, resulting in a subset of 60 nouns, 30 bigender and 30 that appeared either in their gendered form (Study 1) or in their neutralized form (Studies 2 and 3). We then created a two-level variable category by adding a column to the dataset, in which masculine and feminine nouns from Study 1 and neutralized nouns from Study 2 and 3 were labeled as non-bigender, while bigender nouns from the three studies were labeled as bigender.

We then computed the centered score adopted in the analyses of the three studies (cf. section 2.1). Considering that in the following analysis, the variable relative to noun gender is substituted by the variable category (bigender/non-bigender), the absolute score eliminating the polarity of masculine and feminine nouns is no longer needed.

The mean ratings obtained by bigender nouns in the three studies are, respectively: −0.039 in Study 1, −0.035 in Study 2, and −0.037 in Study 3. Average ratings on non-bigender (neutralized) nouns are, respectively: 0.030 in Study 1, 0.004 in Study 2, and −0.049 in Study 3.

To address the questions outlined above, we developed a (nested) linear mixed-effect model with the centered score as the dependent variable, the category of the noun (i.e., the two-level variable described above), and the study (three-level variable: Study 1, 2, 3), together with their mutual interaction as independent variables, and participants and items as random intercepts.⁸ For comparison purposes, Study 2 was set as the reference level of the variable study in the analyses. Nested comparisons showed that (a) judgments on bigender nouns were consistent across studies (Study 1 vs. Study 2: t = −0.78, p = 0.473; Study 3 vs. Study 2: t = −0.52, p = 0.600); (b) ratings on the gender-marked nouns in Study 1 were significantly different from those obtained on the same nouns in their written form using ə in Study 2 (t = 3.69, p < 0.001); and (c) there is a difference in the effectiveness of neutralization between the use of ə in the written and oral modality, as shown by the significant difference between neutralized nouns in Study 3 compared to Study 2 (t = −6.30, p < 0.001). The full model output is available in the OSF repository. Figure 6 shows the effect plot of the interaction model.

The present research aimed to investigate the efficacy of a gender-inclusive language proposal in Italian, specifically the adoption of the symbol ə as a gender-neutral word-ending for role nouns. By means of three studies, we aimed to contribute with a first baseline assessment of the efficacy of this intervention in making a noun ‘neutral’ with respect to gender. This preliminary assessment is a first, necessary, step prior to evaluating any higher-order effects of GIL interventions, such as broader cognitive and socio-psychological effects and their impacts on gender societal disparities.

Study 1 aimed primarily to establish a baseline regarding gender associations related to two types of noun categories in Italian: gender-marked nouns (masculine and feminine) and bigender nouns, which refer to people irrespectively of their gender. Study 2 included the same bigender nouns and the nouns that were gender-marked in Study 1, now neutralized through the adoption of the schwa symbol as a gender-unmarked word ending. Finally, given that this proposal is also tailored for spoken language, we aimed to test its effectiveness in neutralizing gender information also in the auditory modality. As explained in the introduction, the use of the schwa in spoken language could present several issues, as this sound is not part of the phonemic inventory of standard Italian but is a variant of some southern dialects. The goal of Study 3 was to test whether neutralized role nouns were actually perceived as neutral in the auditory modality by native Italian speakers.

In Study 1, the gender associations found for bigender nouns showed that expectations regarding the referent’s gender may emerge even in the absence of grammatical marking. The strength of these expectations for grammatically gender-marked nouns was more pronounced for grammatically feminine nouns than for masculine ones. We interpreted this finding as the effect of the generic interpretation of the masculine form, which can be used in Italian as the unmarked gender form (Cacciari and Padovani, 2007). Notably, this effect is stronger in older participants.

Study 2 and Study 3 showed that the adoption of the schwa as a gender-neutral suffix reduced the expectations regarding role nouns, without fully eliminating them. For cases where the form of the neutralized noun evokes the masculine, or when there is a strong stereotypical association, as with stereotypically feminine nouns, the neutralized form seems largely ineffective. Nouns with strong female stereotypical associations (such as “housekeeper” and “secretary”) were not perceived as genuinely neutral, and neutralized variants of the grammatically masculine suffixes (i.e., those ending in -tore, neutralized as -torə) elicited strong associations with the masculine gender. This latter finding might be attributable to the morpho-phonological form of words ending in -tore, which are strongly masculine-marked, being their feminine form -trice.

Overall, our findings answer these broader research questions: (1) Is schwa in the written and auditory form (equally) effective in neutralizing expectations about the referent’s gender of role nouns? (2) Is schwa effective in reducing gender stereotypical associations on role nouns?

The answer to the first question is “more yes than no”: the analysis reported in Section 6 demonstrates the effectiveness of the ə in neutralizing the gender of nouns, suggesting a probable attenuation of gender associations with specific nouns. Nonetheless, this conclusion needs to be taken with caution. First, bigender nouns were overall perceived as more neutral than nouns neutralized with schwa, both in their oral and written forms. Second, the efficacy of schwa was significantly reduced in older participants: as participants’ age increased, their perception of schwa became less neutral. Notably, this effect emerged from a relatively low age threshold, starting around 30 years of age. Third, schwa in the auditory modality was less effective than in the written modality: in the case of oral schwa, the average ratings shifted significantly toward negative values, indicating a stronger association with the masculine gender. This finding might be motivated by the fact that the Italian spoken language is characterized by significant regional variability and that the introduction of a new phoneme—much more than a new grapheme—could conflict with certain regional variants. Specifically, in some southern dialects, such as Campanian and Apulian, word-final vowels are often reduced to /ə/ regardless of their gender marker (Repetti, 2000), and a word like collaboratore_[MASC] would sound exactly like the neutralized version collaboratorə that we have used in this study, whereas its feminine counterpart would sound remarkably different (collaboratricə). It is, therefore, unsurprising that words that exhibit a marked difference between their masculine and feminine forms may be perceived as grammatically masculine. It is worth mentioning, though, that the majority of the participants in Study 3 were from the north of Italy. Nevertheless, these dialectal variants are well-represented in movies, television programs, and popular Italian shows. Therefore, this is common knowledge among native Italian speakers, even if they do not use such dialects.

Turning to the second question, namely, whether schwa is effective in reducing gender stereotypical associations, the answer is “more no than yes”: the comparison between neutralized forms and bigender nouns demonstrated that the latter are, overall, perceived as being more neutral than the former. Despite being gender-neutral in their form, the evaluation of neutralized nouns in terms of their perceived gender is closely linked to stereotypical expectations, similar to what happens in the case of role nouns in natural gender languages like English (Canal et al., 2015). Indeed, the inclusive forms of ‘housekeeper’ and ‘secretary’ showed female stereotypical associations despite their neutralized form. The effect emerges in spite of the small number of stimuli strongly linked to predominantly female-dominated professions. Therefore, this effect may emerge even more strongly if a larger number of such nouns were included, as done in other studies (e.g., Cacciari and Padovani, 2007; Pesciarelli et al., 2019).

As an overall conclusion, our results suggest that the schwa neutralizes the gender information conveyed by the gender-marked suffix of the noun in the younger population. However, the schwa does not seem to act as an explicit marker of neutrality. Speculatively, for words neutralized by schwa, participants might have simply relied on the noun’s stem to make their judgments.

As with all studies, this research has limitations but also highlights interesting issues worth exploring in future research. One limitation is that the measure used in this work is explicit. It would be very interesting to test the effects of this form of inclusive language in an implicit manner (for example, through eye-movement recording or event-related potentials) in order to obtain a more reliable measure of its linguistic processing. Another limitation is that we presented stimuli in their singular form and in isolation. A clearer picture would emerge by testing the neutralized role nouns embedded in texts, as sentential/discourse context might have an effect that we could not observe in the current work and would add ecological validity to the investigation.

Our findings are also of possible interest in light of the still-alive debate surrounding generic masculine. Indeed, some studies (e.g., Gygax et al., 2008; Redl et al., 2022) have shown that when a noun is presented in its masculine form, a representation of a male referent is automatically activated. The results of our work are far from attempting to contribute to this issue, although the finding that role nouns in their masculine form are perceived as more neutral, or certainly less marked, compared to the same role nouns in their feminine form, contributes interesting data to the current debate.

Finally, it is of paramount importance to consider the tradeoff between gains and disadvantages of burdening the linguistic system by introducing a change in the morphosyntactic and morphophonological linguistic domains. Particularly for those GIL proposals that incorporate symbols, an important issue to consider in future research is the impact of these novel forms on text reading, especially in terms of potential disruptions for individuals with reading difficulties. With regard to this issue, we believe there is a risk of including one category of people while excluding another.