— Language, Gender and Hate Speech A Multidisciplinary Approach

In the 150th anniversary of the foundation of Ca’ Foscari University of Venice, the Department of Linguistics and Cultural Comparative Studies, with the contribution of the University Equal Opportunities Committee, organised the LIGHTS conference ( Language gender HaTe Speec h, 18-19 October 2018) with the spon-sorship of the Italian Linguistics Society (SLI) and its Study Group on Language Policies (GSPL), of national associations, such as Toponomastica Femminile, Rete per la Parità, Donne in Rete, GiULiA, Osservatorio di Pavia, and of the Council of Europe (office of Venice). National and international experts and experts in linguistic, sociological, juridical, and computational disciplines discussed the subject of inclusive communication and verbal violence with focus on the category of gender. Given the relative novelty of the subject in the Italian academic context, the conference aimed to establish a first contact among researchers to create a multidisciplinary network including those who study communication and those who work in communication to put forward future projects with a strong socio-cultural impact. The goal of the selected proceedings is to offer a tool of disseminating knowledge of recent research and affirmative actions with an impact on the Italian society, education, and political institutions in an international perspective. Intended readership includes scholars, university students, school teachers, jour-nalists, policy makers, judges and lawyers and mass media publishers


Language, Gender and Hate Speech
It is a common belief that less attention is paid to gender-inclusive language in Italy than in other countries. 1Italian public figures often voice their skepticism towards language reforms aimed at gender equality.They claim that such linguistic innovations are hardly acceptable but fail to provide evidence beyond personal opinions and ideological stances.This paper illustrates a survey aimed to assess the actual acceptability of feminine job titles among a sample of Italian native speakers.Although the sample is not representative of the Italian population as a whole, the survey shows that feminine job titles are perfectly acceptable for the majority of informants who -rather -prove critical of the inconsistencies deriving from the simultaneous use of masculine and feminine morphological traits referring to a woman.This paper proceeds as follows: section 2 describes the difficulties emerging in any language reform owing to the speakers' conservative attitude.With specific reference to Italy, we provide examples of the opposition to feminine job titles in public debates that are in contrast with data showing the increasing frequency of such lexical innovations in the press over the last twenty years.Section 3 illustrates the survey procedure, the traits under scrutiny and the sample of informants.In addition, we provide a detailed account of one of the phases of the survey, followed by an overview of the overall results.Finally, section 4 draws the conclusion of this study against the background provided by the close link between societal changes and linguistic innovations.

2
Promoting Gender-Inclusive Language is Difficult Any paper on gender-inclusive Italian is invariably bound to start with a reference to the recommendations on the avoidance of gender bias drafted by Alma Sabatini as far back as 1987 (Sabatini 1987).Since her publication is well known in Italy, we shall not dwell upon its proposals and shall only briefly list its three main guidelines: avoid masculine forms meant as (supposedly) neutral reference (e.g.uomo, 'man' to refer to mankind; figli, 'sons' to refer to both sons and daughters); avoid definite articles before a woman's surname (cf.Rossi to refer to a man but la Rossi to refer to a woman); avoid masculine nouns describing jobs and titles when referring to women (l'avvoca-to Maria Rossi).Sabatini's recommendations have been implemented only partially and occasionally in the actual linguistic behaviour of the mass media and the public administration, despite a number of promotional initiatives by educators and academics.Examples include manuals for drafting administrative texts (Cortelazzo, Pellegrino 2003), recommendations by the Italian language services of European institutions (Robustelli 2011), recommendations addressing journalists and other media communicators (Manuelli 2014) and analyses of gender bias in school textbooks (Biemmi 2010).Any resistance to using gender-neutral structures may be explained in part with the difficulties all speakers naturally encounter in welcoming linguistic innovations, as shown by the strong opposition to any orthographic reform such as the recent Rechtschreibreform in German, enforced in 1996 after lengthy debates.After all, this kind of resistance is not a peculiarity of Italian speakers, nor is it limited to gender-neutral language.Another example is provided by the adjective unionale, which was proposed as a substitute for comunitario to refer to concepts, policies and documents connected to the European Union.Cortelazzo (2010) was well aware that it was bound to meet with fierce opposition because it would be considered "unusual" and "ugly".This "ugliness" is the result of mere innovation: in fact, unionale fully complies with the morphological rules of Italian, as shown by adjectives derived from nouns such as nazione>nazionale, regione>regionale etc.
One might say that, regardless of individual political and ideological stances, from a linguistic viewpoint we are all conservatives, as clearly shown by the comparison between law and language repeatedly mentioned by Devoto (1958), Nencioni (1962) and Timpanaro (1963).Despite the numerous features in common, the reform of a language by a central authority is much more problematic than the reform of a legal system.In the case of Italian, this natural predisposition of the speakers' community to a linguistically conservative attitude may be stronger than in other languages for historical reasons.Since the unification of the country, Italian language policies -from Manzoni's monolingual approach to the Fascist strive for language purity -have often been characterized by prescriptivism, which in turn has led to the mistrust for any language reform (Lepschy 1989).This is a possible explanation for the weaker role played by the Accademia della Crusca, compared to its counterparts abroad, from the Académie Française to the Real Academia Española, as shown by the timidity of its recent initiative, called Incipit, aimed to curb the excessive use of foreign words in Italian.2

Examples of Opposition to Feminine Job Titles
Admittedly, the importance attributed to gender neutrality in language is smaller in Italy than in other countries.The media and even the Academia have often opposed the introduction of gender-inclusive strategies with a wide range of arguments.Since the survey illustrated in section 3 of this paper focuses on feminine job titles, a few examples regarding this word class are illustrated below.
A source rich in anecdotes is the blog by the linguist Michele Cortelazzo titled Parole.Opinioni, riflessioni, dati sulla lingua. 3In particular, Cortelazzo focuses on the lack of importance attached to gender-inclusive language by members of Italian political institutions.For example, he stresses the contradictory behaviour held by former minister Elsa Fornero when she declared to the press that she did not like seeing her surname preceded by the feminine article (la Fornero), but then described her position with the masculine form (ministro).According to Cortelazzo, it appears she believed that a position of great prestige was more suitably described by a masculine noun, while, had she been a nurse, a teacher or a shop assistant, she would have probably thought otherwise and used the feminine infermiera (nurse.F), maestra (teacher.F) or commessa (shop assistant.F).
Cortelazzo also mentions the numerous abusive comments addressed to Laura Boldrini, the former President of the Chamber of Deputies.Boldrini sent a letter to her colleagues to foster the use of gender-inclusive formulae and received sharp -even violent -criticism on the social media.Many comments were based on strictly personal preferences, inevitably lacking any scientific foundation.Surprisingly, criticism was also voiced by Giorgio Napolitano, President emeritus of the Italian Republic, who reportedly declared to the press that the masculine nouns ministro and sindaco are perfectly dignified, whereas their feminine counterparts ministra and sindaca (mayor.F) were, respectively, "horrible" and "abominable". 4rovided the limits to the (polite) expression of personal opinions are not exceeded, the anecdotes reported above by Cortelazzo are hardly surprising: the conservative attitude he describes is voiced by Italian speakers who are not experts in linguistics.However, criticism was levelled at gender-neutral language by members of the Academia as well, from Nencioni (1987), who considered Sabatini's recommendations as examples of authoritarian intimidation against the rules of Italian grammar, to Sgroi (2018), who claimed that, if speakers wish to, the Italian language is intrinsically equipped with all the necessary resources to make the referent's sex explicit; thus, there is no The Acceptability of Feminine Job Titles in Italian Newspaper Articles Quaderni del Comitato Unico di Garanzia dell'Università Ca' Foscari Venezia 1 79 Language, Gender and Hate Speech, 75-90 need to introduce novelties that are both scientifically ungrounded and ideologically biased.
Conclusively, one may say that the opponents to the innovations proposed by Sabatini in 1987 base their criticism on: 1. Mere personal opinions referring to the "ugliness" of the proposals put forward for feminine job titles.2. Untenable arguments condemning the purported violation of grammar rules (why is sinistra, 'left' perfectly acceptable but ministra to be rejected?And what about attrice, 'actress' vs. rettrice, 'vice-Chancellor.F' or protetta, 'protected.F' vs. architetta, 'architect.F'?). 3. Undeniable difficulties in implementing gender-inclusive strategies in languages characterized by rich morphology (managing split forms in Italian is certainly more cumbersome than in English).4. The all-out condemnation of changes imposed from above for ideological reasons.
Alma Sabatini was perfectly aware that encouraging linguistic change is a difficult task and noted that language users tend to neglect the great difference emerging between imposing a word from above and simply proposing it (Sabatini 1987, 99).Apparently, critics of gender-inclusive language tend to consider the proposals put forward as instances of linguistic imposition rather than recommendations that are political in nature, as rightly pointed out by Elena Pistolesi (cited in Cortelazzo 2015).

Do Italians Dislike Feminine Job Titles?
Based on the examples above, we can claim that the Italian-speaking community shows a certain dislike for feminine forms denoting professions and titles traditionally attributed to men: even women seem to prefer masculine nouns to refer to their positions.However, the data we rely on consist in explicit statements that do not reflect the actual linguistic choices of Italian speakers.For example, Italians are likely to express their regret for one of the most successful myths about their language, i.e. the decline of the subjunctive mood (Schneider 1999).Consequently, they might affirm that they make abundant use of this verbal mood, even though in reality they do not (or, at least, not always).At present, little evidence is available to evaluate this resistance to the introduction of feminine job titles in Italian.Owing to this gap, tentative results may be drawn from the research conducted by Ondelli and Viale (2010) on newspaper articles translated from other languages and published in Italian newspapers.The influence of source languages (mostly English) that are prone to welcoming politically correct structures may pave the way to a greater number of gender-neutral solutions in translations (e.g.split forms), which are likely to go unnoticed by the readers.For example, the fortune of the feminine noun cancelliera (chancellor.F) is very likely to be connected to the use of Kanzlerin in German, although, after closer scrutiny, this feminine job title may sound much 'stranger' than ministra or sindaca.
A simple search in the online archive of la Repubblica 5 provides evidence of the increasing acceptability of feminine job titles in Italy in recent years.Table 1 below reports the frequency of the words assessora (alderwoman), avvocata (lawyer.F) and ministra during the first twenty years of the 21st century (the search was conducted in September 2018): Table 1 shows that the frequency of these three words has significantly increased in the last ten years, and assessora has literally soared in the last four years.Of course, this test has no scientific value because, on the one hand, it includes only a limited number of word types and, on the other, it accounts for the choice of only one of the main Italian newspapers circulating at national level.In addition, even though quality newspaper articles are currently viewed as the reference model of the new Standard Italian (Antonelli 2011), quality press readers may not be considered representative of the general population from a diastratic viewpoint.Nevertheless, the tendency to a greater acceptability of feminine job titles is very likely to emerge in other mass media too (e.g.television and the Internet).

The Survey
As is often the case, regardless of individual preferences based on ideology or more or less scientific arguments, the main challenge for linguists is posed by the collection of data that may be used to effec-5 An Italian national newspaper available at https://www.repubblica.it.

Giorgia Castenetto, Stefano Ondelli
The Acceptability of Feminine Job Titles in Italian Newspaper Articles tively assess the actual behaviour of a community of speakers.Data collection needs to be managed in order to portray the reality under examination as objectively as possible.
For example, in the case of the small experiment illustrated in section 2.2 above, besides the very limited number of job titles considered, a possible objection may entail the role played by ideological factors: e.g., since la Repubblica is known to adopt a progressive political stance, its journalists may accept feminine job titles more easily than their colleagues working for other newspapers.Similarly, a survey of different written text types may generate other imbalances in the analysis: e.g., in the case of administrative texts, a role may be played by mandatory guidelines imposed by state authorities requiring the use of gender-inclusive language, thus leading to an increase in the number of feminine job titles not reflective of their actual acceptability on the part of the common reader.

Methodological Issues
To conduct a survey aimed at assessing the acceptability of feminine job titles by native speakers of Italian, we adopted and adjusted the model implemented by Ondelli and Romanini (2018) to assess the acceptability of features of the new Standard Italian.The difficulties deriving from the elicitation of data from informants may be summarized in the three main points illustrated below.
a. Unsurprisingly, the need emerges to select a varied population sample representing the average speakers from the sociolinguistic viewpoint, with particular reference to diastractic variation.In the case at hand, diamesic factors do not seem to play a significant role, while diatopic variation, rather than referring to the traditional differences emerging in the northern, Tuscan and southern varieties of Italian, may account for the differences between linguistically innovative large metropolitan areas vs. more conservative rural areas.However, it is the diastratic variation, accounting for factors such as age, sex and education, which is more likely to impact the results of our survey.b.One should avoid the overt exhibition of the linguistic traits under scrutiny, which in turn may explicitly draw the informants' attention on those traits.E.g., if the respondents are presented with a simple sentence or clause in isolation and are asked: "do you like this feminine job title?This is exactly what we are trying to analyse", they may feel they are called upon to judge the correctness of that word and draw upon the explicit norms they were taught at school or base their answers on their ideological and political values.In fact, we are trying to measure their actual linguistic behaviour rather than what they think is linguistically correct or ideologically preferable.c.In the light of (b) above, the text type including the feminine job titles we are analysing is paramount, since it is liable to influence acceptability.E.g., typical features of administrative texts such as the narrative imperfect or the use of the present participle as a verb rather than as a noun or adjective may go unnoticed, while they would stand out in other text types.Similarly, if readers expect to find gender inclusive language in administrative texts, they may consider the presence of feminine job titles less unusual and may not notice them.

Trait Selection, Submission Procedure, Distribution Procedure
The first step of the survey consisted in selecting a range of structures involving job titles to be submitted to the population sample called upon to assess their acceptability.The list of these structures is provided below.
• Feminine job titles whose acceptability on the part of the majority of Italian speakers is questionable as opposed to more traditional forms: sindaca, ministra, la vigile, vigile donna, vigilessa (traffic warden.F), avvocata, avvocatessa, donna avvocato, avvocato donna and la legale (lawyer.F).• Masculine job titles attributed to female referents in the text.
• Both masculine and feminine job titles attributed to female referents in the text (e.g. using both ministro and ministra for the same person).• Grammatical agreement: adjectives and past participles agreeing with masculine job titles attributed to a female referent (e.g.il sindaco Chiara Appendino è soddisfatto, 'the.M mayor.M C.A. is satisfied.M').• Referential agreement: adjectives and past participles agreeing with a female referent preceded by masculine job titles (e.g.il sindaco Chiara Appendino è soddisfatta, 'the.M mayor.M C.A. is satisfied.F').• Alternating grammatical and referential agreements in connection to the same female referent.
To avoid influencing the respondents and elicit spontaneous answers, the informants were asked to read short passages extracted from newspaper articles and write comments on the overall linguistic quality of the texts.In other words, they were asked to note any linguistic trait they considered unacceptable or even incorrect, but no explicit mention was made of feminine job titles.We chose newspaper articles because the language of the press is considered the model for the new Standard Italian (Antonelli 2011).As already observed, although the quality press readers may not be considered representative of the average Italian speaker in the light of the low reading levels of the population, most Italians are familiar with this text type.Furthermore, given the broad range of subjects covered (politics, business, local news etc.), newspaper articles often make reference to job titles of great prestige attributed to women.Four passages were selected from articles published online by the news agency Adnkronos and the newspapers Il Messaggero, la Repubblica and Il Sole 24 Ore.To ensure that all the structures we wished to assess were present, we modified the extracts when necessary.
Finally, the texts were submitted through Surveymonkey,6 an online free service for the creation and administration of questionnaires through the Internet.Students, friends and relatives were asked to disseminate the questionnaire among their contacts via email, the social media etc., in order to include a large (207 subjects) and varied pool of respondents.Unfortunately, upon collecting and classifying the results, the sample proved to be excessively homogeneous, since the vast majority included women, young people (between 20 and 25), people with a high level of education (having completed at least their secondary education) and living in northern Italy.Since the subjects included in those subgroups exceeded one half of the total number of respondents, the average values of their answers are more reliable than those calculated for less numerous groups, in which one single answer may lead to significant variations in the mean values.
Vigilesse was criticised by a very small number of informants, while le vigili was considered unacceptable by 25% and 27% of the men and women in our population sample respectively.In terms of age, they account for 11% of the 14-19 age group, 26% of the 20-25 age group, 26% of the 26-34 age group, 31% of the 35-54 age group and 31% of the 55-64 age group.In terms of education, those who rejected le vigili account for 24% of the respondents who have completed their lower secondary education (eight years in total), 24% of those who have completed their higher secondary education, 29% of those who have a bachelor's degree and 33% of the respondents with a master's degree.Lastly, in terms of origin, 28% reside in northern Italy, 27% in the south and slightly more than 10% in central Italy.
Since the informants were free to provide explanations for their judgments, some wrote that le vigili is unacceptable because vigile is a masculine noun and cannot be preceded by a feminine article.In fact, vigile is an epicene (or common) noun, i.e. it has one single form that can be used for both men and women (as is the case with cantante).Its gender may only be derived from determiners or adjectives, as in il/la cantante or un'interprete molto colta.In other words, according to the Italian grammar, vigile may be correctly introduced by either the masculine article il or the feminine article la.
Finally, the third variant (i vigili donna) was rejected by 39% of female informants and 30% of male informants.Those against i vigili donna made up 62% of the 14-19 age group, 40% of the 20-25 age group, 26% of the 26-34 age group, 22% of the 35-54 age group and 25% of the 55-64 age group.Those who voiced their criticism accounted for 30% of the informants who have completed their lower secondary education (eight years), 36% of those who have completed their higher secondary education, 29% of informants with a bachelor's degree, 33% of those with a master's degree.Finally, they account for 28% of residents in northern Italy, 27% of the informants living in the south and slightly more than 10% of those residing in central Italy.
With reference to the acceptability of the alternating use of different solutions for the feminization of vigili, the lack of consistency was mentioned by only 33 subjects out of 193 (17%).A possible explanation is that the vast majority of the informants considered these different feminine forms as perfectly interchangeable synonyms.
As regards the acceptability of the referential agreement, included in the text within the sentence "i due vigili donna picchiate", only 8% of the informants deemed it incorrect, i.e. 16 out of 193.
Another job title included in the passage was the masculine noun sindaco referring to a woman.Only 9 respondents out of 193 considered it incorrect, which is hardly surprising since the masculine forms seem to be preferred to refer to the offices and professions considered of great prestige.

Overall Results
Our survey cannot be considered a reliable picture of the linguistic norm (i.e. the actual average linguistic behaviour of speakers) in Italy.A larger and more balanced population sample would be necessary, while most of our respondents are young, female, highly educated and resident in northern Italy.However, with reference to the four more numerous groups, the results of our survey on the acceptability of sindaca, ministra, la vigile, il vigile donna, vigilessa, avvocata, avvocatessa, donna avvocato, avvocato donna and la legale may be illustrated as follows: • sindaca is accepted by approximately 4/5 of respondents; • le vigili is accepted by approximately 3/4 of respondents; • ministra is accepted by approximately 5/6 of respondents; • avvocata is accepted by approximately 2/3 of respondents; • avvocatessa is accepted by approximately 90% of respondents; • vigilessa, donna avvocato, avvocato donna and le legali are accepted by nearly all respondents; • vigili donna is more frequently subject to criticism; however, it is accepted by approximately 60% of the informants.
We summarize the general conclusions of our survey below.1. Men and women share the same opinion on feminine job titles, since neither group of informants proved more critical than the other.2. Surprisingly, it was the 26-34 age group and the group of informants with higher education levels who proved more critical of feminine job titles, immediately followed by the informants who had only completed their lower secondary education (licenza media).However, no clear pattern emerges in terms of age and level of education.In other words, younger and more educated informants are not necessarily more open to innovations nor are the older and less educated cohorts less inclined to welcome new feminine job titles.3.In general, feminine job titles today seem to be more easily accepted than in the past, although their masculine variants referring to women still provide a widely accepted alternative, since they were virtually never subject to criticism.4. Today, the linguistic strategies implemented to obtain feminine job titles seem to be less criticized than the introduction of new feminine forms in addition to a pre-existing, more familiar feminine noun, as shown by the assessment of le vigili and le legali or vigili donna and avvocati donna.Le vigili and i vigili donna met with greater opposition than le legali or gli avvocati donna, although the morphological structures are the same.The pre-existing and widely used form vigilessa might explain this difference, since la legale is an innovation.Avvocatessa was considered acceptable by our informants, although it is less familiar than vigilessa and leaves room for the introduction of avvocato donna or even avvocata.5. Neither grammatical (e.g., l'avvocato Raggi è appena arrivato) nor referential (e.g., l'avvocato Raggi è appena arrivata) agreements were specifically opposed; rather, criticism was levelled at their simultaneous presence within the same sentence or text (approximately 1/3 of respondents), since it impinges upon text cohesion and clarity.

Conclusion
Although opinions varied, our survey shows a decreasing trend in the assumed dislike for the feminine forms of nouns describing titles and professions in the light of their purported violation of the rules of Italian morphology and syntax (Robustelli 2013).In contrast, the use of feminine forms for titles and professions provides a solution to the morphological and syntactic inconsistencies considered by the informants as the real problem in terms of a correct use of the language.
In the final analysis, those who describe Alma Sabatini's proposals as prescriptive and unjustified choices produced by ideological bias, devoid of any scientific grounding and contrary to the rules of Italian grammar, seem to disregard the dynamic relationship emerging between usage and rules, i.e. the driving force of linguistic change (Coseriu 1971).The increase in the acceptability of female job titles has probably been strengthened by their frequent use in the mass media.In other words, societal changes (in this case, women's access to positions of power and status) is reflected by linguistic innovations regardless of any grammatical reason for maintaining the masculine forms as the sole option.So, the use of masculine forms as neutral terminology referring to titles and positions in laws and regulations may well be justified but, as Cortelazzo (2017) rightly points out, it is the community of speakers who will eventually decide what is acceptable in their language.

Giorgia Castenetto, Stefano Ondelli
The Acceptability of Feminine Job Titles in Italian Newspaper Articles

Chart 1
Informant distribution according to sex, level of education, age and Feminine Job Titles in Italian Newspaper Articles Quaderni del Comitato Unico di Garanzia dell'Università Ca' Foscari Venezia 1 85 Language, Gender and Hate Speech, 75-90

Table of Contents
Can Gender-Fair Language Combat Gendered Hate Speech?Some

Table 1
Total and annual frequencies of assessora, avvocata and ministra in la Repubblica, 2000-18.The 2015-18 period was calculated as four full years

Giorgia Castenetto, Stefano Ondelli The Acceptability of Feminine Job Titles in Italian Newspaper Articles The Acceptability of Feminine Job Titles in Italian Newspaper Articles Quaderni
del Comitato Unico di Garanzia dell'Università Ca' Foscari Venezia 1 83 Language, Gender and Hate Speech, 75-90

Giorgia Castenetto, Stefano Ondelli The Acceptability of Feminine Job Titles in Italian Newspaper Articles Giorgia Castenetto, Stefano Ondelli The Acceptability of Feminine Job Titles in Italian Newspaper Articles Quaderni
del Comitato Unico di Garanzia dell'Università Ca' Foscari Venezia 1 87 Language, Gender and Hate Speech, 75-90