Participation and deliberation in language policy: the case of gender-neutral language

ABSTRACT

This paper investigates language policy formation through participatory and deliberative methods and, more concretely, the regulation of gender-neutral language in Barcelona City Council (Spain). Through an argumentative approach to policy, the paper examines a specific language policy idea, process and solution, and the accompanying discursive argumentation used by decision-makers. The paper (a) shows that linguistic preconceptions and power relations may constitute a potential barrier to effective deliberation on language, and (b) argues that if local modes of governance (as opposed to the centralised role assigned to language academies) are going to be used to prescribe language practices in institutional contexts, arrangements should be put in place to provide access to a full range of views about language, generate reflective judgments and promote a public exchange of arguments. The paper concludes with a discussion about the implications of this study for the inclusion of deliberation in language policy-making.

KEYWORDS:

• Deliberation
• language policy and planning
• gender-neutral language
• gender-fair language
• language and gender

Introduction

In recent decades, gender-neutral language (GNL) has earned considerable public support and social recognition. This has had an impact on language policy and planning (LPP) at an institutional level, albeit often without the normative endorsement of language-planning bodies such as language academies. Managerial authorities have often agreed with the arguments advanced by advocates of GNL and so it has been accepted by the decision-makers in administration, public bodies and organisations in many countries, with different measures in favour of GNL (e.g. guidelines, style guides and regulations) being encouraged and implemented. At least to a certain extent, feminists and other activist groups and actors have gone from being norm-breakers to becoming norm-makers in language reform and agents of change in policy formation. Yet, GNL is a socially, ideologically and academically contested concept, in Spain and elsewhere, and not everyone open to arguments in its favour wholeheartedly supports it.

The processes of language policy-making have been under-researched even though globalisation, language rights and social inclusion have led to the most fruitful period for LPP in history (Lo Bianco, 2013). Research on policy around ‘feminist language planning’ (Liddicoat, 2011) is particularly sparse, as it has only emerged relatively recently. The present study attempts to bring together GNL and LPP formation with a focus on participatory and deliberative forms of democracy. While the paper is specifically about GNL, its overall focus is on democratic policy processes and their contribution to policy development. These forms of democracy aim to promote egalitarian values by directly involving citizens in public reasoning and collective decision-making (Cohen, 2007). Since earlier versions of democratic deliberation seem unrealistic and idealised when faced with complex political conflicts (see Peled & Bonotti, 2019 for a linguistic context), I draw in particular on approaches with potential applicability to real-life contexts which involve ongoing disputes (e.g. Dryzek & Niemeyer, 2010). Unlike the considerable body of thought on language issues derived from liberal and rights-based models, participatory and deliberative approaches to language have received less attention from researchers (Schmidt, 2014). A number of applied linguists and other scholars have recently explored the theoretical and practical implications of these perspectives in the language domain (Erdocia, 2020; Erdocia et al., 2020; Lo Bianco, 2013, 2017; Ramberg, 2016; Røyneland, 2016).

In line with this scholarly work, I aim to analyse language policy formation through participatory and deliberative means in a real-world case: the adoption of GNL in Barcelona City Council. In 2018 the council organised such an approach to combat discrimination against women and to promote and regulate gender equity in the institution. The regulation includes, among many other measures, the use of GNL for the council’s external and internal communications and by external contractors. This case is noteworthy because, unlike the more frequent voluntary basis of approaches to GNL in other institutions, here its promulgation is of a binding nature.

The purpose of this paper is to analyse the practical argumentation (Fairclough, 2013) behind the original development of the policy, the construction of the process and the adoption of the solution. To do this, I start by providing an overview of GNL, with an emphasis on the Spanish context. I move on to elaborate on my theoretical and methodological principles and then present the analysis and interpretation of my data (policy documents and interviews). I conclude with a discussion of the key findings and the implications of this study for the inclusion of LPP within governance models. In brief, I show that the public dimensions of the policy process have been outweighed in the policy formation process by the ideological positions of the policy decision-makers and I argue that linguistic preconceptions and power relations may constitute a potential barrier to effective deliberation on language. I claim that in cases where particular language practices are prescribed in institutional contexts through local modes of governance (as opposed to the centralised, authoritative role assigned to language academies), then it is important to include measures which support the exchange of a wide-range of opinions and perspectives in relation to language and an environment which is conducive to discussion. This study can fruitfully contribute to illuminating other scenarios in which a strong preference for a given understanding of language may cause disregard for open discussions and legitimate disagreements.

Gender-neutral language

Advocacy for GNL emerged from the linguistic activism and scholarly work of the 1970s and 1980s. As part of a broader movement for women’s liberation, feminist activists challenged male authority in language regulation, claiming that normative language had a sexist and androcentric nature, including in the development of grammars, dictionaries and other planning activities under the authority of male-dominated language academies and other institutions (Pauwels, 2003). Since then, GNL has gained considerable recognition and social support.

Advocates of GNL adopt a relativist stance towards the relation between language and worldview (Alvanoudi, 2020), according to which language is a reflection of reality and shapes it. In other words, language mediates the experience of the world that speakers have and grammar guides them to specific interpretations of experience (see Slobin, 1996). The following statement is a good illustration of this position:

Language does not merely reflect the way we think: it also shapes our thinking. If words and expressions that imply that women are inferior to men are constantly used, that assumption of inferiority tends to become part of our mindset; hence the need to adjust our language when our ideas evolve. (UNESCO, 2011, p. 4)

On this view, there is a direct relationship between the use of gendered language structures and the construction and interiorisation of social gender. It follows that norms that standardise these structures relegate women to a deviant and subordinate status to men (Hellinger, 2001). Put differently, some prescribed practices represent and perpetuate traditional gender hierarchies in society and grant men more power than women (Ridgeway & Correll, 2004). Using words and linguistic forms in a different way renames and reinterprets reality and can, therefore, contribute to effective social change by reducing linguistic stereotyping, the asymmetrical treatment of genders through language and power inequalities.

There are different proposals for alternative mechanisms to promote more inclusive linguistic practices depending on the type of language. In the case of languages with grammatical gender, such as Catalan and Spanish,¹ the masculine forms of agent nouns (words that denote somebody who performs an action) and other agreeing elements are traditionally the unmarked form and thus function generically (e.g. ciudadano ‘citizen-masc’). Feminine forms, on the other hand, refer only to women (ciudadana ‘(female) citizen-fem’). Strategies to avoid the use of generic masculine (see Lledó, 2005 and Grijelmo, 2019 for Catalan and Spanish respectively) mainly include (1) various forms of gender-feminisation – the insertion of masculine-feminine pairs or double gender marking (ciudadanos/as) or, to a lesser extent, use of generic feminine; and (2) gender-neutralisation – the replacement of generic masculine nouns by unmarked gender-indefinite ones (ciudadanía ‘citizenry’) or the use (in Spanish) of a new non-gendered morpheme e (ciudadanes) or graphic character or letter such as @ (ciudadan@s) or x (ciudadanxs) to replace the masculine form (note that the last two are unpronounceable). Most of these mechanisms do not adhere to standard normative language, and which would be the overall preferred alternative strategy is still a matter of discussion amongst feminist and transgender activists. There have been recent calls stressing the use of more gender-inclusive means (e.g. the use of the non-binary ending e) rather than the use of masculine-feminine pairs, to avoid misgendering.

To be sure, many argue that the gender system is a purely linguistic phenomenon (e.g. Junyent, 2013; Roca, 2005) and that claims for GNL rest on a flawed interpretation of the relationship between grammatical gender and sex. Most of the alternative strategies,² they say, are redundant, exclusionary or unnecessary (e.g. Real Academia Española, 2020) and should be a matter of personal choice; the usual claim is that most of the standard norms are neither sexist nor discriminatory. For instance, Escandell-Vidal (2020) states that ‘maintaining the use of the unmarked form for inclusive purposes [in Spanish] is not a way of defending and perpetuating the silencing of women’ (p. 246). Other arguments against the adoption of GNL refer to issues of impracticality or a lack of understanding, or involve appeals to authority, historical authenticity and freedom of speech, amongst others (Blaubergs, 1980; Parks & Roberton, 1998). That said, an increasing body of experimental research on perceptions of gender and sex (see Sczesny et al., 2016, pp. 2–3) points to a male bias when masculine generics are used, especially in relation to social roles and professional occupations (see Carreiras et al., 1996 for Spanish). However, results from experiments are mixed and vary considerably, making it difficult to address this question definitively (Mickan et al., 2014).

Nowadays in Spain, GNL is a matter of an ongoing debate and the situation is still far from social, political or academic agreement. One example of this lack of consensus can be seen in ‘We are women, we are linguists, we are many and we say enough’ (Som dones, som lingüistes, som moltes i diem prou) (Junyent, 2021), an edited volume with contributions from 70 female linguists who hold differing (but mostly critical) stances towards GNL. The publication of the book gave rise to a heated debate in which the editor, Carme Junyent, professor of linguistics at the University of Barcelona, became the target of a torrent of dismissive remarks and insults on social networks. At the national level, one of the latest public episodes of this debate is the response from the Real Academia Española (RAE) to an enquiry from Carmen Calvo, vice-president of the Spanish left-wing coalition government, about the project to make the Spanish Constitution linguistically inclusive. The RAE (2020) stated in its report that the use of masculine-feminine pairs when referring to mixed or generic groups is acceptable; however, they made unambiguously clear that the language of the current Constitution is ‘grammatically impeccable’ and that the use of masculine plural as a generic form (e.g. los españoles ‘Spaniards-masc’ to refer to Spanish men and women) is ‘correct’ and ‘inclusive’. The stance taken in the report was not unexpected, since the RAE’s position has not undergone significant changes in recent times (cf. Bosque, 2012). What was perhaps more significant was the attempt to separate the statement from an ideological position by stressing that it was a linguistically-based decision: Santiago Muñoz Machado (2020), the president of the RAE, expressed in a press briefing that ‘incorporating it [GNL] or not into the Constitution will be a political decision that is not the responsibility of the Academy’. Needless to say, the RAE’s resistance to (and the Institut d’Estudis Catalans’s disregard of) more explicit inclusive alternatives has led to criticisms that its patriarchal structures, its conservative approach and its ideological bias reinforce sexism (Becker, 2019; Moreno, 2012). It is normally in the progressive centre-left and left of the political spectrum where GNL receives more ample support.

Theoretical framework and methodology

In this paper, GNL is tentatively defined as the prescribed forms of communicative practices that attempt to avoid bias towards a particular sex or gender and to promote social change. GNL results from a set of rationalised and embodied beliefs based on the perceived nature of language and responding to a grievance against standard normative language. This working definition draws upon two conceptual underpinnings:

1. Problematisation: GNL is an articulated conception of language and linguistic behaviour that problematises the standard norms that emanate from normative language institutions. This includes language practices that, consciously or not, follow ingrained norms. Hence, the advocates of GNL challenge the authority of language regulators.

2. Transformative change: The goal of GNL is a broader societal change mediated by the use of alternative linguistic conventions. GNL is constituted by moral values and is often tied to social and political ideologies. In this sense, GNL is linked to different degrees of activism, political activity and power dynamics. The transformation of societal tendencies inevitably implies political action through multiple channels, including the struggle for institutional control that enables policy change.

I adopt a critical orientation to policy analysis and, more specifically, I take a problem-driven approach (Fairclough, 2013). My analytical task is to problematise a particular course of action leading to language regulation in institutional settings. I examine the way the policy idea, process and outcome in the Barcelona City Council were constructed and whether and how this outcome and the process leading to it were influenced by a specific or preconceived conception of language and language use (and if so, why). My problematisation includes the participatory process, namely an open process that enables the political engagement of citizens, with a focus on evaluating the inclusion and exclusion of individual and collective actors, viewpoints and deliberative practices within the policy process. To do this, I concentrate on the actors in management and governance, who seek to introduce transformative changes at the institutional level and in existing social life through the institutionalisation of GNL practices.

I conceptualise the politics of policy change as the ways in which organised networks of social and political actors and activists strive to set out their ideological programmes in the institutional sphere. Like political struggles, disputes about language are mediated by the institutional setting in which they take place. Many administrations (e.g. regional governments, city councils, universities, etc.) have become privileged sites in which social demands on language linked to wider identities, cultures and political projects (e.g. GNL) can gain full policy legitimacy. I consider that, when GNL reaches the institutional sphere, it becomes part of the logics of production and contestation of hegemony (Howarth, 2009). This is because, notwithstanding the increasing support and use of GNL at various levels, some of the claims that its advocators make remain hotly contested. Proposals to adopt GNL stem from ideological premises of feminist and progressive movements and are often (but not always) part of political projects that counter those of conservative parties by (re)defining representations of social realities and controlling social practices. Although my approach includes conflictual elements, it does not exclude cooperative practices such as negotiation, compromise and building alliances. I subscribe to Fairclough’s (2013) view that such an approach can include a commitment to deliberative democracy as a normative ideal and desirable objective.

Deliberative democracy is a form of government in which equal citizens participate exchanging reasons to justify decisions in public discussions and debates with the aim of reaching conclusions that are binding on all citizens (Gutmann & Thompson, 2009). Deliberative practices can induce preference change and yield rational collective outcomes. Deliberation is not the same as participation. Participation is the engagement of citizens and stakeholders in institutional processes which are open. For instance, Cook et al. (2007) refer to political participation as the ways in which citizens discuss in institutional settings about issues that concern their community, city or country. Deliberation is about reasoning and collective decision-making, not simply discussing (Cohen, 2007). Although some deliberationists maintain that deliberation alone should lead to unanimity, many others argue that the ideal of consensus can be an unrealistic one when faced with deep political and ideological disagreements. Dryzek and Niemeyer (2010) argue that ‘meta-consensus’, or agreement based on the legitimacy of competing preferences and the validity of contested judgements, is a more realistic goal of deliberation. Along with this view, I consider that democratic politics should not be reduced to a search for consensus.

My approach lies within the ‘argumentative turn’ in public policy (Fischer & Gottweis, 2012), an alternative perspective to analytic approaches that impose scientific frameworks on the process of decision-making. The argumentative orientation is context-sensitive, focuses on real-world problems, and emphasises communicative and deliberative practices. I adopt an interpretive policy analysis for the evaluation of discursive argumentation. Argumentation, which is central in the policy process (Majone, 1989), is a reasoned articulation of explanatory and justifying statements that is shaped by discourse and leads to a conclusion. When applied to policy-making, which is intended to find solutions to real-world problems, argumentative practices look at the validity of reasons and motives for decision-makers’ actions. In other words, arguments are beliefs, opinions and values about a subject matter that advance decision-making. I draw on the concept of practical argument (Fischer, 2003), which seeks ‘to establish that particular acts are good and should have been performed’ (p. 189) to solve a problem. Practical argumentation refers then to the effectivity of discourse when purporting to persuade others about the appropriateness and reasonableness of a policy decision, including its legitimacy. Consequently, the analysis of practical argumentation should include both empirical and normative enquiries.

I use a critical discourse analysis framework to assess decisions and arguments in my data. My analysis draws inspiration from the idea that policy decision-makers pursue a certain pathway over others in the light of three premises: goals, circumstances and values (Fairclough & Fairclough, 2012). The arguments for taking a particular action to solve a problem stem from the assumption that the action might enable the agent to reach their goals following their values and in the particular circumstances in which they find themselves, including institutional constraints. The agent claims that they ought to take that specific action, as it is the right means to modify the existing status quo. In short, argumentative discourses provide agents with reasons for action. I engage in a two-step analysis (see Wodak, 2015): I begin by mapping out the thematic content of my data and then move on to an in-depth analysis that identifies practical arguments (goals, circumstances and values) and evaluates the justification given for that particular deliberative course of action.

My approach involves working with two sources of data: policy documents and interviews with decision-makers.

First, I analyse a set of sixteen documents, which cover the different stages of the processing of the Regulation for Gender Equity (see Table 1). This documentation is provided by the transparency site of the Barcelona City Council (https://ajuntament.barcelona.cat/transparencia/es/reglamento-igualdad-de-genero). My set consists of (1) the policy outcomes; (2) a range of administrative, legal and procedural documents that led to the approval of the regulation; and (3) other documents that were referenced in the first two sets and provide complementary information about the context, making and implementation of the regulation. For the selection of the material, I chose the most relevant documents concerning language.

Table 1. Policy documents included in the analysis (with reference codes).

Stage/source	Document type (with reference codes)	Number of documents
(1) Policy outcomes	(1.1) Regulation for Gender Equity (original and final versions)	2
	(1.2) Guide to Inclusive Communication	1
	(1.3) Guide to Non-Sexist Use of Language	1
(2) Policy process	(2.1) Report	1
	(2.2) Amendments	4
	(2.3) Participatory process	2
(3) Others	(3.1) Equality Plan 2015-2019, (3.2) Recruitment Plan, (3.3) Verification of Compliance with Contract Terms, (3.4) Guidelines for Drafting Municipal Regulations, (3.5) Plan for Equity of Women and Men 2020–2023	5

Second, I conducted three in-depth, semistructured interviews.³ The three interview participants are female and are senior figures from various departments in the Barcelona City Council, who were directly involved in the elaboration and processing of the policy.⁴ Interviews were conducted in Spanish by videoconference and lasted for an average of one hour. They had a set of predetermined questions that opened up the dialogue and allowed respondents to produce argumentative discourses and to take argumentation in different directions. Each interview contained three phases:

The first phase was an introductory one and aimed to situate the practices of policy-making around language in a wider institutional context. Participants were asked to outline their role in the processing of the regulation and to explain the reasons for this initiative and the policy process.

The second phase concentrated on that part of the regulation that concerns language. As GNL is a broad concept to operationalise in an interview, I chose two specific and contentious cases (masculine generics and gender-neutral forms) to better situate the conversation. The purpose of this part was twofold: to assess the decision-making processes and bodies, discursive networks, linguistic standpoints and the scope of deliberative practices on language; and to analyse the argumentation for the particular policy course of action that was taken.

The third phase focused on one specific aspect of the regulation of GNL, its usage by the council’s employees and by private companies that seek to establish a contractual relationship with the council. The reason for this is that this regulation is distinct from other council-level institutional approaches to GNL in Spain (e.g. in Madrid), which are voluntary. This last phase seeks to evaluate the arguments for this particular policy solution in contrast to those of institutional actors who have pursued other strategies towards the implementation of GNL.

The processes and outcomes of qualitative research are not without any influence from the researcher’s subjectivity. Researchers often embody in some fashion the values of the object under investigation. Indeed this is partly the case for the present author, who sympathises with arguments for GNL and has integrated it into his communicative practices in English and Spanish (albeit unsystematically in the latter case). That said, the author has reservations about the extent to which the use of GNL can force social change and, finally, has no definite position when it comes to the prescription of GNL practices. In line with Heller’s (2011) stance towards sociolinguistics research, my role as an applied researcher consists of noticing interesting shifts in language management, producing a valuable account of them and encouraging dialogue between academic and non-academic stakeholders.

Policy course of action

I organise the analysis into three sections: policy outcomes, policy idea, and participation and deliberation. The focus of the first section is the analysis of policy documents. In the second section, I combine the analysis of documents and interview data to find out how and why the policy was originated. In the third section, I concentrate on the interview data to examine the participatory process. I provide essential contextual factors and sociopolitical background to enable an examination of the policy process and the justification given for the particular policy course of action.⁵

Policy outcomes

Goals are ‘future states of affairs which agents imagine in accordance with their values and try to bring about by means of action’ (Fairclough & Fairclough, 2012, p. 240). The goal of the regulation is to eliminate sex discrimination and establish a legal regime whereby gender equality is ensured. To do so the regulation lays down the principles and values of gender equality that guide the actions of the city government (effective gender equality, non-discrimination, and intersectionality and gender mainstreaming) and delineates the measures to be taken to incorporate a gender perspective. GNL is often listed in the policy documents along with other equity measures, such as gender parity, reconciliation of work and family life, and measures against sexual harassment. This is important because GNL now becomes as relevant as other broadly supported measures to combat sexism in the workplace. Yet, GNL is not covered extensively in most policy documents, which devote more attention to describing policy actions than to justifying them. For instance, only three out of 25 articles in the regulation deal with GNL. Sexist language is defined as ‘structures and words that tend to exclude or invisibilise women through language’ (3.5, p. 24), which includes masculine generics. GNL, instead, ‘avoids bias towards one sex and treats everyone with the same respect without hiding, excluding or prioritising either sex’ (p. 24).

In addition to goals and values, the structure of practical argumentation also involves circumstances or empirical aspects of the context (Fairclough & Fairclough, 2012, p. 240). In the present case, the results of a recent survey of council staff indicate that reported use of GNL remains low (3.5, p. 71). The regulation adopts two measures to transform these particular circumstances: the use of GNL for the council’s communication and by external contractors. The regulation states that written, graphic and audiovisual documents and those relating to the provision of council services ‘must use inclusive and non-sexist language’ to ‘ensure an equal visibilisation of women and men […] and promote the visibility of sexual and gender diversity’ (1.1, p. 12). Gender equity also requires the use of GNL in public contracts. The mandatory social clause for public procurement contracts establishes that companies working with the council need to ensure that they do not use sexist language and images in their services. Importantly for the understanding of the position of political parties on the regulation, there was only one dissenting voice against this compulsory approach: the liberal party Ciutadans proposed an amendment – which was rejected – to replace compulsory social clauses (including GNL) by an additional scoring system (2.2).

In sum, the regulation incorporates GNL as part of the council’s efforts to combat discrimination against women. The policy sets out an ambitious line of action with measures normalising GNL among council employees while they are carrying out their duties and making its use compulsory for the private sector when they are dealing with the council.⁶ The policy implies not only that a certain linguistic conceptualisation of inclusiveness has been endorsed by local authorities and therefore integrated into a legal-institutional framework but also that linguistic practices other than GNL are now restricted – in particular, the use of standard masculine generics in both Catalan and Spanish.

To establish this last point conclusively, the policy process also resulted in two guides, for both staff and the public, in which the use of non-discriminatory language is exemplified and explained:

• - Guide to Non-Sexist Use of Language (1.3): An easy-to-understand document containing ten points with justifying statements such as

  Excerpt 1

  If you always use masculine generics, you make women invisible and you reinforce stereotypes. Avoid them!

  Using language in a non-sexist way is important: it visibilises women and breaks stereotypes that foster inequality. It is a gesture that will help us build a fairer and more equal world.

  Other points include the use of double gender marking when lexical gender neutralisation is not possible.

• - Guide to Inclusive Communication: An all-encompassing document that focuses on many forms of discrimination (racial, ethnic, cultural and colonial discrimination, and discrimination against members of the LGBTQ+ community or people with disabilities or mental health issues) and linguistic stereotyping. Gender is also addressed but, unlike in the other guide, predominantly from a non-binary perspective. The focus of linguistic examples is on gender-indefinite nouns. Although feminine-masculine pairs are within the options provided, the document warns that ‘by doubling up on gender we exclude people who do not identify as man or woman’ (1.2, p. 15) and further comments on the gender-neutral Swedish pronoun hen and its use and inclusion in the Swedish Academy’s dictionary. Despite this, the guide does not consider alternative non-binary innovations in Spanish such as the use of @, x or e.

Even though inclusiveness is the overarching principle for both guides, each of them has its specific theoretical orientation. The first, focusing on discrimination against women, is shaped by feminist theory and operationalises gender into two discrete linguistic categories (masculine/feminine). Various approaches inform the second one, including transgender theory and therefore this guide tends to include women and men within a broader spectrum of other gender identities. According to the interviewees, the publication of both guides was the agreed decision that followed an intense debate between feminist and LGBTQ+ groups to define gender identities.

Although GNL and inclusive language are two different concepts, this distinction is not always reflected in the way in which documents and speakers refer to linguistic forms that are distinct from the standard language. In my data, the Spanish or Catalan equivalent of ‘inclusive language’ is the most commonly used term, but is mainly used to refer to linguistic practices that avoid gender discrimination, rather than any other form of discrimination. Given this focus on gender, I use the term GNL throughout this article.

Policy idea

The policy action emerges as an idea that might enable agents to achieve transformation, ‘starting from the circumstances they find themselves in and in accordance with their values’ (Fairclough & Fairclough, 2012, p. 236). Having examined the outcomes, I now concentrate on the circumstances and values, that is, how and why the policy began to be developed. By policy idea, I refer to the identification of a problem (standard linguistic gender norms), together with its proposed solution (implementation of GNL) and the programmatic ideas, namely, the guiding plan of action between problem and solution, which involves choosing from among the available policy instruments (Campbell, 2002).

The reason for the inclusion of GNL in the regulation of gender equity is that unbiased language is argued to be one of the ways in which equity can be brought about in society. The idea that language is a key dimension for social transformation is particularly visible in the reflection of Eva:

Excerpt 2

In Barcelona, we started from the idea that in order for the principle of equity to really permeate into society as a whole, different layers, political levels and so on have to be covered. In this sense, language is a very important layer or dimension (…) What language criteria do we establish for equity to be much better reflected and to accomplish more equity? Because language also modifies the world. This double relationship between the world and language is based on this principle [of equity]. (Eva)

Equity is the normative principle that underlies the goals of the policy but this comment is a good example of how beliefs and values about language inform decision-makers’ conceptions of what policy goals should be: because language is believed to modify the world, language practices should be regulated accordingly.

Taking up the question of how the idea was originated, interviewees were very explicit and unambiguous, as the following comment from Marina exemplifies:

Excerpt 3

Since 2015, we’ve had the new government of Barcelona en Comú and their work on feminist matters, to recognise diversity, visibilise through language and, generally, through communication at all levels from a gender perspective … so we’ve had a three-year-long trajectory with a very clear leadership on these issues [GNL and equity] and the political stance of the city government on that was very, very, very clear. (Marina)

Here, the emphasis is on some of the social and political foundations of the leftist programme of the incoming government. However, GNL is not an entirely new issue in the political agenda of the council. To put this in context, different rules at regional, national and European level had been the basis for further action at a local level. A number of provisions on gender mainstreaming were approved only months before the end of the term of the previous centre-right government of Convergència i Unió, so 2015 was the point of departure for gender policies. For instance, the linguistic criteria included in the guidelines for drafting municipal regulations, which were published in 2015, dictate that

Excerpt 4

any sex-based discrimination needs to be eliminated because the implementation of non-sexist language in the administrative domain is one of the principles for public powers. (3.4, p. 27)

Also in 2015, the previous city government drafted a plan for gender equity, and two of that plan’s actions refer to language. An important feature of these actions is their commitment to deliberation and consensus. In this sense, one action aimed to establish the criteria for language by:

Excerpt 5

Looking for experts in matters of gender in language and sharing different perspectives;

Achieving a consensual agreement on the policy on inclusive language use and communication. (3.1, p. 205; emphasis added)

The following government did not implement this plan in the end, but GNL had already entered regulatory policy discourses and was set as a policy solution.

Moving on to consider the programmatic idea, it is noticeable that the 2018 regulation reflects the same objectives around linguistic inclusiveness as the 2015 documents, but with a significant recontextualisation of the original procedural idea. In particular, the new approach did not include a consideration of different perspectives on GNL and policy decision-makers did not pursue a consensual agreement or reasons that are persuasive to all parties as Cohen (1989) puts it (see the following section for a discussion of the actual participatory process). Admittedly, given that language management situations typically call for ‘decisions-in-conflict’ (Ramberg, 2016), the ideal of consensus can be an unrealistic one when faced with linguistic and ideological disagreements such as this one. ‘Meta-consensus’ (Dryzek & Niemeyer, 2010), or agreement based on the legitimacy of competing preferences and the validity of contested judgements about language matters, is probably a more realistic goal of deliberation. In relation to the council, the reasons for the shift in the current policy approach ultimately appear to be an acceptance of the idea that policy-making is a non-neutral enterprise:

Excerpt 6

Author: Was there much debate or articulation of different positions? For example, a person who says ‘look, I think that the generic masculine includes women’ and makes that argument from a certain point of view (…). For instance, I’m thinking about the linguist at the University of Barcelona, Carme Junyent, who has publicly positioned herself against ‘inclusive language’.

Marina: It’s true that we haven’t held a debate in these terms at the municipal level. That is to say, we’ve never, we’ve never sat down to debate in a calm, technical way. And … this specific issue … well, perhaps it would’ve been good to have it [a debate on language]. But, uhm, the political stance of our city government dictates the option to be taken, that is, it’s a political option so the executive and technical departments … what we do is to apply that political programme. This is what happens in public policy, isn’t it? There’s an ideological option which won the elections and public policies are made in accordance with that ideological option.

This comment captures the connections between policy ideas, interests and institutions (Campbell, 2002) and shows how ideological expediency and political opportunities resulting from the institutional contingencies caused by elections may determine the choice of a particular course of action over others. Plain and simple, the time and the circumstances were right for the adoption of GNL.

Participation and deliberation

In line with the city government’s commitment to developing participatory institutions, the policy process incorporated a participatory and deliberative mechanism. The participatory model aimed to avoid technocratic and bureaucratic procedures in policy development and integrate the demands of the feminist movement by coproducing the regulation with them (2.3). Therefore, participation was not only about involving and consulting citizens. It consisted of decision-making and policy-forming in conjunction with feminists and other groups. The groups who were invited to participate were essentially women and LGBTQ+ activists and organisations for gender equity in Barcelona. Most of these groups had a pre-existing relationship with the council and there was a moderate or high degree of trust among them (Beierle & Cayford, 2002). Political parties, unions, women lawyers’ associations, research groups, scholars in gender studies and citizens also took part in the process. Importantly, the selection of the participants did not include linguists. The driving group was formed of representatives from council departments working on feminism, women’s and LGBTQ+ rights and gender mainstreaming. This group nominated a monitoring group consisting of a number of local women’s groups and LGBTQ+ associations.

The process lasted six months (July-December 2018) and had five phases (information, analysis, proposals, follow-up and assessment). There were face-to-face and virtual sessions, both among council employees and the broader community of Barcelona. The interviewees reported that the participatory process primarily involved council employees or members of groups who were expressly invited to take part, and that participation by men and contributions from private citizens were very limited. The interviewees argued that as the regulatory policy is mainly for internal use, other groups and individual citizens did not feel called upon to intervene in the process.

The conceptual debates that formed part of the policy process revolved around issues currently under discussion within the feminist movement (e.g. the delimitation of gender categories and the consideration of various forms of inequity). The interviewees attested that proposals on GNL generated a broad consensus in the participatory process and that participants did not deliberate on this specific topic:

Excerpt 7

We opened up participation to feminist groups and movements, and other women's organisations in the city, and everyone agreed. In other words, they valued the [regulation’s] article [about language] positively … so we didn't have any problems or difficulties there […] we did not have to make efforts [to integrate different perspectives]; all this was part of the strategy of the city council. (Marina)

The interviewees explained that language issues were peripheral because the deliberation was on the regulation project as a whole and only one of the articles focused on GNL. This is one of the reasons that was given for why critical or sceptical voices on GNL were not incorporated into the process. Other reasons put forward included the fact that priority was given to groups directly affected by gender policies (mainly women but also those who identify as LGBTQ+) and, as seen in the previous two extracts, the ideological character of the initiative, which is derived from a political programme. As a result, those elements of the draft document relating to GNL remained unchanged throughout the policy process (1.1, 2.1).

Deliberative practices are not a neutral enterprise because it is unrealistic to expect participants to adopt a neutral perspective on language (Wee, 2011). Yet, unlike in other LPP contexts in which deliberation reflected disparate views and different orientations (see e.g. Røyneland, 2016 regarding the revision of the Norwegian minority standard Nynorsk), the deliberation process in relation to GNL did not seem to ensure a ‘discursive representation’, namely the balancing of reasons for and against a proposed action (Dryzek & Niemeyer, 2010). The interviewees remarked that the absence of widespread participation and the lack of different viewpoints from outside the designated groups do not undermine the legitimacy of the policy process nor the enforcement of its outcomes. This is despite the fact that they are aware of the mixed reactions that the GNL regulation provokes:

Excerpt 8

Some sectors don’t like this, it’s true. But well, this is what public administration is about, isn’t it? Public administrations regulate. (Carme)

This view is not consistent with the idea of legitimacy in political theory, which applies not only to the decision-making process but also to the composition of decision-making bodies (Røyneland, 2016). Within the council, not everyone had the same opinion about GNL and there were opposing views among staff members:

Excerpt 9

Some people have reservations in the city council about the issue of inclusive language. Yes, especially from departments dealing with linguistic issues: people who make linguistic corrections, well philologists [‘filólogos y filólogas’], who in some way, well … have always shown some resistance … I think it’s connected to the … to linguistic orthodoxy, right? (Marina)

But people with discordant voices did not speak up during the process. Marina explains the reasons in the following terms:

Excerpt 10

Because it wasn’t politically correct, so to speak, nobody dared to question the compulsory nature of the gender perspective in the use of inclusive language, images and communication. That is, it wasn’t the moment to question that, you know? (Marina)

This comment suggests that there was an adverse environment for stances in favour of the standard norms and that, again, the participatory process decidedly reinforced certain ideas and values. Moreover, there were hierarchies and power relations at play behind the argumentation that took place. The following example by Carme serves to illustrate this point:

Excerpt 11

Some female colleagues think that masculine generic includes women. Luckily … well, luckily or whatever, people working in gender matters in this council do not agree and, actually, the Direction of Communication is in charge of the instructions about the use and the kind of communication we need to do. (Carme)

Here, the hegemony of the meanings about inclusiveness stems from key institutional areas that have the capacity to influence decision-making.

Another important issue is the process of decision-making related to the choice of appropriate alternative linguistic conventions. As noted above, the guides propose the use of masculine-feminine pairs when gender-neutralisation is not possible, even though they admit that this option can be exclusionary for non-binary people. Yet, the guides do not go so far as to propose the use of anti-canonical innovations such as the non-gendered morphemes x and e. Arguments for disregarding these forms in the institutional realm – impracticality in certain settings and the possibility that the forms will not be understood – sometimes resemble those brought forward in other contexts to argue against the adoption of the most commonly used forms in GNL (e.g. masculine-feminine pairs). This can be seen in the following example: when asked about the exclusion of the x and e forms, Carme responds that

Excerpt 12

we opted to look for generic terms. I always try to avoid double gender marking [‘los/las’]. If I’m home or on the family chat I use x. Personally, I have some issues with using e … I don’t know. I don’t like how it sounds […] For sure in an institution you can’t use e; you can’t use x, right? Because it’s also for accessibility reasons, we need to understand each other … it’s true that, of course, language generates much, much controversy, you know it well. (Carme)

This is an indication of how values and perceptions about institutional constraints interact when taking action to solve a policy problem (Fairclough & Fairclough, 2012). The current perceptions around the appropriateness of alternative forms seem to indicate that non-gendered morphemes are still a long way from being legitimised in formal settings, even by groups who are open to non-standard forms of language.

Discussion: lessons from the case study

In this paper, I have investigated the institutionalisation of GNL through participatory and deliberative methods. I have shown how the interaction of goals, values and circumstances influenced the institutional design of the participatory process. More explicitly, I have provided evidence that this particular policy process and its outcomes were directed by a predetermined political and ideological goal to implement an unargued, and apparently unarguable, orientation towards language based on the conviction that the policy promotes women’s inclusiveness and visibilisation. This case is noteworthy because the problematisation of language, in the form of how normative language deals with gender inclusiveness, did not lead to treating the very nature of linguistic discrimination as a problem to be effectively discussed.

While the policy model was open to public participation and problems under discussion were putatively contestable, the process did not seem to include different visions of language provided by divergent voices and discourses. There was a disproportionate representation of one-sided understandings of inclusiveness through language (that the use of alternative forms, e.g. double-gender marking, would lead to greater linguistic and social inclusiveness) in the process. In ideal deliberation, parties are equal: everyone with deliberative capacities has equal standing in the deliberative process and the existing distribution of power does not shape the chances of participants to contribute to deliberation (Cohen, 1989). In the present case, the outcome was not equal in the sense that not all parties had their position and claims counted. To be specific, the process should have included linguists and speakers who claim that most of the standard norms (e.g. masculine generics) are neither sexist nor discriminatory. The ideology of gender equity of some decision-makers and the result of the debate between feminist and LGBTQ+ groups about the definition of gender identities seem to have prioritised binary gender equity. This seems to be one of the reasons for non-binary genders to occupy a less prominent position in the outcome (e.g. guides for the use of non-discriminatory language). As seen above, another reason is that the grammatical binary of the language could only be effectively resolved by what planners depict as unacceptable strategies (the use of non-binary endings @, x and e when lexical gender neutralisation is not possible) in institutional settings.

The analysis of this case study shows that practical argumentation is at the interface between agency and structure (Fairclough & Fairclough, 2012, p. 244): decision-makers’ argumentative discourses about the policy idea and the course of action taken are built upon reasons such as ideological decisions, political agendas and the functioning of public power. Considering this, the apparent consensus reached on the proposals for language reform in the regulatory policy may well be a provisional hegemony based on the exercise of control over institutional resources rather than an indefeasible agreement derived from the ‘unforced force of the better argument’ (Habermas, 1996, p. 306). Moreover, the power relations behind the argumentation of decision-makers suggest that the participatory model did not break with the hegemonic logics in the institutional sphere (Fairclough, 2013). This point seems to uphold the idea that interests and power, which are inherent to the essence of politics (Shapiro, 1999), constitute a potential barrier to deliberation on language. The decision-makers acknowledged the controversial nature of language use and regulation and the existence of different sensibilities and linguistic practices both within and outside the council. Yet, there were no counter-discourses nor counterpublics in the policy process, which is essential for deliberative governance (Dryzek & Niemeyer, 2010). The process ultimately ended up in a discursive arena of equity matters other than the relation between language and discrimination. Thus, this case shows how decision-making power operated to not consider language as being a debateable problem and, in turn, focused the discussion on other genuine problems. Hence, all this inevitably affects the credibility of the deliberative approach to language issues that was taken.

It goes without saying that sex-based discrimination is unacceptable and gender inequity needs to be combated and eradicated. That said, inserting a prescriptivist approach for GNL, which is still a contested approach in the Spanish context (Junyent, 2021; Lomotey, 2018), into an all-encompassing legal framework with other broadly supported measures to confront sexism, such as equal treatment and pay, may not be a sensible choice. In fact, this case is in line with other well-intentioned open participatory processes (e.g. Erdocia, 2020) in which institutional actors merge language issues into a normative framework along with disparate social, political and cultural elements, thus blurring the specificities of language-in society as an object of enquiry, treatment and discussion. Language, I believe, deserves to be allocated its own deliberative arena in social and political debates on which (socio)linguistic issues occupy a central position.

What are the implications of this study for the inclusion of democratic policy processes and their contribution to policy development? Language policy is all about choices and decision-making, that is, about who make the decisions and how these are reached (Spolsky, 2009). I suggest that if local modes of governance (as opposed to the centralised, authoritative role assigned to language academies) are going to be used to prescribe language practices in institutions, arrangements should be put in place to provide access to a full range of views about language (Ramberg, 2016; Røyneland, 2016). Systems of governance should also include an appropriate environment for the public exchange of arguments and for making reflective judgements about language (Wee, 2011) that may lead to the transformation of linguistic assumptions and the resolution of language-related grievances. Such an approach implies freedom from bias and institutional constraints, among other things. This is a reasonable condition to be expected for public power to legitimately require compliance with the regulatory practices on GNL communication if a decision is made to that effect. Most certainly, the dynamics of such a system could constrain the role of linguists (who, unlike in the present case, should contribute to the discussion) but would not undermine the legitimacy of decision-making (Ramberg, 2016). If there is equal participation in the deliberative process, this provides a strong case for compelling external bodies to comply – in the present context, for private companies who deal with the council to use GNL – as the fact that the outcomes result from actual deliberation serves to justify making the regulations mandatory. From a deliberative democracy perspective, this would be a legitimate exercise of coercive public authority (Gaus, 2003), one that could prove to be a more compelling argument than the discretionary power of public administration that decision-makers invoked in the present discussion.

Concluding remarks

This case presents a real-world language problem in which there is a strong value preference for a given conception of language and language use (the acceptability of certain alternative forms) and this is supported at the expense of other interpretations and uses (primarily standard norms but also other innovations, e.g. the use of @, x or e). It remains to be seen if deliberative and participatory modes of operation in a more ‘open’ democracy can give rise to a satisfactory and workable solution – be it linguistic or pertaining to the regulatory framework (mandatory/recommended/voluntary prescription of language reform) – for all the competing publics. It is not yet clear whether deliberation-centred language policy-making can promote a shared way forward in cases of divisive or perhaps irresolvable linguistic issues, one that all members of the public can embrace. In fact, there may be some cases in which language does not conform to a model of rational discussion oriented towards mutual understanding and reasonable compromise (Gardiner, 2004). As Krause (2008) notes, emotions, passions and affect can undermine reasoned judgements in public deliberation. Further research is needed to determine the extent to which the affective dimension plays a role in deliberation in relation to language policy (Ramberg, 2016).

The present case study is perhaps one such case in which the only normative outcome may be ‘decision, not consensus’ (Fairclough & Fairclough, 2012, p. 241; emphasis original). In any case, achieving a position of mutual recognition and respect for conflicting beliefs (Dryzek & Niemeyer, 2010) is probably an important first step in the regulation of GNL. I believe, along with Lo Bianco (2013), that these points deserve further scrutiny as they ultimately call for a research agenda on LPP with a problem-solving orientation, one that safeguards the specificities of language from all-encompassing procedures supposedly ready to be satisfactorily implemented without considering contextual factors (Erdocia, 2020). Admittedly, deliberation has been successfully applied in bottom-up public policy processes on language in contested multi-ethnic settings and deliberative practices can generate more reflective kinds of language policy-making (Lo Bianco, 2017). That said, this study serves as a cautionary example of how actual governance practices may sometimes result in less than successful processes.

I conclude by claiming that the transition from a top-down tradition of the imposition of expert knowledge on language to a more democratic procedural approach (or the combination of the two) should not be at the expense of dismissing appeals to authority or silencing dissent from a politically, ideologically and morally valued linguistic trend (in this case, GNL). Standard norms are certainly a dominating force but also involve legitimate credentials based on the authority of history and experts (see Van Leeuwen, 2008). Put another way, collective managerial decision-making on language should not circumvent our ongoing scholarly discussions about sexism and language, nor overstress the value-laden aspects of language disputes. This study contributes to pointing out the sometimes scant consideration given to the intricacies of language by debate-centred policy-making. But the reason for this, I think, is not to be found in our scholarly disagreements (see Erdocia & Soler, 2021). Most likely, this may exemplify another case of the practical inability of applied linguists and LPP scholars to reach relevant stakeholders and policy-decision centres.

Acknowledgements

My sincere gratitude goes to my informants, without whom this article would not have been possible. I would also like to express my thanks to Jennifer Bruen, who read an early version of the article and provided insightful suggestions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Faculty of Humanities and Social Sciences, Dublin City University.

Notes on contributors

Iker Erdocia

Iker Erdocia is Assistant Professor in the School of Applied Language and Intercultural Studies at Dublin City University. His research sits at the interface of language, policy and politics and aims to promote debate on language matters between academic and non-academic stakeholders.

Notes

1 While Barcelona City Council assigns a preferential status to Catalan over Spanish, the documents under analysis are in both languages so I use them both as a reference here. I prioritise Spanish for the contextualisation because my interviewees referred to examples in this language in their argumentation.

2 Some lexico-grammatical dissymmetries are widely depicted as sexist practices (see Bengoechea, 2011, pp. 36–37).

3 An informed consent to participate was obtained. The Dublin City University Research Ethics Committee approved this research (DCUREC/2019/219).

4 To guarantee the anonymity and confidentiality of participants, the description I provide here is deliberately limited. I use the pseudonyms Carme, Marina and Eva to refer to the participants.

5 All extracts in this section have been translated by the author.

6 Note that I do not include the case of language practices in private contexts (e.g. among council staff members) in my following argument, for which a line of demarcation should be set. Since the regulation excludes them, this point is beyond the aims of this article.