Discriminatory language in the workplace: unmasking prejudices and stereotypes / El lenguaje discriminatorio en contextos laborales: desenmascarando prejuicios y estereotipos

Abstract The link between language, communication and open discrimination has been the focus of numerous studies. For many years, language has been a vehicle of prejudice and negative stereotypes. However, in the last few years, overt verbal aggressions have diminished thanks to the implementation of equality policies and awareness-raising initiatives. Despite that, negative attitudes against protected groups might not have disappeared and may instead be transmitted subtly, such as through uncivil language. The goal of this research is to study the relation between language and subtle discrimination against lesbian and gay (LG) workers. With this goal in mind, 39 in-depth interviews were conducted. The analysis showed that LG people are victims of subtle discrimination and that language is one of the most common vehicles of prejudice and negative stereotypes. Jokes, inappropriate comments and questions are the expression of negative attitudes, which are probably present at an implicit level. Due to the ambiguity of these expressions and the settings where they have been studied — the workplace — these acts could be defined as uncivil behaviours. Practical implications, limitations and directions for future research are discussed.

Palabras clave: lenguaje discriminatorio; trabajadores/as gays y lesbianas; comportamientos incívicos en el trabajo; discriminación sutil; prejuicios y estereotipos negativos In recent years, there have been numerous efforts to educate and create more inclusive societies and organizations. Thanks to the fight against discrimination, many behaviours which openly harmed the rights of minority or protected groups (e.g., women, the LGBTQ community, individuals with disabilities, etc.) in the past have been diminishing. However, recent studies reveal that the affective processes, which are behind prejudices (negative attitudes towards the members of a given group), and cognitive processes, which are behind negative stereotypes (beliefs about these groups) (Mor Barak, 2017), have not disappeared but instead are expressed in a different way and are potential sources of subtle discriminatory behaviours (Cortina, 2008;Jones, Peddie, Gilrane, King, & Gray, 2016;Moya, 1997). Researchers' interest (Di Marco, Hoel, Arenas, & Munduate, 2018;Gómez & Huici, 1999) has shifted towards these behaviours.
The process described above has been verified with language, which has traditionally been studied as a vehicle for openly transmitting negative attitudes towards protected groups; these behaviours include insults or verbal aggressions towards people in these groups. Today, open verbal aggressions towards protected groups occur less frequently, especially in certain settings such as the workplace, where there are usually rules regarding mutual respect. Nonetheless, other studies have shown that language can also act subtly (e.g., Bianchi, Piccoli, Zotti, Fasoli, & Carnaghi, 2017;Collins & Clement, 2012), giving rise to uncivil behaviours which are not openly aggressive but are equally discriminatory (Di Marco et al., 2018;Jones et al., 2016).
This study analyses the discriminatory aspects of uncivil behaviours towards lesbians and gay men (LG) in Spanish workplaces. Specifically, we will focus on subtle behaviours which are expressed through the use of language during communicative interactions. We have analysed these groups because of the heavy stigmatization they have experienced in all societies until quite recently, a stigmatization of which they are still victims in some of these societies. Recent studies stress the sexual stigma (Herek & McLemore, 2013), meant as lower status and power among people who belong to these groups. Even though in Spain the struggle of lesbian, gay, transsexual and bisexual (LGTB) people has made major inroads, giving rise to different legislative reforms (e.g., the law allowing for same-sex marriage, which also allows non-heterosexual couples to adopt children; Spanish Law 13/2005), data on discrimination against LGBT people in Spain nonetheless reveal a contradictory picture. Specifically, 56% of Spaniards believe that discriminatory behaviours towards LGBT people are still widespread (European Commission, 2015). Surveys conducted with LGBT people reveal that 31% of the workers from this group have experienced some discriminatory situation in the workplace (López, Generelo, & Arroyo, 2013). In Madrid, the capital of Spain, a homophobic aggression is reported every other day (Rosati, 2016).
The current gap between the ideal situation embodied in the laws and reality may be rooted in the presence of implicit prejudices and negative stereotypes (Cortina, 2008;Dovidio, 2001;Fiske, Dupree, Nicolas, & Swencionis, 2016;Hebl, Bigazzi, Mannix, & Dovidio, 2002;Jones et al., 2016). That is, even though most members of Spanish society reject expressions of discrimination, these prejudices and stereotypes may still be shaping people's behaviour without them being aware of it.
Starting with a literature survey of the findings on the relationship between language and discrimination on the one hand and workplace incivility and open and subtle discriminatory processes on the other, we will explore the results of the study carried out in Spain to achieve the goal of this research. Finally, we will highlight the practical implications for organizations that want to build a non-discriminatory and inclusive workplace.
Language as a vehicle of discrimination Even though the majority of societies reject and condemn all kinds of discrimination, the international data on perceived discrimination among protected groups are still alarming (e.g., Comisión Europea, 2015). To explain this dissonance, numerous studies state that prejudices and negative stereotypes, which are behind these behaviours, have often shifted to the implicit level, transforming the way they are harnessed and expressed (e.g., Dovidio, 2001;Jones et al., 2016). For example, people who have been raised in extremely macho environments may have shifted their opinion on the role of women in society throughout their lifetime thanks to the influence of the people with whom they have lived or interacted as adults. Nonetheless, the values acquired previously may not have fully disappeared and may be present implicitly. This process can even affect these people's behaviour in the way they treat women who do not align with traditional gender roles.
In line with the above, in recent years numerous studies have examined the nexus between language, communication and discrimination (e.g., Cocchiara, Bell, & Casper, 2016;Dueñas, Pontón, Belzunegui, & Pastor, 2016;Ng, 2007;Sachdev, 2007). Several perspectives have been adopted in these studies: while some of them have focused on the effectsboth implicit and explicitof prejudices and stereotypes expressed through language in the victims and on the quality of interactions (e.g., Collins & Clement, 2012), others have probed how a certain kind of accent or dialect affects the receiver's perceptions and decisionmaking (e.g., Gowen & Britt, 2006). Yet other studies have examined how language and the use of labels, especially derogatory ones, affect the people targeted by them (e.g., Bianchi et al., 2017).
Prejudices and stereotypes can shape language without people being aware of it (Collins & Clement, 2012;Sutton, 2010), giving rise to implicit discrimination. According to Ng (2007), the discriminatory potential of language can take on different aspects. First, the lack of knowledge of a given language may be the legal justification for treating certain groups in a discriminatory fashion. One example is an employer's prejudices towards foreign candidates, which they shield behind the pretext of the candidate's lack of knowledge of the language of the majority group.
Secondly, language can reflect historical inequalities among the social groups in a given society. One example of this is sexist or racist language, which reproducesoftentimes without the interlocutors realizing itthe stereotypes and power inequality which have long characterized, and continue to characterize, intergroup relations between men and women or among different ethnic groups.
Thirdly, if we consider language as a social act, it could be a discriminatory behaviour in and of itself through choice of words or certain grammatical constructions (e.g., active versus passive constructions), which tend to focus the interlocutor's attention on a given aspect or content within the discourse. Language can also be a discriminatory act in monolingual or bilingual contexts: in the former, if one chooses to use technical jargon which is unclear to the majority even if all the persons involved in a given communicative interaction share the same language; in the latter, when one of the languages not spoken by all the interlocutors is chosen.
The fourth and final aspect that Ng (2007) takes into account in his study is the normalization and standardization of discriminatory language. Stereotypes enter the daily language, reproducing the power inequality among groups and taking root in popular beliefs. Through primary and secondary socialization in the family, at school and in peer groups, we learn expressions and ways of speaking whose discriminatory nature is concealed. One example is the use of the masculine gender to refer to a group of men and women, a consequence of women's invisibility in the past.
The discriminatory forms of language are often manifested subtly, and there are doubts regarding the level of awareness of the people who perpetrate these acts. A joke, for example, could conceal its sexist content behind humour and not be perceived as damaging by the people who tell or hear it. Nonetheless, it could be propagating traditional gender roles or denigrating women. Given the ambiguity of these behaviours and the lack of clarity regarding the intention to damage the interlocutors, it is possible to classify these acts as uncivil because they violate the rules of mutual respect established in a given context without being openly discriminatory. Uncivil acts in the workplace began to be studied in the late 1990s (Andersson & Pearson, 1999;Pearson, Andersson, & Wegner, 2001). Even though workplace incivility encompasses verbal and other kinds of behaviours, it may be the expression of prejudices and negative stereotypes, as we explain below.
Subtle discrimination and workplace incivility Subtle discrimination can be manifested in all spheres of life, including the workplace. In the late 1990s, Andersson and Pearson (1999) began to study behaviours which are not openly aggressive but which nonetheless show a lack of respect towards the people receiving them. Their studies referred to workplace incivility, the hallmarks of which are the low intensity of the acts carried out and the high degree of ambiguity when attributing to the perpetrator the intentionality of those behaviours. Shouting, insulting a colleague or making fun of a colleague are all acts that could be considered inoffensive, be attributed to other causes (e.g., the perpetrator having a bad day) or be considered an expression of a more complex phenomenon which may fit within the behaviours defined as negative (Andersson & Pearson, 1999;Pearson, Andersson, & Porath, 2000;Pearson et al., 2001).
Recent studies have found that workplace incivility can be an expression of prejudices and negative stereotypes towards people belonging to vulnerable groups (e.g., Cortina, 2008). Therefore, uncivil acts at work may be aimed at expressingeither consciously or unconsciouslythe prejudices and negative stereotypes that have previously been internalized by the person. Researchers talk specifically about 'selective' uncivil behaviours (Cortina, Kabat-Farr, Leskinen, Huerta, & Magley, 2013;Cortina, Magley, Williams, & Langhout, 2001), that is, those primarily targeted at vulnerable groups. Indeed, studies conducted in the United States (Cortina et al., 2013(Cortina et al., , 2001 found that women were the recipients of significantly more uncivil acts than men, and the same was found in the case of Black workers compared to their white counterparts.
Workplace incivility can be manifested in different ways. The scale developed by Cortina et al. (2013Cortina et al. ( , 2001 refers to several acts which entail being isolated, being the recipient of yelling or insults, being the target of disparaging language, being the victim of inappropriate jokes, etc. As is clear, the inappropriate use of language is considered the vehicle of uncivil behaviours and ultimately of prejudices and negative stereotypes. The purpose of this research is to identify the verbal acts perceived as uncivil by LG workers which might potentially be an expression of subtle discrimination.

Method
This study is based on a broader project undertaken with the objective of analysing the experience of LG persons in Spanish workplaces. With this goal in mind, 39 in-depth interviews were held in different Spanish cities. While the interview touches upon different topics (e.g., the coming-out process, the influence of sociocultural factors in the experience of LG people, etc.), this study considers the results related to language as a vehicle of uncivil acts.
The choice of in-depth interviews as a data-collection technique is due to the sensitivity of the topic at hand (King, 2004a). This data-collection technique allows a trusting relationship to be created between the interviewer and interviewee, helping to shed light on a phenomenon from the perspective of the people involved in it (King, 2004a). Given the difficulty of finding participants, we used the snowballing approach (Miles & Huberman, 1994), which activates existing social networks starting with the first participants. Thus, the description of the research was published on the website of the National Federation of Lesbians, Gays, Transsexuals and Bisexuals (Federación Estatal de Lesbianas, Gays, Transexuales y Bisexuales -FELGTB) of Spain. As the word got out, the first volunteers got in touch with the research group, who travelled to the places where the participants lived. After the interviews, we asked the interviewees to get in touch with their friends and acquaintances who may be interested in participating. A total of 24 lesbians and 15 gay men were interviewed. They had different levels of education (18% secondary school, 56% bachelor's degrees and 36% post-graduate degrees), and their mean age was 36.18 years old (SD = 8.72). They were all working or had been working in the past six months, and they represented different activity sectors, including construction, education, finance, marketing, health, etc. (see Table 1).
The interviews were held by the first author and lasted between 60 and 90 minutes. They were all held in neutral settings (e.g., cafés) and never at the workplace, so that the participants could freely discuss their experiences at work. After an initial set of questions to break the ice (e.g., 'What is your job?', 'How many people work with you?'), the interview explored the participants' work experience with regard to their sexual orientation via questions aimed at exploring the coming-out process and any uncivil behaviours they have perceived (e.g., 'What is it like to work in this organization in terms of your sexual orientation?', 'Are your colleagues aware of your sexual orientation?', 'Have you heard any comments about LG people at work?', etc.).
The data collection ended when saturation was reached (Morse, 2000), that is, when new data did not contribute any new information to the research.
The interviews were recorded and transcribed verbatim. Conversational transcription was done in order to not lose any information coming from nonverbal communication (D'Agostino, 2007). This transcription allowed us to use conventional signs to reproduce in writing the pauses, silences, hesitations, etc., of the participants so they could be borne in mind during the analysis process. The data were analysed using the software Atlas.ti 7, a tool that helps systematize the relevant data. The technique used for the analysis is called Template Analysis (King, 2004b), which enables codes to be found a priori (before holding the interviews), thanks to the literature review, as well as a posteriori (after holding the interviews). For example, the literature suggests that jokes, pranks and direct aggression can be discriminatory behaviours expressed via language (e.g., Douglass, Mirpuri, English, & Yip, 2016), among other possibilities, and for this reason, these elements were included in the list of codes developed a priori. Once the interviews were analysed, other codes were added to the initial list, and those which were not reflected in the information analysed were eliminated or changed. For example, we decided to eliminate the 'direct aggressions' code given that most of the experiences recounted by the participants entailed more subtle negative behaviours. The rigour of the data analysis was guaranteed through systematic teamwork. Therefore, the list of codes (a priori) was developed separately by each of the authors. The results of this individual work were compared before beginning the data analysis. The authors reach agreement on the codes identified in the majority of cases, and an adjustment of the labels attributed was only needed once. Cases in which there was any disagreement were resolved through debate. After identifying the first list of categories, the first and second authors analysed the interviews individually, leading to changes in that list. Throughout the data analysis, the authors held weekly meetings to discuss the results and resolve any cases of disagreement via debate.
During the analysis process, the most important categories related to the research question of this study came clearly into focus. Below are the results of the analysis process, with an emphasis on the results related to discriminatory behaviours towards the LG workers associated with the use of language. The names used in the quotes are fictitious in order to guarantee the anonymity of all the participants.

Results
The analysis of the interviews showed that LG people are the victims of uncivil acts in the workplace in Spain. These acts are expressed in different ways, such as through isolation, but they mostly occur through the use of language. Among the uncivil behaviours via language, there is a set of experiences which underscore the existence of prejudices and negative stereotypes, which are manifested subtly in social interactions.
Among the most common uncivil acts, we find the use of derogatory labels. The words 'marica' and 'maricón' (both of which could be translated as 'faggot') are commonly used to refer disparagingly to a gay person, as recounted by the informant Alejandro. In a previous job he held in the construction industry, after reprimanding a subordinate for working without the required equipment, he got the following response: A faggot is going to come and give me orders! (G, Alejandro) As can be seen in this quotation, not only was the response disparaging, but it also implicitly questioned the supervisory abilities of the participant due to his sexual orientation. As Ng (2007) reminded us in his classification, language can reproduce historical power inequalities: in the past, it was unthinkable for an openly homosexual person to occupy a position of power. This belief is reproduced in the words of the worker above, who when angered instinctively expressed the stereotypes and prejudices he possesses via language.
Generally speaking, women recount fewer cases of direct uncivil attacks through disparaging language. The derogatory terms associated with lesbians (e.g., 'bollera' ['dyke'], 'marimacho' ['butch']) are used less often than male labels: No one has attacked me for being a lesbian. I'm always afraid of that, but it's never been direct […] In guys, I have seen them be attacked and they've been told 'That faggot…' (L, Violeta) However, the more infrequent presence of these expressions towards lesbians may be associated with the invisibility they have experienced throughout the course of history (e.g., Juliano, 2012), especially when their physical appearance did not align with the common image many people have about lesbian women (e.g., an unfeminine woman with short hair, etc.). While in the past male homosexuality was the object of an open process of stigmatization, lesbianism was not even considered (Juliano, 2012). Language could implicitly be reflecting this invisibility.
On the one hand, labels for lesbians are heard less frequently and are still quite disparaging; however, the words 'marica' and 'maricón' have become normalized in everyday language, as claimed by the following participant: They're set phrases which even my father uses every now and then. 'This is nonsense', 'Stop being silly, what a bunch of nonsense' […] expressions which you always find on the street. (G, Álvaro) 1 According to Álvaro, these terms are not only used to refer to non-heterosexual people. While still disparaging, these forms of expression have become part of our everyday language and are often associated with something negative or someone who is wrong, regardless of the sexual orientation of the person who is being described with this adjective. As the next participant explains: As we can see, several participants stress how over time the words 'maricón', 'mariconada', etc., have lost their original meaning and are less associated with sexual orientation. Nonetheless, the negative charge which these terms have always carried has not vanished. Unconsciously, the prejudices and negative stereotypes anchored in these expressions have become visible via language.
Another uncivil behaviour that the participants have witnessed is spreading jokes whose main characters are LG people. These jokes are told even when the teller knows that there are non-heterosexual persons around them. According to the LG interviewees, their colleagues justify this by the humorous aspect of these behaviours without taking into account the fact that they are expressing negative attitudes towards and negatively affecting the LG people present.
[In my previous job] they would tell humiliating, ugly jokes and there I wasn't as strong as I was later in my other job to say, 'What are you saying?' [I would hear] unfair things and unpleasant comments and I was incapable […] I didn't support them, I remained silent. (L, Elena) It is important to bear in mind that the degree to which LG persons reveal their sexual orientation varies according to the person, context and situation (Clair, Beatty, & MacLean, 2005;Ragins, 2008), and this determines the LG workers' reactions to the comments and jokes that they hear at the workplace. In the previous case, the participant had not yet come out and her image did not reflect the one that society associates with lesbians. Admonishing a colleague for having offended LG groups may be an act of civility, but it could give rise to doubts about the other person's sexual orientation. On the other hand, not reacting after an event of this kind creates a sense of impotence that affects LG workers.
As mentioned above, it is possible to consider language a social act (Ng, 2007). This act can sometimes be revealed to be discriminatory. One example is questions that are inappropriate or out of place in the workplace (e.g., questions about one's private life), as illustrated by the next participant's story, after she informed her boss of her sexual orientation and forthcoming wedding: […] So, then we started talking. 'Why are you getting married? Do you think it's necessary?' This is a topic of conversation that does not usually arise when you're talking to a 'normal' The questions asked may seem innocent and not meant to harm the other party, yet they are the symptom of the failure for non-heterosexual orientations to be normalized. Heteronormativitythat is, the norms and values that define heterosexual orientation as normative and socially accepted (Losert, 2008;Warner, 1991) still exists in Spanish society, and anything that does not align with these norms is the target of questions and curiosity. A common and seldom-noticed event for heterosexual couples, like getting married, becomes a focal point of attention for LG people.
As the same participant recounts, in some cases the comments can become even more intrusive, touching on aspects of their private life which are rarely shared in the workplace. This could be related to the image of promiscuity (Rosenfeld, 2009), which was associated with non-heterosexual people for years and which still exists in some sectors of society today.
The absence of language has also been shown to be a discriminatory social act. One example is LG people who cannot share information about their family or free time because they implicitly convey information about their sexual orientation (such as when they talk about activities they have done with their partner on the weekend). The absence of questions to LG colleagues can be experienced as rejection: [There are] times when I feel rejected [referring to her boss]; he prevents me from saying, that is, from speaking. (L, Inés) The rejection expressed through absence can also be considered an uncivil act and can convey prejudices and negative stereotypes: given the lack of normalization of non-heterosexual people, the preference is to avoid any topics that refer to the other person's sexual orientation, even implicitly.
Finally, language once again becomes uncivil and simultaneously discriminatory simply via grammatical structures (Ng, 2007). Dorleta, an employee in a public institution, mentions this regarding the interaction among her colleagues and a lesbian user: I didn't deal with her directly, but when I heard the comments […] they used the word 'lesbian' and then descriptive adjectives: 'That lesbian, she's imposing, she's really aggressive, she's rude!' […] It was like associating in the same sentence her sexual orientation … I mean, you don't say, 'This heterosexual woman is aggressive, who did she think she was'. You don't say it, so why do you insert the word 'lesbian' in the middle? (L, Dorleta) We can see how in the sentence itself a user's aggressiveness is associated with her sexual orientation, implicitly becoming a feature that characterizes lesbians as a group. Once again, language is implicitly the bearer of heterosexuals' stereotypes about LG groups.

Discussion
The results presented above show that LG workers perceive themselves to be the victims of uncivil acts. These acts are largely expressed through language, which transmits prejudices and negative stereotypes related to non-heterosexual orientations. The use of derogatory labels for LG people, referring to homosexuality as something negative, as well as jokes, inappropriate questions or the absence of questions about daily life, and/or the grammar constructions themselves are the outcome of prejudices and negative stereotypes about language and vice versa (Collins & Clement, 2012). In the majority of cases, prejudices and negative stereotypes are present implicitly: language acts as a 'barometer' (Sutton, 2010) that reveals people's real attitudes and beliefs towards their LG colleagues.
These behaviours are problematic because it is difficult to identify their discriminatory side. A joke or the absence of questions about one's personal life could be seen as inoffensive acts. A joke could be excused by its humorous veneer, and the lack of questions could be interpreted as excessive respect for the LG colleague's privacy. Furthermore, an inability to identify the discriminatory nature of certain behaviours prevents us from being able to all join forces in fighting against these expressions. Failing to recognize an enemy is an obstacle to combating that enemy and a threat for organizations that want to educate and train their workers in respect for differences.
Several factors can shape the relationship between language, prejudices and negative stereotypes, among them the sense of belonging to a given group and social norms (Collins & Clement, 2012). Thus, previous studies (e.g., Collins, Biernat, & Eidelman, 2009) have shown that people show more prejudices when they talk about members of other ethnic groups. Obviously, each person's ethnic group cannot be changed, but a sense of belonging can be created via other factors, creating organizational groups that include a set of workers with shared objectives (Cortina, 2008;Hewstone, Rubin, & Willis, 2002;Pettigrew, 1998).
Social and organizational norms agreed upon by the majority can also moderate the relationship between language, prejudices and negative stereotypes (Collins & Clement, 2012). This has major implications for organizations that want to develop an inclusive, discrimination-free setting. Establishing rules of respect and conveying them through continuous training of workers is the first step in limiting the adverse effects of discriminatory language. Therefore, making these kinds of behaviours and their consequences for their victims visible, setting punishments for people who do not respect these rules and/or providing an anonymous complaint system for workers who feel like the victims of prejudice (Capell, Dolan, & Tzafrir, 2013;Di Marco, Arenas, Munduate, & Hoel, 2015) are good practices that could foster respect for antidiscrimination rules (Dovidio, Abad-Merino, & Tabernero, 2017). Supervisors are key actors throughout this entire process and crucial to ensuring that the people who feel affected perceive a climate of respect and trust when the rules of mutual respect fail .
Social norms can also be effective when dealing with workers whose prejudices and negative stereotypes have become subtle so that they do not 'lose face' with the people around them. That is, sometimes subtle discriminatory behaviour can be carried out by people who are not aware of the discriminatory facet of their acts, while other times people mask their real attitudes towards a given group behind unimportant behaviours. Social and organizational norms can put a stop to these behaviours (Tajfel, 1984).
We have been able to underscore how easily discriminatory messages are insinuated into daily language. However, even their absence can be a symptom of prejudices and negative stereotypes, as in the case of the lack of questions about LG people's everyday lives. So where is the boundary between language that is and is not discriminatory? This question links up with the reflection on the rootedness of heteronormativity in society and workplaces (Losert, 2008;Warner, 1991). This concept refers to tacit norms which mean that one does not question the sexual orientation of a new acquaintance, colleague, etc., and that heterosexuality is considered 'normal'. These norms determine some people's surprised reactions when they discover that a colleague is LG and make them question another person's sexual orientation only when they do not align with the image of femininity or masculinity that is predominantly accepted in society. In a context of heteronormativity, questions about the personal life of an LG colleague are avoided in order to prevent 'embarrassing' situations. Likewise, discriminatory language towards these groups becomes normalized, especially in job sectors which are traditionally highly masculinized (e.g., banking, finance, construction, sport, etc.). It is essential to reflect on this phenomenon in order to understand behaviours that are apparently inoffensive but are actually harming LG people's well-being (Pérez, Correa, & Castañeda, 2013).
It should be borne in mind that uncivil discriminatory language affects not only the victims (e.g., the intention to leave the job increases, the job satisfaction and the well-being decrease, etc.) and the organization (e.g., lower productivity, higher rotation, etc.) but also the colleagues who witness these actions. This type of language, just like all uncivil behaviours, can be contagious inasmuch as they are capable of unleashing spirals of violence. These processes can worsen the working climate with consequences for both people and organizations (Jones, Arenas, Nittrouer, Alonso, & Lindsey, 2017;Pearson et al., 2001).
It is necessary to highlight several limitations of this study, including the lack of data on bisexual people. Furthermore, the mean age of the participants is low, so this study may not reflect the reality of older LG persons. Future studies should bear these limitations in mind.
Future studies should also explore what heterosexual workers think of the discriminatory charge of certain kinds of language. Are they aware of the damaging potential of certain words? Do they know that certain expressions are the result of implicit prejudices and stereotypes? Furthermore, it is important to study the effectiveness of training systems and the interventions carried out. Is it possible to modify implicit stereotypes and prejudices in the short or middle term?
The process of building inclusive organizations is ongoing and takes place on a day-to-day basis . Engaging in periodic training involving supervisors, human resources managers, executives and perhaps workers' representatives to make prejudices, stereotypes and the social norms and values transmitted through language more explicit somehow guarantees that the key actors in the organization are socialized with such deviant processes, which may jeopardize the attempts to build and maintain an inclusive workplace.
El rigor del análisis de datos ha sido garantizado a través de un sistemático trabajo de equipo. Así pues, el listado de códigos a priori ha sido desarrollado por separado por cada uno de los/as autores/as. Los resultados de este trabajo individual han sido comparados antes de empezar el análisis de datos. Si bien en la mayoría de los casos los/as autores/as coincidían en los códigos identificados, siendo necesario sólo un ajuste acerca de las etiquetas atribuidas. En los casos en los que se ha presentado algún desacuerdo han sido resueltos a través del debate. Tras identificar el primer listado de categorías, la primera y la segunda autoras han analizado las entrevistas de forma individual, modificando dicho listado. A lo largo del análisis de datos, los/as autores/as han tenido reuniones semanales para comentar los resultados obtenidos y para solucionar los casos de desacuerdo a través del debate.