1. Introduction
Example sentences are key tools for linguists to demonstrate linguistic phenomena, both to substantiate claims within research and to present concrete examples to students in the classroom. Their utility in the discipline is ubiquitous, but they are not without issues. Research shows that example sentences in textbooks and journal articles systematically over-represent men and perpetuate harmful gender stereotypes, such as presenting feminine-gendered arguments as more likely to be non-subjects, less likely to be named, and more often referred to with kinship terms in relation to masculine-gendered referents (Macaulay and Brice Reference Macaulay, Brice, Bucholtz, Liang, Sutton and Hines1994, Reference Macaulay and Brice1997; Bergvall Reference Bergvall1996; Pabst et al. Reference Pabst, Cépeda, Kotek, Syrett, Donelson and McCarvel2018, published as Cépeda et al. Reference Cépeda, Kotek, Pabst and Syrett2021; Richy and Burnett Reference Richy and Burnett2019; Kotek et al. Reference Kotek, Babinski, Dockum and Geissler2020; Kotek et al. Reference Kotek, Dockum, Babinski and Geissler2021). They also typically introduce protagonists with Anglocentric proper names, especially with common defaults like John and Mary, though all-gender or gender-neutral names are increasingly common, as advocated for by organizations such as the Linguistic Society of America in their Guidelines for Inclusive Language (Linguistic Society of America 2016; but see Kotek et al. Reference Kotek, Dockum, Babinski and Geissler2021 for why this practice may not be ideal). Unsettling this practice is necessary as part of creating a more affirming discipline and increasing the presence of scholars who have historically been under-represented in the field. As educators, it is our responsibility to ensure that our classrooms are not only spaces where our students’ diverse backgrounds are respected and valued, but also ones where students see themselves reflected in course content. It is our hope that by recognizing themselves in linguistics, students can in turn recognize linguistics as a place where they can thrive and make meaningful contributions in their own right.
The field of linguistics is becoming increasingly attuned to these and other language-related injustices, and considering the role of representation in example sentences is embedded within a larger disciplinary project of interrogating how linguistics can be more affirming of the intersections of human diversity (see, for example, recent calls to action by Rickford and King Reference Rickford and King2016, Leonard Reference Leonard, McDonnell, Berez-Kroeker and Holton2018, Conrod Reference Conrod2019a, Charity Hudley Reference Hudley and Anne2020, Sanders et al. Reference Sanders, Umbal, Konnelly, Hernández and Emma Butterworth2020, Calhoun et al. Reference Calhoun, Anne, Charity, Mary, Jazmine and Brittney2021, and Mallinson Reference Mallinson2024; see also many contributions in Charity Hudley et al. Reference Hudley, Anne, Mallinson and Bucholtz2024a and Reference Hudley, Anne, Mallinson and Bucholtz2024b for further discussion). When teaching, linguists may not realize that they are relying on their own biases in coming up with examples, particularly in spontaneous classroom situations. These kinds of biased examples can be highly impactful in classroom settings, where they serve to centre normative Anglo identities that have long been the unmarked norm in linguistics. To this end, we aim to share our progress on the construction of the Diverse Names Database, a database of names from a variety of languages and cultures. In the following sections, we outline the goals for the project in relation to the above concerns, introduce our process of developing and adjusting the design of the database, and present some additional issues for consideration and ideas for using the database for teaching.
2. The Diverse Names Database
The Diverse Name Database is one of the outputs of a three-year pedagogical initiative in the Department of Linguistics at the University of Toronto. The primary goal of this initiative was to intentionally cultivate resources to create more affirming linguistics curricula, bring more equity, diversity, and inclusion (EDI) into the linguistics classroom, and address language-related biases more generally in teaching beyond linguistics. In the summer of 2019, we applied for a grant through the University of Toronto's Learning & Education Advancement Fund, totalling almost CAD$45,000, spread out over three years. The majority of the grant was earmarked to pay for the labour of two graduate students (co-authors Konnelly and Umbal) as Lead EDI Teaching Assistants, working a combined 245 hours per year during the academic year. For more details of the development of the project and an overview of some of its other pedagogical outputs, see Sanders et al. Reference Sanders, Umbal, Konnelly, Hernández and Emma Butterworth2020, Reference Sanders, Konnelly, Umbal, Hudley, Mallinson and Bucholtz2024a, and Reference Hudley, Anne, Mallinson and Bucholtz2024b.
As an intervention to avoid the common inequity of names chosen for linguistic example sentences, we developed the Diverse Names Database (DND), a database of names from 78 languages, categorized three ways by gender (all-gender, feminine-leaning, and masculine-leaning), confirmed with fluent speakers and/or experts on these languages. An excerpt from the DND is presented in Figure 1.
Excerpt from the Diverse Names Database
The initial idea for the DND emerged from the general goal of our initiative to diversify linguistic data in linguistics courses and to create pedagogical resources that address and unsettle existing biases in linguistic curricula. The development of the DND parallels informal efforts by some instructors who have shifted their own practices in their classes, and in some cases, have created their own resources of names, which are not always widely available or lack stable locations.
Our goal was to create an easily accessible spreadsheet with names for three gender groups: names that can be used for referents of any gender, ones that are feminine-leaning, and ones that are masculine-leaning. One underlying design principle was simplicity and ease of use: we wanted to create a tool that could be consulted quickly, circumventing the need for those on-the-fly decisions that often result in defaulting to John and Mary. Given that we are educators at an institution where English is the primary language of instruction, we designed the DND to have the three gender categories crossed with the 26 letters of the English alphabet, resulting in 78 total names. This was to ensure an even distribution of names across initial letters, as in some circumstances it is common to abbreviate names to a single letter. For example, in predicate logic formulas in semantics, it is common to use single letters as individual constants and predicates, so that H(a) may be a logical formula representing Amal is happy. Because of this common convention across languages, a subset of the DND or an entirely different version would be warranted in situations where the language of instruction is not English.
In finding names for the DND, we largely employed a scavenger methodology, trawling as many sources as possible and prioritizing understudied languages and a broad range of language families. We also sought as many expert consultants as possible to confirm the gendered judgments of the names. We also included phonetic transcriptions from our consultants wherever possible. The resulting names in the DND represent more than 30 language families from over 110 countries, and the DND has been one of the more successful products of our larger project, with at least a dozen linguists reporting to us that they have used names from it. The DND is available for viewing and download as an Excel spreadsheet on the public online Linguistics Equity, Diversity, and Inclusion Repository (LEDIR), located at https://ledir.ling.utoronto.ca (Sanders et al. Reference Sanders, Konnelly and Umbal2024b). As an online resource, it can also be readily updated with new names or information as needed.
3. Using the Diverse Names Database
The basic function of the DND is as a tool to be used as part of course material preparation and instruction. Whether for new exercises or updating existing resources, users can draw from the database as they see fit, balancing gender representations and language families to suit the specific parameters of their exercise. Though the DND is accessible primarily through its dedicated page on the LEDIR website, it is amenable to adaptation in other formats, such as a mobile app along the lines of the Diverse Names Generator (O'Leary et al. Reference O'Leary, Williams and Lee2023).
It is important to note that names from the DND or similar resources should not be unthinkingly inserted into examples. Simply using diverse names will not remove language-related prejudices and will not create affirming classrooms on its own. We invite linguists who tend to default to names such as John and Mary in their course materials to reflect on why this might be the case. Is it simply convention? Is it because it is what was modelled for us in our own experience as linguistics students? Is there something about such names that feels safer and more familiar? If so, why might that be the case, and how does that intersect with our own positionality in the classroom? Understanding the answers to these questions is crucial to any serious attempt to address the underlying biases and inequities that pervade the field.
We also advocate for taking that reflection as an opportunity to investigate where else in our course materials we are relying on hegemonic familiarities and what other shifts in content or presentation could be made that would affirm a wider range of experiences for our students. We recognize that it may feel like a daunting and insurmountable task to reconsider and revamp what we may have thought were tried and true materials, especially during a busy semester when energy and motivation are often low. In this case, identifying even one aspect of the course per year that would benefit from greater attention to linguistic injustice can lead to an accumulation of shifts over time and make the process feel far more manageable. This may require some additional care when adapting materials prepared by someone else. One possible solution is to use a revised version of an example or piece of material with a footnote explaining what has been replaced from the original, and, perhaps more importantly, why the change was made.
We argue that it is crucial to remain in open dialogue with students about intentions and interventions to create more affirming courses, and that students be included as part of the conversation if they so choose. At a basic level, this could simply be to disclose to students that names from the DND database are being used, as a way to introduce them not only to the database itself but also to the more general concept of disrupting conventionalized practices that are problematic. Students could also be invited to be directly involved with selecting names from the DND or even in building their own version. Our students are the next generation of linguists, and they should be collaborators in constructing more affirming classrooms. This requires they be made aware of issues in the field and how those issues can be challenged. By being transparent and explicit about our own involvement in linguistic advocacy, we problematize the notion that EDI initiatives must be unnoticeable in order to be effective or justified. This transparency also creates the opportunity to learn from student expertise. For example, a colleague at another institution who used the DND in their courses sent out a survey link to students, inviting them to submit more names that they would like to see included. Such crowdsourcing is a way to engage students in expanding the scope of the DND and to localize it to the particular interests of the students.
4. Reflections and considerations
Though the DND presents opportunities for greater inclusiveness and affirmation (both in terms of gender and cultural representation), it also raises additional issues and potential drawbacks that must be carefully weighed. For example, though we have three distinct gender categories in the DND, this does not entail that users of the database are required (or even expected) to map certain pronouns onto the use of certain names. Singular they could be used for any of the names in the database, and we strongly advocate creating materials with an eye for gender diversity, not just in terms of minimizing over-representation of masculine referents, but also of increasing gender diversity in a broader sense as well.
Further, we encourage linguists who wish to use gender as an axis for teaching certain phenomena, such as co-reference restrictions, to do so thoughtfully and with great intention. Consider the impacts of presenting students with a sentence containing a supposed mismatch between a name and gendered pronoun and claiming the sentence is ungrammatical. Many transgender communities have advocated against such a view of linguistic forms as irrevocably and inherently gendered (see Conrod Reference Conrod2019b and Reference Conrod, Hall and Barrett2020 for a nuanced consideration of the difference between grammaticality and pragmatic felicity with respect to pronouns and misgendering). Rather than simply assuming shared intuitions about gender agreement, one solution could be to include explicit contextual information for each example that specifies the pronouns of reference used by each fictional character, much like what linguists often include for other pragmatic clues necessary to interpret a given reading.
Avoiding gender essentialism is key as well, and to this end, the content of the sentence also matters. For example, research cited above (Macaulay and Brice Reference Macaulay, Brice, Bucholtz, Liang, Sutton and Hines1994, Reference Macaulay and Brice1997) shows that example sentences disproportionately display men as (i) engaged in more intellectual activities, (ii) employed and with a wider range of occupations, and (iii) more often directing sexual attention or violence to other referents (most typically women in sexual scenarios). It remains common, for example, to illustrate concepts such as transitivity with verbs that are either violent (such as hit) or sexually suggestive (such as kissed), and certain phenomena, such as donkey anaphora, are notably exemplified by violent acts, the canonical version being Geach's (Reference Geach1962) example every farmer who owns a donkey beats it (which many linguists have since reformulated to every farmer who owns a donkey feeds it, as in Crain et al. Reference Crain, Ni, Shankweiler, Conway, Braze and Schütze1996 and Rawlins Reference Rawlins, Baumer, Montero and Scanlon2006). As a follow-up to the DND, we also intend to construct a database of sentence frames and predicates, so that linguists can come up with example sentences that are attentive to these tropes and construct examples that avoid them.
As noted above, the DND was created in the context of an institution where English is the primary language of instruction, and we generally selected names that contained English or English-like phones and followed English phonotactics. This decision in particular has instigated a substantial amount of reflection on our part, and it is certainly worthy of critique with respect to the utility and implications of the DND. On the one hand, it is hard to ignore the salience of the phonotactics of the language of instruction, and names that incur extreme violations of those phonotactics run the risk of being misinterpreted or worse, being exoticized and evoking overt commentary, mispronunciations, and/or jokes that could be distressing for students to hear. One of the co-authors, for example, experienced a student interpreting a name from the DND in an unintended way. This created confusion that sidelined discussion of the linguistic point being demonstrated, requiring clarification that the name was indeed a name and that there was no typographic error. This is another compelling reason for explicitly discussing the DND with students. We also recommend that instructors use this explicitness as an opportunity to model how to engage with names that may be unfamiliar to them. While pronunciation may not ordinarily be an issue in printed examples (such as in tests and assignments), we still suggest that linguists attempt to genuinely learn the pronunciation of the names that they are using rather than simply pasting them into course materials. There are also benefits to using familiar names; as Calhoun et al. (Reference Calhoun, Anne, Charity, Mary, Jazmine and Brittney2021: e15) note: “While it is important to expose students to less familiar linguistic structures, introductory linguistics courses are also an opportunity to meet students where they are academically, intellectually, and socially.”
However, names that do not conform to English phonotactics can be an important tool for unsettling the language's institutional and global social power. In effect, there is a careful balance to be struck, and instructors need to be prepared to respond when the balance tips one way or the other: either reinforcing English as a hierarchical standard by using only names that fit English phonotactics, or inadvertently creating situations of othering communities whose names do not follow English phonotactics. Similar principles would apply in other contexts where a different language is the dominant language and/or language of instruction, although English holds a notable status as an academic lingua franca beyond individual institutions. There is no single right answer, and any attempt to incorporate greater cultural representation will necessarily require being prepared to deal with possible issues as they arise, and importantly, advocating for why it matters to get people's names right.
The issue of phonotactics must be distinguished from that of mispronouncing students’ names. Linguists must reflect on how we can be in stronger solidarity with students whose names have been deemed “hard to pronounce”. We must take the time to learn and pronounce their names without seeking shortcuts. Having their name pronounced correctly can be highly meaningful and affirming to a student, as names are at the heart of our identities and personhood. Mispronouncing names, particularly those of people of colour, devalues their cultural heritage and renders these individuals invisible through the imposition of linguistic assimilation.
In their study of racial microaggressions in K–12 classrooms, Kohli and Solórzano found that students of colour experienced substantial cultural disrespect in regard to their names, and that teachers “played an especially significant role in this type of racism” (Reference Kohli and Solórzano2012: 451). Against the backdrop of historical and continuing racism, Kohli and Solórzano argue that these incidents are racial microaggressions, subtle daily insults that, as a form of racism, support a racial and cultural hierarchy of minority inferiority. Bucholtz similarly describes this as a practice of indexical bleaching, itself a technique of deracialization. Through renaming, de-naming, and misnaming, the Anglicization or “phonological mutilation or wholesale erasure” (Reference Bucholtz, Samy Alim, Rickford and Ball2016: 276) of racialized students’ names evokes feelings of profound loss and public humiliation. Educators wishing to make use of the DND should therefore be mindful of how they themselves can take their responsibility not to misname others seriously (for suggestions on how to put this in practice, see Bucholtz Reference Bucholtz, Samy Alim, Rickford and Ball2016: 286–287). They should also be prepared for confusion when challenging the dominant sociolinguistic order of standardized English: if a name elicits remarks, address it. The issues of racism and Anglocentrism will persist beyond any intervention the DND can provide, and linguists should be attentive to how they may manifest in their classroom dynamics as part of an anti-racist teaching practice more generally.
Put simply, there is much more that linguists need to discuss about example sentences as pedagogical phenomena in linguistics, and proper names are in many respects just the tip of the iceberg. The DND is not something that can simply be sprinkled on top of linguistics pedagogy. It is insufficient on its own to resolve the numerous issues related to the use of names and example sentences, and it still runs the risk of contributing to problematic aspects of pedagogy if used without care. In short, the DND must be integrated with intention, with regard for classroom dynamics, and with a commitment to anti-racist teaching more generally. Regardless, our hope is that the DND will, at least as a point of departure, be a supportive resource for constructing more diverse, inclusive, and affirming examples in assignments and other course materials.
5. Conclusion
We offer the DND as one of many possible tools to help change the underlying structures of our teaching, as part of a larger process of deconstructing how we teach linguistics. In the classroom, linguists may not realize that they are relying on their own biases in coming up with examples, especially in spontaneous situations in the front of the classroom. However, we must recognize that these biases are a part of our scholarly and pedagogical practice and address them, because if we do not actively work to combat linguistic discrimination, we are helping to perpetuate it.
We do not suggest that tools like the DND are sufficient to increase diversity within linguistics on their own. As a discipline, linguistics and linguists have a long history of enacting injustices (Anderson et al. Reference Anderson, Bjorkman, Denis, Doner, Grant, Sanders and Taniguchi2022: Chapter 1.4). Linguistic scholarship has been and continues to be infected by racism (Charity Hudley et al. Reference Hudley, Anne, Mallinson and Bucholtz2020), audism (Higgins and Lieberman Reference Higgins and Lieberman2016, Cullinan et al. Reference Cullinan, Hutcheson, Wolfram, Hill, Lucas and McCaskill2020), transphobia (Bender et al. Reference Bender, Warner and Baković2017, Conrod Reference Conrod2017), and colonial extraction (Errington Reference Errington2008, Leonard Reference Leonard, Leonard and De Korne2017) that have imposed Western epistemologies onto language communities, especially Indigenous communities (Smith Reference Smith1999). The exertion of power has been further entrenched through disciplinary gatekeeping, whereby some linguists narrowly define what counts as “linguistics”, marginalizing or even outright dismissing research that pushes against such circumscriptions. It is this artificially narrow scope to which Charity Hudley and her colleagues famously posed the question, “why is your linguistics so small?” (Reference Hudley, Anne, Mallinson and Bucholtz2020: e312).
Many scholars in the field of linguistics are devotedly interrogating its barriers and will likely continue to do so for some time. We join our colleagues across subdisciplines who are working together on how to put theory into practice (Charity Hudley et al. Reference Hudley, Anne, Mallinson and Bucholtz2020) in moving towards an affirming and anti-racist linguistics that is deeply engaged in reconciliation (Sterzuk and Fayant Reference Sterzuk and Fayant2016, Kanatawakhon-Maracle Reference Kanatawakhon-Maracle, Anderson, Bjorkman, Denis, Doner, Grant, Sanders and Taniguchi2022), holds our professional organizations accountable to truly act in solidarity (Natives4Linguistics 2021), and makes community-based collaborations truly collaborative (Leonard and Haynes Reference Leonard and Haynes2010). This also entails being expansive about what linguistics (and language) is and could be and taking seriously the role of linguists as scientists and advocates of linguistic practices and the communities to whom those practices are most significant. It is our hope that the reflections we offer in this article represent an additional opportunity to evaluate our disciplinary practices, to recognize the diversity of tactics necessary in dismantling a small linguistics, and to begin building up from what is left.
Acknowldgements
We thank Naomi Nagy, Keren Rice, and the members of the Language Variation and Change Research Group in the Department of Linguistics at the University of Toronto for early advice and support in this project. We thank Anne Charity Hudley for initial inspiration and continued wisdom throughout the life of the initiative. We thank audiences at the 2020 and 2021 annual meetings of the Canadian Linguistic Association, the Linguistic Society of America's Racial Justice, Equity, Diversity, and Inclusion in the Linguistics Curriculum webinar series, the 31st Semantics and Linguistic Theory conference, the 13th Toronto Undergraduate Linguistics Conference, the University of Toronto Faculty of Arts & Science Teaching and Learning Community of Practice, and the University of Toronto Mississauga 2022 Workshop on Linguistic Equity and Justice. We thank editors Catherine Anderson and Meg Grant and anonymous reviewers for valuable feedback on earlier drafts of this article. Finally, we thank Marisa Brook, Maura O'Leary, and Ai Taniguchi for various discussions about the DND.
Open access
