An ethics for linguistics? What, why, and how?

1 Why ethics, why now?

Academics, teachers in higher education and researchers in general, tend to view ourselves as having entered this profession out of love for our subject matter (in our case, languages and linguistics), and a passion for knowledge and for sharing that knowledge with others. Yet, academia has, a little too often in the past few years, been in the news for the wrong reasons. These reasons have included data fabrication, sham papers and peer-review fraud, issues of academic freedom and freedom of speech, transgressive behaviour and cancel culture (at the individual level) as well as academic boycotts (at the collective one), but also concerns for diversity and inclusion that have generated, among others, calls to “decolonize the curriculum” from different quarters. These issues affect not only individual career trajectories but also entire fields and their subject matters, and the perception of those fields by the public at large. Moreover, they have not been limited to specific fields or countries; rather, ethics concerns are shared by and affect academics across the globe.^[1]

The response has, in many cases, involved some kind of sanction or expanded guidelines, spelling out (once more) the standards to which the academic enterprise should be held. To some, this increased amount of regulatory oversight might feel like an ethics ‘on overdrive’. Yet, this response is justified, once it is realized that what is at stake is not only an individual researcher’s ability to continue carrying out research – whether publicly or privately funded – but also how this research is viewed by society, in terms of its reliability and usefulness, factors which in turn feed back into the public image of academia and the extent to which future generations are attracted to it. The debate over ethics is, in other words, nothing less than an existential one for the future of academia, giving us all sufficient reason to be at least informed about (if not actively involved in) it. One can hardly continue to carry out research in the 21st century without considering the impact of that research on others, on society at large, on the environment, but also on oneself. While within a single journal issue it is impossible to go into all of these aspects in depth, D’Arcy and Bender’s broad definition of ethics as “the ethical considerations of producing research, knowledge, dialogue, process, curriculum, and technology that is more beneficial than harmful to all those affected by the work (scientists, participants, users)” (2023: 63) seems like a good starting point and is also the one we will be adopting here.

Just as the importance of ethics itself cannot be over-emphasized, the timing is also important. We live in a world where threats to democracy, wars, and widening inequalities abide; but also post-truth, fake news, the effects of climate change, and loss of trust in research are daily occurrences. If some of these problems are old, others are new and created by the technological advances of the past couple of hundred years. Our world is different from the post-WWII world in response to whose needs the ethics principles underpinning modern science were first articulated. In the current socio-political climate, rights conquered over the past century and until recently considered universal (e.g., rights to national sovereignty, women’s rights) are challenged and relativized. By redefining these rights as political, they become contestable and removable rather than universal. Where do we, as scientists, stand on these matters? This special issue stems from the sorely felt need for researchers, linguists in particular, to leave their comfort zone and deal with the consequences, value, and impact of their research on life on earth now and in the longer term. This is no longer about the future of academia and the impact of our work alone (goals in which science communication plays a crucial role too). It is about the ethical responsibility of scientists to the people in their communities. As this ethical and moral responsibility is becoming clearer, so are some ways of responding to it. More than cataloguing the problems, this special issue aims to provide new ideas about how some of these problems may be addressed in order to help move this discussion forward.

2 Why linguistics?

The title of this special issue, “Ethics in Linguistics”, is taken from a 5-day workshop organized at the Lorentz Center in Leiden, the Netherlands, in May 2022, where preliminary versions of some of the papers featured here were first presented.^[2] This title naturally brings up a further question: is there anything specifically linguistic about ethics, and if so, what? On the one hand, if ethics and ethical behaviour are a shared concern of academics across the board, then the same standards ought to apply to all and it ought to be enough to familiarize novices (and ourselves) with these standards in order to carry out our work with integrity. On the other hand, even among linguists, it is not uncommon to hear that ethics is only relevant to some linguistic subfields – notably, those engaged in research with “human subjects” – while others can continue to do things as they always have or with minimal adjustments (e.g., those imposed on them externally by Institutional Review Boards or IRBs; see below). Even among those working with “human subjects”, ethics is sometimes treated as incidental inasmuch as they focus on language structure, studied separately from social considerations. Ethical considerations (at least with respect to participants) can then translate into simply ensuring that everyone’s data is given equal weight and taken into account during the analysis.

However, as Dobrin (2025) puts it, “[b]ecause using language is part of virtually everything people do, there is hardly a human activity in which there would be nothing of interest for linguists to study”. This omni-presence of language means that studying it has consequences that easily extend beyond those for people who are currently alive to take in the impact of technology on the present and future of humanity as well as consequences at the societal, national, and species levels. This expanded range of consequences sets linguistics apart from other fields of inquiry that do not deal with a subject as integral to our individual and social lives as language(s). It also creates a need for solutions customized to the nature of language and linguistic inquiry that are not simply copied over from other fields. A better question to ask, therefore, is: who is impacted by our work and how? Is it individuals (be that ourselves or others)? Or rather groups of people, entire “speech communities” (Bloomfield 1933; problematic as the term might be; see Eckert 2014), including our workplaces? Or is it rather less tangible aspects, such as perceptions (including stereotypes) and claims (to territories, resources, …) that our work can affect? And does it do so directly or indirectly? If the answer to any of these questions is yes – and, it seems to me, it could hardly be otherwise, for even the mere act of naming the variety one is working with implies a decision about who has the power to do that (Vrzić 2025; Chandra et al. 2025) – then all linguistic work entails at least some of the above kinds of consequences and, with them, ethical risks we should be aware of and prepared to minimize.

Importantly, the notion of minimizing risk, as D’Arcy and Bender are quick to remind us, “does not mean no risk […] regulatory oversight […] is not about removing all risk but rather about minimizing opportunities for harm as much as possible” (2023: 53). Yet, the idea of (more) ethical oversight is often met with resistance because of a widespread belief that the work of a linguist should be descriptive, not prescriptive, from which the further view that the committed linguist is a bad linguist can easily follow. This view, however, is rather short-sighted. Not only do others, such as language educators, forensic scientists, lawyers, environmentalists, and so on, routinely look to linguists for insights into how language(s) work in order to use them most effectively for their purposes, but also within modern (socio)linguistics it has been clear for some time now that linguists are inevitably also social actors who had better acknowledge and carry this responsibility openly. In this vein, Labov (1982) proposed two principles, the principle of error correction (responsibility to correct public perceptions about language varieties and their speakers when they are mistaken) and the principle of debt incurred (responsibility to use the outcomes of research to benefit the community of speakers who enabled the research in the first place). While not every linguist may see themself as a linguistic activist in the sense of De Korne (2021), that is, as “social actors […] contribut[ing] to the elusive goal of linguistic equality or justice through their language activism practices”, our work has consequences we can neither always control nor afford to ignore. The principles of error correction and debt incurred are often cited as cornerstones of how to handle the social responsibility that flows from those consequences.

But there is a second set of reasons, beyond these wide-ranging consequences, why linguists specifically ought to take an interest in ethically conducted research. This second set of reasons concern the quality of the research, and in particular research validity. One may link here to the principles of Open Science (Liu et al. 2023; Spellman et al. 2017), that is, ideas about accountability, documenting one’s methods, and replicability of findings, which are increasingly gaining ground as benchmarks of good-quality research beyond the medical and natural sciences where they originated. These benchmarks can be implemented differently in different linguistic subfields. While replicability, for instance, is a hallmark of some psycholinguistic and applied linguistic research,^[3] it cannot be among the goals of linguistic anthropological research that analyses language in the rich contexts where it occurs. Yet, even in this second case, closely documenting one’s methods provides useful insights into the dynamics that shaped those data and can be useful when it comes to comparing the findings of different projects engaging with similar questions and/or communities as well as for any future parties wishing to make sense of a project and its history in context, including dealing with so-called ‘legacy data’, especially materials that only become available once the researcher has died.^[4]

Αn ethics of accountability thus depends upon a notion of researcher positionality, that is, positioning oneself vis-à-vis the phenomena and communities one is studying and being self-reflective and explicit about one’s own epistemological commitments and reasons for doing this research. It is an ethical prerequisite to acknowledge how who we (the researchers) are is reflected in and constrains some aspects of our findings while enabling others. In other words, rather than making ourselves invisible – which runs the risk of creating an aura of infallibility that is incompatible with science as the continuous quest for the truth – positionality involves taking responsibility for our findings by making ourselves and our role as researchers visible and acknowledging the existence of other viewpoints, not represented in our own research. While at first sight this may sound redundant, naïve or, what is worse, self-serving, the simple act of acknowledging the existence of other frameworks or points of view, without necessarily engaging with them, should be seen – and practised – as an act of intellectual humility that emphasizes the limits of our knowledge and what we can and cannot be held accountable for.

Another way in which attention to ethics can enhance the validity of our research is by encouraging us to acknowledge the positionality not only of ourselves qua researchers but also that of our participants qua research participants. Through often subconscious processes of self-censorship, self-presentation, and the like, larger social groupings such as ‘nations’, ‘genders’, and so on inevitably have an impact on individual speaker judgements, however these are obtained. Research participants, in other words, are not positioned outside of social hierarchies themselves but rather operate from within them and may re-enact them through their judgements. Speaker judgements, in this sense, are also socially positioned and should be acknowledged as such – which, once more, brings us back to closely documenting the methods and circumstances in which they were obtained. The important question, from an ethical point of view, is not what type of theory the data are used to build, but rather that the full complexity of the data, including the positionality of their producers, be acknowledged and preserved in the research record. Chandra et al.’s (2025) statement that “[t]he speakers are, for G[enerative] L[inguistics], the ultimate authorities and the owners of their data” can thus be taken to highlight the relevance of ethical considerations (including ways in which they can be met; e.g., through detailed documenting of methods) to all kinds of linguistic research, no matter under which theoretical umbrella this is carried out. Once more, we do not all have to conceive of language in the same way in order to uphold the same overarching set of ethical standards (in fact, requiring that we all conceive of language in the same way would be unethical in itself, implying that another’s way of conceiving of language is less valid than our own). But we do have a responsibility to treat participants (and ourselves) as the complex and socially positioned individuals we truly are.

A final way in which ethical principles can enhance the validity of linguistic research is by enabling more inclusive theories and perspectives beyond those based on WEIRD (Henrich et al. 2010) and abled populations, and mainstream English ways of speaking at the expense of MIND (Minority, Indigenous, Non-standard(ized), and Dialect) varieties (Kirk 2023). At first, this might seem like an elementary lesson that any linguistics student learns with their first exposure to an unfamiliar language and which has been integral to progress in the field since its earliest beginnings in the comparative reconstruction of IE (Jones 1807), the work of American structuralists with Native American languages (Bloomfield 1933), and that of early ethnographers in the Pacific (Malinowski 1922). Yet, still today, and despite strides in documentation work, by far the most researched language is English and the majority of our theories are based on the structure of this (and a handful of other Western) language(s) and exemplified using English-language examples (on the associated dangers see, among others, Ameka and Terkourafi 2019; Evans and Levinson 2009; Kidd and Garcia 2022). Without going into the reasons for this,^[5] ethical considerations can play a role here by encouraging (or even requiring) us to break with convenience and develop more inclusive methods of participation in language research (e.g., Choy et al. 2022). This includes expanding the ‘canon’ of standard bibliographical references and granting more visibility to the work of scholars from other research traditions who may have been silenced or made invisible for a variety of reasons until now (Gibson et al. 2024). As with the voices of the participants who provide us with data for analysis, the (often emic) voices of all researchers must be given equal weight and taken into account during our theory building.

3 The origins of modern research ethics (and why we need to move beyond them)

Yet, even if all linguistic research has ethical consequences that extend beyond the here and now and apply equally to all branches of linguistics and not just those engaging in research with human subjects, as argued above, it does not necessarily follow that linguistics needs its own set of ethical principles. This need rather follows from the origins of modern research ethics outside of linguistics, which have left their imprint on how ethics is conceived of and practised across many academic fields, including linguistics, today.

These origins lie firmly within bio-medical research, specifically, the medical experimentation abuses of Nazi doctors during World War II, which led to the creation, in 1947, of the Nuremberg Code, the first attempt to deal with ethical issues in research from a legal perspective. This was succeeded in 1964 by the Declaration of Helsinki, which was formulated to meet the international need for a more specific code of ethics for bio-medical research. Parallel to these developments, events in the US, not least doctors’ abuses during the 40-year Tuskegee Syphilis Study (1932–1972; for a recent assessment, see Reverby 2009), gradually led to the establishment of institutional review boards (IRBs) at US institutions receiving federal funding for research. A landmark in this process was the 1979 Belmont Report, which established three general principles that continue to be applied to “human subjects” research, as research involving people came to be known. These principles are: (1) Beneficence: to maximize benefits for science, humanity, and research participants and to avoid or minimize risk or harm, (2) Respect: to protect the autonomy and privacy rights of participants, and (3) Justice: to ensure the fair distribution among persons and groups of the costs and benefits of research.

Yet, even the designation “human subjects” for participants in scientific research points to the way they are perceived in bio-medical research, which is not always appropriate to describe their role in other types of research. For instance, in contemporary linguistic documentation projects, participants are often referred to as “consultants”, highlighting their higher level of native language expertise compared with that of the linguists who are there to learn from them. More generally, the practical implementation of the principles of Beneficence, Respect, and Justice in bio-medicine has led to requirements which are not always in tune with the goals of other fields. Two of these, the requirement to anonymize participant contributions, and the requirement to destroy records after a period of time, can, depending on circumstances, be inappropriate or downright counter-productive for research in some humanities fields, including linguistics, where future co-operation with particular communities sometimes depends on publicly acknowledging their contributions to the research (see also Schifano et al. 2025), and records of earlier ways of speaking are necessary for long-term diachronic research. These shortcomings are not unrelated to the fact that, partly because of their origins in bio-medical research, guidelines and codes of practice emanating from these fields do not sufficiently address cross-cultural differences affecting how ethics is understood and implemented in different cultural settings.

4 From macro-ethics to micro-ethics

The influence of the bio-medical origins of modern research ethics on how ethics is conceptualized and practised in linguistics is most immediately seen in various sets of disciplinary guidelines published by professional associations and in the decisions of Institutional Review Boards (IRBs) or other institution-level committees charged with approving research involving “human subjects”. Both disciplinary guidelines and IRBs are discipline-wide or institution-wide agents that operate top-down and may, in this sense, be considered components of macro-ethics, defined as “the procedural ethics of IRB protocols based on general ethical principles, which are also incorporated in professional codes of conduct” (Kubanyiova 2008: 504, after Guillemin and Gillam 2004).^[6] Parallel to macro-ethics, researchers these days are increasingly acknowledging the need for a micro-ethics that will address, in a bottom-up fashion, “the everyday ethical dilemmas that arise from the specific roles and responsibilities that researchers and research participants adopt in specific research contexts” (Kubanyiova 2008: 504, after Guillemin and Gillam 2004). This distinction, which may be likened to that between statutory law (legal decision-making based on existing bodies of laws and regulations) and case law (legal decision-making based on precedent) legal systems, is useful because it highlights the complex nature of research ethics and the fact that – unlike national legal systems which opt for one or the other – both of these approaches are needed to adequately handle the ethical questions that arise in the conduct of our work as academic researchers.

5 This issue

The eight contributions to this special issue cover many, though not all, of the issues raised above.^[13] Giving priority to the specific challenges that can arise in linguistics research, the first three articles approach these from three rather different perspectives: that of formal linguistic research, of participant-centred linguistic research, and of linguistic research using pre-existing digital data. By highlighting the ethical dimensions of each of these, this trio of articles illustrate the point made earlier that ethics is not a niche concern of just some branches of linguistics but rather shared across the field. Moreover, by showing that ethical questions can arise not just during the planning stages, where they can be addressed macro-ethically, but throughout and even after a project has been completed, they underline the need to cultivate an ‘ethical mindset’, whereby researchers are trained to make and justify their own micro-ethical decisions in a self-reflective way.

In “Ethical concerns for theoretical research in linguistics: issues and best practices”, Chandra, D’Alessandro, and Putnam consider ethical dimensions of formal (specifically generative) linguistics and show that, despite a possible perception that ethical considerations are irrelevant to this type of work due to its focus on grammars rather than speakers, ethical concerns arise here as well. Focusing on two of these, the labelling of language varieties and speakers’ attitudes to the varieties they speak (see also Armostis 2025), the authors draw on their own work with Italo-Romance varieties, Hindi/Urdu, and Heritage Germanic in the US to unpack some of the complex decision-making linguists, sometimes all too routinely, engage in. Drawing attention to the ethical repercussions of these choices, the authors call for increased awareness and self-reflection, which apply to linguists no matter what their approach to theory.

Moving away from grammar-building as the goal, the next three authors, Atkins, Mackenzie, and Jones, introduce the term “participant-centred linguistic research” to refer to the “close engagement with participants, beyond collecting data produced ‘by’ and ‘about’ them” prioritized in approaches such as linguistic ethnography and anthropology. In “Ethical practice in participant-centred linguistic research”, they draw on their interactions with three different populations (medical professionals, members of an online community, and youth LGBTQ+ groups) to illustrate the complex relational issues that can arise in this type of work and the continued negotiation of rights and obligations between parties that these issues call for. Combining these experiences, they propose four micro-ethical principles, two concerning consent and confidentiality, and two covering roles, relationships, and power dynamics, that emphasize the dialogic and dynamic nature of these processes.

In the third article, “Ethical challenges in collecting pre-existing digital data for linguistic research”, Dalmaijer, Stommel, Pas, and Spooren tackle the thorny issue of using already available digital data for one’s research purposes, documenting several of the ethical pitfalls that can arise even when – or rather, precisely because – the researcher has not been involved in collecting the data in the first place. Drawing attention to what they call the “non-observer’s paradox” (the fact that not being able to observe the data prior to collection blinds the researcher to possible ethical considerations raised by these data), they – like the authors of the first two articles – highlight the importance of monitoring one’s research processes beyond macro-ethical compliance to ensure the overarching goal of minimizing risk. Another important dimension that this article (and, in part, the previous one) serves to highlight comprises the complexities of interdisciplinary work, where the ethics requirements of different fields may at times be at odds with each other, requiring, once more, on-the-spot, customized solutions. Dalmaijer and colleagues emphasize the role of technology in both enabling and constraining such solutions.

Staying with the theme of collaborating with stakeholders outside linguistics, the next two articles shift the focus to collaborating with communities, and how the interests, needs, and priorities of non-academic stakeholders can be taken into consideration. More than just an ethical desideratum, this is necessary to help foster the trust needed for longitudinal research in a community setting. As such, it generates responsibilities not only to participants but also to future generations of researchers in the field. In “Ethical considerations and good practices in linguistic work on endangered languages: the case of a research programme on Cypriot Arabic”, Armostis describes the ethics tensions that can arise during the various stages of documentation and revitalization of an endangered minority language and how these can be resolved through close consultation with community members, whose opinions may sometimes be given priority. Respecting the emic viewpoints of the users and bringing in experts from neighbouring fields (education, software development, etc.) are ways of practising intellectual humility which ultimately benefit the goals of the project itself.

The next article, by Schifano, Allen, Nellan, Restrepo Garcés, and Kasstan, provides a useful counterpoint on working with communities by offering a rare insight into the perspective of community members themselves: what reasons and expectations do they have for and from engaging in such projects? A first observation is that community members are not reluctant and passive participants in this type of research. Indeed, they might actively seek out such collaborations as a way of advancing their own advocacy goals. While not the focus of Schifano et al.’s contribution, this generates, of course, a different kind of ethical challenge for a linguist. To the extent that the community’s goals align with (or at least are not opposed to) those of the linguist, such “alliances” (always entered into with cognisance) are not necessarily bad. More often than not, however, communities feel that decision-making is disproportionately practised by the academics, who do not always include (a sufficient number of, or those considered ratified) community members in designing their studies. As a result, and despite progress in this regard, projects can still have an extractive flavour, which can sour the collaboration and undercut possibilities for future collaborative work. Researchers’ ways of seeking an entry into a community are central to this: internal community dynamics play out in the selection of partners, with repercussions for the engagement of community members and the quality of outcomes. To improve this situation, the authors make a number of concrete recommendations that aim to place communities and academics on an equal footing under a broader protocol regulating ethical collaborative research design and implementation.

This macro-ethical note is picked up by the next contribution, which continues on the theme of working with communities, this time from the perspective of giving back. In “Ethical budgets in (psycho-)linguistic fieldwork”, Sarvasy shifts the focus to psycholinguistic fieldwork, which is different from the kind of long-term descriptive fieldwork dealt with in the last two articles because of its short-term, impersonal, and transactional character. Psycholinguists, whether in or outside the lab, typically run experiments which call on participants to carry out predefined tasks, in the design and implementation of which community members are hardly ever consulted in advance. Under these circumstances, avoiding a sense of extractive exploitation and implementing principles of reciprocity can be hard. After introducing the TRUST Code for Equitable Research Partnerships, which regulates short-term cross-cultural research, Sarvasy takes up the thorny issue of what share of allocated budgets goes directly back to the communities that make this research possible in the first place and argues for increased regulatory oversight in this regard. Sarvasy, in effect, argues for an intervention at the macro-ethical level: since budgets are drawn up and approved at the institutional level, this is not something that should be left up to individual researchers. Instead, fair budgets should be used as a criterion for competitive project selection, in order to push for the benefits of research, including local capacity-building for running psycholinguistic research, to be more equitably distributed among those who make it possible.

Regulatory oversight is also the solution called for by the next article on “The role of environmental ethics and sustainability in research ethics for linguistics” by Eisenbeiss, Torregrossa, and Hopper. Here, the authors address a facet of linguistic research that hardly ever surfaces in discussions about ethics, namely the environmental impact of our work, reminding us that this is also an aspect of our social responsibility as linguists. Outlining a multitude of ways in which linguistic research (from data collection to processing and results dissemination) has consequences for our physical environment, they make a wealth of practical recommendations for how we may keep these consequences to a minimum, focusing on energy consumption and sustainable travel. While these issues are shared across academic fields – or rather, precisely because of that – they can only be effectively handled by the concerted efforts of institutions and professional organizations. Including environmental issues in disciplinary guidelines can thus help increase their visibility and put them on the agenda.

A shared insight of the articles discussed so far is that ethics is a dynamic and multi-faceted process requiring creative solutions to sometimes well-known problems. This raises the question of how we can best prepare students to develop the kind of an ‘ethics instinct’ that will guide them to the best decisions in this regard. In the closing contribution, “Training for ethical linguistics: a model for building “Responsible Conduct of Research” into department culture”, Dobrin takes up this question and advocates for a model of ethics training known as Responsible Conduct of Research. The main components of this model are regularly held seminars, organized at the departmental (or other appropriate local professional community) level, in which fellow linguists share ethical dilemmas encountered in their work, as a way of generating reflection about ethics in a safe environment. Emphasizing the role of emotion in generating behavioural change, Dobrin argues that this kind of learning ‘by proxy’ (through others’ experiences) can lead to the necessary paradigm shift from compliance ethics to an ethical sensitivity which should become part and parcel of belonging to our linguistic community of practice.

By drawing on the personal experiences of contributors based in different institutions and countries, the articles in this special issue provide a fairly broad overview of ethics practices and concerns in different localities (though not globally, as we had originally hoped), showing, as one student in my ethics course put it, how much more there is to this topic “beyond its stilted conception as box-ticking”. This is also why we chose to publish this content in a journal that follows the Diamond Open Access publication model, making it publicly available at no cost to authors and freely available to readers. A final word of thanks is due to all the anonymous reviewers, who, by making time in their agendas to review about a topic sometimes considered peripheral to “primary research”, testify to its importance. It is my hope, and that of the contributors, that this special issue will serve to energize ethical thinking and, very concretely, that readers will find here ideas relevant to their own work.