Hurdles and horizons of linguistics for social justice

Translation and multilingualism are often associated with social justice, for translation breaks down communication barriers and multilingualism indexes inclusiv-ity. Angermeyer challenges the assumption that translation and multilingualism necessarily advance social justice by pointing out the context dependence of their contribution. Not only are translation and interpreting not always an effective remedy to linguistic inequality, translation and interpreting practices can themselves be a source of such inequality. Angermeyer posits that interpreting practices can be discriminatory when they are provided in ways that prioritize the needs of the institution over those of users who are served by it, pointing to asymmetrical interpreting modes in institutional interpreting as evidence. He also demonstrates that an act of inclusivity could itself be discriminatory — for example, multilingualism could be used punitively to enforce stereotypes by singling out speakers of certain languages as potential offenders of public order. This response paper complements and complicates Angermeyer ’ s intervention. While sharing concerns about problems that arise from certain modes of court interpreting and about the pu-nitive use of multilingualism, this paper invites consideration of wider contexts, including different factors that affect the delivery of a fair trial and the role of private actors in shaping a linguistic landscape. It also highlights some recurring con ﬂ icts and gaps in the discussion of linguistic justice

provide service to multilingual users, they tend to adopt different interpreting modes depending on who the addressee is.For instance, in courtroom and police interviews, consecutive interpreting is often used when interpreting from a subordinate language (understood as a language variety that differs from that used in an institution) into the language of the institution.This mode of interpreting requires the speaker to pause for the interpreter, who would then translate out loud what has been said.By contrast, simultaneous interpreting done via chuchotage (i.e.whisper mode) is typically used when the interpreter translates the institutional language for the benefit of a participant who speaks a different language.Since the interpreter has to whisper the translation to the participant while listening to other interactions at the same time, accuracy tends to be lower, and errors do not get challenged because the interpretation is only heard by one participant who has limited competence in the institutional language.The lack of a record of the interpretation makes it impossible to determine whether the rendition falls short of the required standard (Ng 2022).
It seems therefore self-evident that speakers of subordinate languages are disadvantaged because, as speakers, their narrative is fragmented by pauses, and, as audience, the information they receive through simultaneous interpreting is more likely to contain omissions than if consecutive interpreting is used.Furthermore, as Angermeyer points out, subordinate language speakers are not a ratified participant of the proceedings when they receive chuchotage service.When they do happen to speak 'out of turn', their utterances often do not get translated.
Angermeyer demonstrates that the mode of interpreting adopted in institutional processes is indicative of power relationships among participants.This insight parallels my earlier work on legal translation (Leung 2018).Legal translation is often the result of teamwork by professional translators who are invisible to outsiders in the law-making process.I have found that even in legislative translation, institutional preference for the translation to orient towards the source language or the target language reflects power dynamics between the respective language communities (Leung 2018).In decolonizing territories, a source-oriented translation strategy (involving techniques such as borrowing, transliteration, neologisms, and literal translation; see Baaij 2018) is commonly adopted where the law is translated from the language of a departed colonial power into a vernacular.In contrast, a target-oriented approach (focusing on comprehensibility and domestication; see Venuti 1994) is associated with raising consciousness about linguistic equality and the need to improve accessibility to public information.Although translation approach does not map neatly onto power relations, the latter always informs the former, even where both language versions enjoy the same legal status.Existing theories of legal translation tend to be concerned with linguistic meaning, textual functions, and legal effect (Vermeer 1996;Šarčević 1997;Reiß & Vermeer 2014) and ignore how translation practices may be sensitive to not only the legal status of the texts but also power relationships and social structures.
Returning to the case of institutional interpreting, granted that modes of interpreting are asymmetrical, whether this amounts to injustice requires further contextual analysis in a particular institutional setting.Angermeyer notes that more egalitarian modes of interaction include 'mutual accommodation, second language acquisition, translanguaging, or the use of a contact language' (p.845).Except in the situation where one or two subordinate languages are the focus of remedial efforts due to historical injustices, these are not solutions that public institutions can adopt while dealing with a linguistically diverse population.Leaving it to judges, police officers, immigration officers, and doctors to translanguage with subordinate language speakers is unlikely to better safeguard their rights than engaging interpreters.It is difficult to ensure that jurors, who do not have speaking rights in court, would have sufficient understanding of what is going on, if other court participants could freely draw from their multilingual repertoires.Indeed, Angermeyer's recommendations for 'more just translation in institutional settings' are more modest, including 'respecting speakers of subordinated languages as interlocutors and by making sincere efforts to check their comprehension' (p.854).
Courtroom discourse is hierarchically structured among its legal and lay participants (Heffer, Rock, & Conley 2013), regardless of whether interpreting is involved.This means that even lay participants who speak the institutional language do not have the ability to speak at their will, or freely check their understanding while the attorneys are speaking in legalese with the judge.While this may itself be a problem, it is not a problem that interpreting solves or creates.The goal of interpreting, though hardly ever fully achieved, is putting participants with limited proficiency in the institutional language on equal footing with those who are conversant with it (Ng 2022).
Designing linguistic accommodation for multilingual speakers requires systems thinking, which includes crossing disciplinary boundaries and considering linguistic and non-linguistic factors.Systems thinking entails the evaluation of trade-offs with every possible solution, because even social goods can conflict with one another.In the case of the right to silence, asking the suspect to restate their information in their own words to ensure their understanding could prevent the suspect from waiving their rights unknowingly.The injustice that can be averted is well worth the extra effort involved in a comprehension check.In a court of law, where proceedings can go on for hours and days, things are more complicated, and is the focus of my discussion here.
Given the problems with chuchotage, one obvious solution is to use consecutive interpreting throughout a trial.The Court of Final Appeal in Hong Kong contemplated this solution but did not endorse it, because consecutive interpreting would considerably lengthen court proceedings (HKSAR v. Chan Hon Wing 2021).Consecutive interpreting was recommended by the Supreme Court of New Zealand following Abdula, 1 but is not consistently practiced (see Ng 2022).The adoption of consecutive interpreting may be more feasible in some jurisdictions and some courts than others.For courts with a large backlog of cases, such as immigration courts in the US, more participatory discourse structures may mean further delays of justice.Delays do not only cost the court's time, which is funded by taxpayers, but could also lead to financial hardship, extended loss of freedom, and psychological stress for individuals involved.In other words, we have to ask whether a solution that addresses one source of injustice creates another set of injustices.The provision of interpreting services to those who cannot understand or speak the language used in court safeguards the right to fair trial, as enshrined in Article 14 of the International Covenant on Civil and Political Rights (ICCPR). 2 The same article also stipulates that defendants should be 'tried without undue delay' (Article 14(3)(c)).In short, both linguistic accommodation and timely justice are elements of a fair trial.
An alternative to consecutive interpreting is the adoption of electronic simultaneous interpretation, similar to conference interpreting.Ideally, a team of interpreters would work in a booth and take turns to interpret through headphones (Ng 2022).This elegant solution, which requires infrastructural investments, is understandably costly.However, the popularity of virtual courtrooms (Rossner & Tait 2021) may open less costly technological solutions to the challenge.It is not difficult to imagine that audio input from each participant could be recorded and accessed separately in a virtual courtroom.Virtual courtrooms of course bring about a different set of challenges, and again, this is where systems thinking becomes necessary.
While the limitations of chuchotage are undisputable, and better methods of linguistic accommodation should be explored, the idea that the source of the problem here is the prioritizing of institutional goals over the communicative needs of its participants might have overly simplified the situation.The court's primary goal is administering justice.This entails-among other things-making sure that trials are conducted fairly, interpreting and applying the law, deciding on pertinent legal issues, and keeping other branches of government in check.There are legitimate institutional goals, such as the timely delivery of justice, that ought to be prioritized.Such goals do not necessarily conflict with the interests of its participants, who have both communicative and non-communicative needs.In fact, in the interest of due process, it aligns with the court's institutional goal to adopt a method of linguistic accommodation that allows for recordkeeping and auditing of the quality of interpreting.
The second type of injustice Angermeyer portrays is found in the written translation of public signs.Challenging the presumption that multilingual signage signals diversity and inclusion, and that the presence of minority languages advances linguistic justice, Angermeyer shows that the inclusion of subordinate languages could be an act of discrimination when the displays aim to warn against deviant behaviour.

885
Poor quality of translation is further evidence of the lack of respect for and willingness to engage with the language communities involved.
One example is a sign from an Italian city, warning that spitting on the ground is prohibited, which was printed in Italian, Chinese, Arabic, and English.Readers of this sign may infer that Chinese, Arabic, and English speakers habitually spit on the ground.The Arabic text is written backwards, showing a lack of care and of involvement of Arabic readers in the translation and production of the sign.Angermeyer provides further examples of the punitive use of Turkish on a train and in a department store in Germany, as well as examples of ungrammatical and incomprehensible Hungarian signs in Toronto produced by machine translation.Piller has similarly shown that multilingual prohibitory signs in hotel rooms and schools in Australia single out groups of users by their language, invalidating certain group habits and exerting a form of cultural domination (Piller 2016).Labelling this phenomenon 'punitive multilingualism', Angermeyer demonstrates that multilingualism is not inherently morally superior to monolingualism.This echoes Pennycook's (2000:1) study of colonial language policies, which led to a warning against 'a simplistic liberal analysis' whereby pluralism is good and monism is bad.
The meaning of inclusion or exclusion depends critically on context, in terms of whether the act creates a positive or negative impact on the community.Another layer of contextual dependence concerns the social status of a language in the community.Although Angermeyer has labeled them subordinate languages, not all languages used outside of the institution have the same social status, and some languages are more subordinated than others.Marginalized communities are likely to experience a stronger negative impact from punitive multilingualism than others.
Even where its use may not be considered punitive, multilingualism can be a strategic means to an end that has little to do with the liberal ideology of diversity and inclusion.Multilingual welcome signs in shopping malls and duty-free shops, and multilingual greetings on a new phone or on a screen saver, are a strategy for consumer management and marketing, and are likely chosen for their commodifying value.Stereotypes are also at play here-for example, Chinese consumers are often profiled as big spenders and many international airports now have shop assistants who readily speak in Mandarin to Asian-looking shoppers.
My monograph on multilingual legal orders illustrates that even official multilingualism encoded in constitutions or legislations is often adopted for strategic purposes.Strategic pluralism in official language policy could be used to serve political and economic functions, such as containing minority nationalism, or enhancing trading opportunities (Leung 2019).I argue that such strategic pluralism promotes a kind of equality that is shallow in nature, but it may serve a positive norm-setting function.
Compared with asymmetric interpreting practices discussed in the last section, punitive multilingualism is even harder to correct, as it is ideologically laden.It is driven by forces larger than institutional communicative practices, and may be 886 Language in Society 52:5 (2023) tied to anti-immigration sentiments, pressures of globalization, nationalism, and inter-group conflicts.Where official language policy provides guidance to the public use of language, it usually focuses on ensuring that language displays are sufficiently inclusive of languages in the community, without restricting what additional languages may be included.Given that the punitive use of multilingualism is heavily context-dependent, it is difficult to imagine a top-down, regulatory solution.
What is also interesting in Angermeyer's examples is that they involved both public and private institutions.While governments can be expected to provide linguistic accommodation to its multilingual citizens, private companies do not have the same relationship with their customers.Unsurprisingly, private companies choose to provide services in selected languages based on market demand and profitability.As a result of globalization and the digital revolution, the private sector plays an increasingly important role in shaping our global linguistic landscape.The world's biggest public fora are operated by private companies, which make more decisions about the online speech environment and moderate more speech than any government in the world (Leung 2022).Social media companies' uneven efforts in content moderation have disproportionately affected some speech communities more than others.One study shows that Italian and Spanish speakers are more likely to be exposed to misinformation than English speakers (Avaaz 2020).The Facebook Papers, a set of internal documents leaked to the media, revealed that eighty-seven percent of the company's global budget for time spent on classifying misinformation is earmarked for the United States, while only thirteen percent is set aside for the rest of the world-even though North American users make up only ten percent of the social network's daily active users (Frenkel & Alba 2021).Other than misinformation, hateful speech and other verbal abuse are also more rampant in minority languages on social media platforms, which could lead to real life harm.The lack of content moderators who speak the local language meant that Facebook was not able to effectively control the spread of hate speech on its platform during the Rohingya crisis in 2017, which resulted in genocide and other atrocities (Human Rights Council 2018).
Much work in the area of language and social justice has focused attention on seeking rights that are claimed vis-a-vis the government (thus work on 'language rights' and 'linguistic human rights').When it comes to the private sector, there is an abundance of work that provides perceptive observations about language and neoliberalism, such as work that describes the commodification of language (Heller & Duchêne 2012;Holborow 2015).But the scholarship in language and social justice is relatively silent about what to do with stereotypes and biases displayed by the private sector that do not amount to legal violations (such as discrimination as legally defined), and the vulnerability of linguistic diversity and language communities to changes brought about by technological innovations.These may be manifested as punitive multilingualism in public signs, as illustrated in Angermeyer's work; in media representations and popular culture, such as Disney movies (Lippi-Green 2011); and in other acts of symbolic violence on the relevant speech communities.Such symbolic violence entails the power to construct reality and impose a definition of the social world on others (see interpretation of Bourdieu in Kramsch 2021:99).Other than stereotypes and biases, the private sector also makes market-driven decisions that contribute to shifting linguistic capitals and shaping linguistic landscapes.Industrial innovations in cross-linguistic data management technologies have led to diminished interest in instructed language learning (Gramling 2021).The technology sector is interested in solving multilingualism as a problem but not promoting multilingualism as a social good, despite their outsized cultural and political influence.In short, much of the linguistic justice literature focuses on linguistic rights, and existing efforts at theorization do not seem well equipped to deal with the role of private actors.There may not seem to be readily available tools in linguistics that allow us to bridge this gap immediately, but this is an obvious space for impactful innovations and collaborations.

L I N G U I S T I C S A N D S O C I A L J U S T I C E
Linguistics and its cognate disciplines, such as anthropology, have a long history of providing fertile ground for the perpetuation of inequalities (Charity Hudley, Mallinson, & Bucholtz 2020).The study of language as a means of classifying human diversity was applied in race 'science', pseudoscientific theories that served as justifications for racial hierarchies (Ashcroft 2001).Many concepts and methods in linguistics continue to reflect the politics of European colonial expansion and nationalism (such as 'mother tongue', Hutton 1999; and creole studies, DeGraff 2020).Exploration of the relationship between language and justice invites both social investigations and introspection within the discipline.
Formal linguistics sees itself as a scientific enterprise with a universalist ideology, focusing on innate and abstract structures that underlie all natural languages.Sociolinguistic research similarly aims to discover and describe generalizable linguistic phenomena, such as language variation and change, pragmatic principles of interaction, and so on.However, since sociolinguistic work is grounded in actual language use, sociolinguists inevitably have to contend with the social realities of language, especially how linguistic difference serves as a basis for social differentiation and inequality.There has been increasing scholarly interest in a critical sociolinguistics (Singh 1996;Heller, Pietikäinen, & Pujolar 2018) that does not only aspire to understand language in society but also to interrogate and challenge hegemony and inequities (Coupland 2016).The idea of linguistic justice also recently emerged among those who work across the disciplines of linguistics, political science, philosophy, and law (Van Parijs 2011;Mowbray 2012;Piller 2016;Baugh 2018).
A significant portion of such works focus on how national governments and law treat minority language groups.Linguistic oppression has resulted from atrocious language and educational policies imposed by the dominant class, for example, by punishing pupils who speak minority languages on school grounds (Jaspers & Rosiers 2022), or by removing indigenous children from their communities and placing 888 Language in Society 52:5 (2023) them in English-only residential schools (such as in Canada, Grant 1996;andin Australia, Read 1999=2020).Others have looked at how language may be used as a proxy to class, gender, and racial inequalities (Block 2015), and how characteristics of the linguistic variety one speaks could become a basis for linguistic discrimination (Holborow 1999;Henry 2010;Lippi-Green 2011;Weissler 2022).
Scholarly interest in the relationship between language and social justice has led to a plethora of relevant concepts such as linguicism=linguistic discrimination (e.g.Baugh 2018), linguistic oppression (e.g.Phillipson & Skutnabb-Kangas 1995), linguistic genocide (e.g.Skutnabb-Kangas 2000), linguistic imperialism (e.g.Phillipson 1992), linguistic (in)equality (e.g.Boisvert & Thiede 2020), linguistic justice (e.g.Van Parijs 2011; Mowbray 2012), and linguistic human rights (e.g.Skutnabb-Kangas & Phillipson 1995;Skutnabb-Kangas 2000).Although these conceptual tools all seem to have the potential to help advance social justice by labelling the problems and imagining potential solutions, tensions underlying them are rarely addressed.For example, essentializing tendencies lead to stereotypes and linguistic discrimination, but are required in thinking about language rights or linguistic human rights (Wee 2018).A related example is the discourse of language endangerment, which is often driven by a concern for marginalized languages and an ambition to achieve equality, but adopts concepts and theories (such as essentialism, organicism, homogeneism) that are available for constructing inequality based on difference (Heller & Duchêne 2007:11).Notable efforts that escape or confront such tensions exist.A burgeoning scholarship in applied linguistics advances a social justice agenda without reinforcing essentialism.One example is works by Ortega (2019Ortega ( , 2020) ) in the language learning context that treat bilingualism as continuous, gradient, and probabilistic.Another is linguistic citizenship, which emphasizes agency and participation and has been proposed as an alternative to top-down interventions (such as a rights-based approach) that tend to reproduce existing power structures (Stroud 2001).By contrast, it has been pointed out that the essentialist view of language is part of the lived experience of multilingualism; to that extent, it has social utility and should not be dismissed entirely (Gramling 2021).
Both of Angermeyer's case studies involve clashes between named languages as a social category, which form the basis of institutional organisation and structure how people understand the world, and a universalist conception of language that drives a sizable amount of scholarly work in language and social justice.Amidst growing resistance in linguistics against viewing language as a countable and bounded entity, the term 'multilingualism' is under immense discursive pressure, while other descriptors of linguistic practice (such as 'translanguaging') are deemed more accurate and equitable.Portraying social phenomena such as punitive multilingualism requires us to engage with individual or institutional understanding of language and language communities, which may be scientifically inaccurate but is, as Gramling suggests, part of the lived experience of language.Although an essentialist position is indefensible as a linguistic fact, anti-essentialist sentiments should not prevent us from engaging with socially meaningful concepts.
Another conflict is what linguistic justice means and whether linguistic justice is the same thing as linguistic equality.There is neither consensus on what amounts to linguistic justice, nor normative guidelines for realizing such justice (Mowbray 2012:2), but our ideas about justice are likely formed by the experience of injustice.Therefore, drawing from philosopher Nancy Fraser, Piller (2016) proposes that linguistic justice may be conceived as the overcoming of linguistic injustices.At the same time, Piller criticizes political scientists' approach to linguistic justice, which is centered around the parity for languages with a name (such as a redistributive approach that would tax those with linguistic privilege; see Van Parijs 2011).The idea of linguistic equality suffers from the same problem, 'for language simply does not fall into discrete units that one can treat equally' (Leung 2019:258).Governments in multilingual jurisdictions sometimes talk about linguistic equality, but such equality only exists in a shallow form as it is always confined to parity among a limited number of languages with official status (Leung 2019).Linguistic equality in a universalist sense is a sociolinguistic impossibility, and efforts to further linguistic equality often create inequality for those whose language is not recognised.
Efforts in correcting injustices through linguistic research require linguists to move beyond thinking about the discipline as a descriptive science and to develop strategies for social change (Lewis 2018).Many advocates of language rights and linguistic human rights already do this, as do scholars in language and law (including forensic linguistics) who have worked with or testified in front of stakeholders.In sum, within the discipline of linguistics, more work needs to be done to reconcile and debate conflicting assumptions and concepts.Looking beyond the discipline, interdisciplinary work will help place linguistic social goods in the context of non-linguistic social goods and develop a more integrated approach to social justice, and engaging with a more diverse range of actors (including not only communities and governments, but also private actors) will be critical to impacting the global linguistic market today.A few years ago, I was asked to consult on a project focused on identifying best practices for translation and interpretation for immigrant families in US public schools participating in the development of an individual educational plan (IEP) for their child who had been identified as having a disability.I was asked to consider two questions: (i) the necessary qualifications for an interpreter providing support during IEP meetings, and (ii) whether it was sufficient to have an interpreter present or whether all the legal documents that were being discussed should also be translated into the home language.After an extensive review of the literature, I wrote a report that recommended that interpreters should have expertise in special education and that all documents should be translated into the home language of the families.While I believe that this is an accurate reflection of the existing literature, I was left with a few nagging concerns.For one, the literature defined expertise in special education primarily through a medical model that treated disabilities as biological abnormalities that needed to be fixed or accommodated in order for students to become more 'normal'-an ideology rooted in colonial logics and with strong Language in Society 52:5 (2023) 893 T O WA R D A R AC I O L I N G U I S T I C P E R S P E C T I V E O N T R A N S L AT I O N This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/),which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.