Introduction

Discussions about freedom of speech are not new. However, in the past decade, they have become increasingly polarized. One of the manifestations of this polarization is the so-called ‘free speech crisis’. This crisis is exemplified by attempts—some successful—of student unions or other groups to prevent certain speakers from lecturing or giving talks on university campuses (‘no-platforming’Footnote 1). In some cases, students staged protests or threatened to disrupt lectures, leading to cancellation of events or rescinding of invitations to said speakers.

While instances of ‘no-platforming’ are not unique to the UK or the United States, they have attracted increasing attention owing in part to highly mediatized cases in these countries (Smith 2020; Riley 2021). Examples in the UK include the self-styled ‘controversial’ psychologist Jordan Peterson, whose offer of a visiting fellowship at Cambridge University’s Divinity School was rescinded in 2019 after students and staff complained about his conservative and anti-trans views.Footnote 2 Another is philosopher Kathleen Stock, who resigned from Sussex University after being allegedly bullied by colleagues and students for her views on transgender identity.

Regardless of the position one takes on ‘no-platforming’, these cases have an important thing in common. While the opposition to platforming is usually based on speakers’ publicly expressed views, the act of ‘no-platforming’ itself is confined to platforms or spaces of speech within specific institutions: universities. This highlights the role of the university as the institutional, situational, and spatial context of speech-acts. In other words: it is not only about what is being said, as about what and where it is being said.

This distinguishes recent discussions about free speech regulation at universities from debates about free speech on other platforms. Most platforms regulate acceptable speech: recent examples of this issue include film and television, the Parliament, or social media (e.g. Goldman 2019; Are 2020; Diaz and Hecht-Fellela 2021). In some cases, the regulation is left to platforms themselves, providing it does not violate other laws. In other cases, governments decide to intervene. UK Government’s recent Higher Education (Freedom of Speech) Act [HEFoS] is an example of the latter approach. Developed in 2020 and introduced to the Parliament in 2021, the Act was adopted in 2023 (HEFoS 2023).

Such interventions are not neutral; they raise questions not only about the ideology and effects of regulation but, equally importantly, about political ontology—that is, properties and powers of different actors, including institutions and individual, in the political realm. Isaiah Berlin distinguished negative freedoms—freedom from—from positive freedoms, freedom to (Berlin 1969 (1958)). In Berlin’s view, liberal governments should limit themselves to legislating on negative freedoms, seeking to protect the freedom of individuals to choose what they do, rather than trying to prescribe what they should do by legislating on positive freedoms. The starting question, then, is what kind of power is ‘free speech’ expected to confer—and on whom.

There have been several critical engagements with the Act and its background (Waltham-Smith 2021; Bacevic 2022), and more are likely to come.Footnote 3 These discussions map onto the broader, and longer, debate on provisions for and limits to speech at universities (e.g. Smith 2020; Lackey 2018; Nelson 2010), as well as other political and philosophical debates about freedom of expression. It is not my intention to summarize them here. Instead, the article proposes a critical reinterpretation of the most recent iteration of ‘free speech wars’, situating them in the context of universities as constitutive institutions of Western liberal and colonial modernity (e.g. Leigh 2022; Bacevic 2020; Pietsch 2013).

The article argues that the ‘free speech crisis’ has more to do with the kind of speech that should be allowed (or welcomed, or invited) at universities, and less to do with the kind of speech that should be allowed in general. This is not to deny that some forms of speech may be always harmful; however, the emphasis is on universities as institutions that can enable, contribute to, compound, or mitigate these effects. This requires paying particular attention to universities as institutions both producing and produced by speech.

There has been growing attention in scholarship over the past years to the concept of ‘platform university’ (e.g. Carrigan and Jordan 2022; Carrigan and Fatsis 2021; Robertson 2018) and the way in which universities as institutions act as bundles of services equally providing and restricting access to and for different forms of capital (Komljenovic 2021; Bacevic 2018). This article, however, pivots away from the assumption that there is anything stable about institutions of liberal modernity—such as universities—to focus on the mutual constitution of these institutions and speech-acts (Bacevic 2019, 2018). By interrogating the assumptions underpinning platforming at universities, the article engages critically with the theories of performativity of speech (e.g. Austin 1962; Butler 1997; Ahmed 2017). Drawing on Austin’s insight that speech-acts do not have universal veillance, it argues that the characteristics of speech-acts—including their performativity—depend on the existence of universities as particular political, social, and economic institutions in late twentieth and early twenty-first century Western liberal democracies. In this sense, it considers speech acts as mutually constitutive with the spaces in which they become possible (cf. Searle 2010; Bacevic 2018).

The first part of the article presents an overview of the legislation governing freedom of expression at universities in the UK, situating contemporary ‘free speech wars’ in the political and social context of post-WWII Europe, including the massification and expansion of higher education, and growing concerns around weakening job security in the academia. This allows us to consider how they depart from the assumptions concerning the relationship between universities and the state which, among other things, gave rise to concepts of academic freedom and university autonomy. This backdrop informs the argument, in the second part, that the meaning of speech-acts in the context of universities requires the extension of theories of speech-acts used to discuss ‘harmful’ speech (e.g. Butler 1997; MacKinnon 1993; Langton 1993). This provides the foundation for the discussion of performativity of ‘free speech’ at universities, and thus the effects of blurring the boundary between the concepts of free speech and academic freedom, in the third part. The fourth part engages with the pedagogical effects of speech conceptualized in this manner, and its implications for how we think about the relationship between universities as institutions of liberal modernity and violence.

Freedom of Speech and Academic Freedom

In England, freedom of speech is defined in the Human Rights Act of 1998 as the ‘freedom to hold opinions and to receive and impart information and ideas without interference by public authority and regardless of frontiers’ (HRA Article 10, para 1). The Act further stipulates that:

‘This right shall include freedom to hold opinions and to receive and impart information and ideas without interference by public authority and regardless of frontiers. This Article shall not prevent States from requiring the licensing of broadcasting, television or cinema enterprises’ (ibid); and that ‘the exercise of these freedoms, since it carries with it duties and responsibilities, may be subject to such formalities, conditions, restrictions or penalties as are prescribed by law and are necessary in a democratic society’ (HRA Art 10, para 2.).

The Human Rights Act stipulates some restrictions on freedom of expression as both possible and necessary—including ‘the interests of national security, territorial integrity or public safety, for the prevention of disorder or crime, for the protection of health or morals, for the protection of the reputation or rights of others, for preventing the disclosure of information received in confidence, or for maintaining the authority and impartiality of the judiciary.’ (HRA Art 10, para 2).

One area in which freedom of expression overlaps with other protections concerns hate speech, as a form of speech promoting or encouraging discrimination or violence against protected groups. Hate speech has received extensive attention in a report by the Law Commission on tackling hate crime (LC 2021a). Current regulation on hate crime in the UK explicitly protects freedom of expression, in the sense in which it requires speech qualifying as hate speech to explicitly include not only abusive or offensive content, but also a stated intention of harm. The Law Commission’s report further recommended ‘new protections for expression targeted at cultural practices, individual countries and their governments, and discussion of immigration, citizenship and asylum, [as well as] protection for gender critical views, and the use of language which expresses them’ (LC 2021b, para 24).

This hints at some areas of contention in terms of freedom of expression that overlap with the prominently mediatized cases of ‘no-platforming’. Both Jordan Peterson’s and Kathleen Stock’s views concern transgender people; Stock has openly described her position as ‘gender critical’ (e.g. Stock 2021). In this sense, the key conflict seems to be between limitations on the freedom of expression in order to protect the rights of others (in this case, transgender people), on the one hand, and the rights of individual speakers—more specifically, academics—to express gender-critical and related views that might offend or harm transgender people, on the other (cf. Simpson and Srinivasan 2018). Framed in this way, the right to express views that might be offensive to certain individuals and/or groups trumps the right to be protected from those views, as long as these views are not directly harmful, that is, as long as they do not qualify as hate speech. But is no-platforming a violation of freedom of speech?

Following Berlin’s (1958) distinction between negative and positive liberties, freedom of speech can be considered a negative freedom: the emphasis is on freedom from persecution/discrimination based on expressed views, not on the freedom to express those views. Legal protections for freedom of expression in the UK similarly emphasize negative rights, in the sense in which they protect against coercion or discrimination based on expressed views, rather than guarantee the opportunity to express those views. In this sense, no-platforming would not meet the conditions necessary for it to constitute a violation of freedom of speech. After all, intellectuals like Jordan Peterson or Milo Yiannopoulos hardly lack audiences or access to platforms. That is, unless we start from the assumption that there is an entitlement to or an obligation on the part of universities to platform certain speakers. This, however, is traditionally the purview of another set of rules and obligations: those governing university autonomy and academic freedom.

The concept of academic freedom is usually associated with Wilhelm Von Humboldt, and with the establishment of the University of Berlin (Karran 2009; Rüegg 2011). A compound of ‘Lehrfreiheit’ and ‘Lehrnfreiheit’—the freedom to teach and to learn—the concept of academic freedom was initially designed to delineate decisions related to the running of universities from decisions related to the running of the state. In the Prussian kingdom, the Enlightenment belief in the benevolent power of reason meant that the monarch was content to leave the university to run itself (Ash 2006). This was not only an altruistic act of self-control: science and technology were instrumental drivers of colonial expansion, and thus autonomy and independence awarded to universities were tacitly given in exchange for their contribution to the smooth running of empires.

Academic freedom, in this sense, was always a precarious balancing act between universities and political power, including that of the state (Rüegg 2011). Throughout most of modern history, the two were conceptually distinct but mutually supporting. Scientific ‘regimes of truth’ were as necessary for the constitution of Western empires as were practices of measuring and controlling populations, waging war, and predicting the weather (Shaffer and Shapin 2011 (1985); Kamola 2019). Of course, in the early modern period, ‘men of science’ were found outside as well as within universities; but with nineteenth century industrial and colonial expansion, it became necessary to formalize and expand the training of both scientific and administrative cadres to manage growing populations. With this, universities increasingly became co-constitutive with modern state-building projects.

Contemporary concepts of academic freedom and university autonomy are, to a significant degree, heirs to this history (Karran 2009; Nokkala and Bacevic 2014). While the concept of university autonomy was aimed at institutional protection, academic freedom, as a corresponding concept, was primarily aimed at individual academics (Karran and Malinson 2017). Combined, these concepts were meant to provide durable protection for the independence of universities in Europe and beyond.

A turning point in the history of academic freedom came with the signing of the Magna Charta Universitatum [MCU] in 1988. The Magna Charta Universitatum was the first international and inter-university document to outline the protections for academic freedom and university autonomy. Drafted shortly before the fall of the Berlin Wall and the reunification of Europe, it reflected the concern about the implications of political and social change for the role of universities:

‘The undersigned Rectors of European Universities, gathered in Bologna for the ninth centenary of the oldest University in Europe, four years before the definitive abolition of boundaries of the European Community; looking forward to far-reaching co-operation between all European nations and believing that people and States should become more than ever aware of the part that universities will be called upon to play in an international and increasingly international society…’ (MCU, Preamble).

It is not a particular surprise that the reframing of political boundaries between nation-states (the reunification of Germany, the collapse of the Soviet Union) and between respective higher education systems (the Bologna Process and the creation of the European Higher Education Area) generated anxiety about the direction of the relationship between universities and the state. In this context, the political purpose of the Magna Charta Universitatum was dual. In liberal regimes, such as most of those in Western Europe, it was meant to provide protection for the independence of the academic profession in the face of growing pressures from international markets (Beiter et al. 2016). In hitherto Communist regimes, such as most of Eastern Europe (the MCU was signed shortly before the fall of the Berlin Wall), it was meant to act as an international guarantee of academic freedom in the face of, or against, potential state interference. Hence its unequivocal endorsement of the independence of universities from both political and economic power:

‘The university is an autonomous institution at the heart of societies differently organized because of geography and historical heritage; it produces, examines, appraises and hands down culture by research and teaching. To meet the needs of the world around it, its research and teaching must be morally and intellectually independent of all political authority and economic power.’ (MCU 1988, Para. 6).

The intersection between political authority and economic power in the UK primarily took the form of privatization of higher education and the outsourcing of financial governance to the market, processes usually referred to under the umbrella term ‘neoliberalism’. While there has been a veritable boom of scholarship on neoliberalism in higher education in the past decades, including some critical work on the usefulness of the term (Peck 2013; Venugopal 2015; Bacevic 2019), most scholars agree that the period was characterized by weakening job security and reduced bargaining power of labour unions. In the UK, this led to what is sometimes referred to as loss of ‘tenure’, or the guarantee of employment for academics on permanent contracts (Halsey 1992).

From this perspective, it comes across as less paradoxical that the introduction of academic freedom in UK higher education legislation came at the same time as the abolishment of the guarantee of permanent employment. The Education Reform Act (ERA), the first substantial piece of education legislation introduced by Margaret Thatcher as Prime Minister, defined academic freedom as ‘the freedom of staff to question and test received wisdom and to put forward new ideas and controversial or unpopular opinions without placing themselves in jeopardy of losing their jobs or the privileges they may have’ (ERA 1988). Legally, ERA reinforced the position of statutory commissioners to ‘assist in the rewriting of university statutes to allow for the dismissal of academic staff on the basis of redundancy’. In practice, ERA shifted the balance of power away from academics and towards university management. Whereas previous grounds for dismissal of academic staff required a serious violation of code of conduct, triggering a lengthy disciplinary procedure (and often did not happen at all) – for the first time, university management were in the position to say staff were no longer needed.

Early discussions about the concept of academic freedom in the UK revolved around the sufficiency of academic freedom as protection against unfair dismissal. Association of University Teachers—one of the predecessors of UK’s University and College Union (UCU)—called the Education Reform Act a ‘very pronounced encroachment by the State’ (cited in Salter and Tapper 1994, p. 67). This concern mimicked broader developments in university—society relationships at the time. On the one hand, in the UK and the US, weakening job protections meant academic staff could reasonably fear being penalized (or made redundant) based on their publicly stated positions.

On the other hand, less democratic or openly repressive regimes often had explicit restrictions on freedom of speech, or at least a more tightly controlled link between official political ideology and teaching at university. It is in this spirit that the 1997 UNESCO Recommendation on the status of higher education teaching personnel expressed the ‘concern regarding the vulnerability of the academic community to untoward political pressures which could undermine academic freedom…considering that the right to education, teaching and research can only be fully enjoyed in an atmosphere of academic freedom and autonomy for institutions of higher education’ (UNESCO 1997, Preamble, paras 8 and 9).

This suggests that academic freedom, in the form in which it was initially introduced in the UK as well as in international treaties, was not a corollary of freedom of speech: it was a counterweight to the declining security of academic employment, what Halsey dubbed the ‘decline of donnish dominion’ (1992). In regimes where there were actual limitations on the freedom of expressions, academic freedom was used to circumvent them. Importantly, academic freedom was in this period almost always paired with university autonomy (Nokkala and Bacevic 2014; Karran 2009). It was understood that the legal protection for academics also rested on the existence of strong and stable institutions.

In this context, it is interesting to note that the text of UK’s Higher Education (Freedom of Speech) Act makes no reference to existing international recommendations and treaties concerning academic freedom, including the UNESCO Recommendation or the Magna Charta Universitatum. This is despite the fact that 46 UK universities, many of which were among the original 388 in 1988, are signatories to the Magna Charta. Similarly, the entirety of the text contains no explicit reference to university autonomy: the concept only appears in Minor and Consequential amendments to the Higher Education and Research Act (2017) (HEFOS Section11, para 4). Of course, it is possible that the drafters of the Act were unaware of the existing protections for academic freedom in domestic and international treaties; alternatively, it possible that they were keen to ignore it because of the perception of its association with rights and protections related to European institutions, including the European Court on Human Rights [ECtHR] (see e.g. Jay 2022 for an analysis of conflation between EU and EC tHR in British politics and media discourse). However, even allowing for these considerations does not fully explain the reframing of protections for academic freedom under the label of ‘free speech’.

This leads us to two further questions: what are the (assumed/intended) effects of the exercise of ‘free speech’ at universities? What kind of political actors are imagined, enabled or made possible through this kind of speech?

How to do Things with (Free) Speech

The idea of speech as performative—as being able to do things—is usually associated with the work of J.L. Austin and with the school of ordinary language philosophy. Austin developed the concept of ‘speech-acts’: utterances that do something by being stated, for instance, ‘I pronounce you man and wife’ or ‘I order you to go back’. But in addition to doing what they say, that is, their illocutionary effect (by saying ‘I order you’ I have effectively ordered you), speech-acts also have perlocutionary effects: that is, those that can succeed or fail depending on the context. The perlocutionary effect of ordering someone to go back depends on the social situation, position and relationship of the people involved: I cannot order my friends to go back, but I can order my subordinates to go back.

Judith Butler provided one of the most thoughtful and engaging twentieth-century elaborations of the concept of speech-acts. Butler’s initial engagement was with gender as performance: that is, there is no ‘essence’ to gender, but ‘femininity’ and ‘masculinity’ are constructs made through iterative acts that become read in gendered ways (Butler 1993). In this context, ‘speech-acts’ include not only spoken language but any kind of use of symbolic communication: gestures, body posture, etc. Around the same time, philosophers such as Catherine McKinnon (1993) and Rae Langton (1993) extended the concept of speech-acts to apply to representations of gender and desire in pornography, arguing that the performance of ‘sex’ in pornography not only depicts, but constructs, sexuality and gender, in particular heteronormative desire. Butler, however, took exception to this reading. Excitable Speech (Butler 2021 (1997)) is an engagement with speech-acts in a way that seeks to clarify how, exactly, speech-acts are performative.

Excitable Speech does this by extending the discussion of performativity to racism, violence, and inequality in the US more broadly. Butler centers on the following question: how can speech harm? Butler takes exception to the MacKinnon/Langton view that pornography is simultaneously speech and conduct: that is, that the mere (f)act of pornography constitutes women as subordinated to men, and as unequal participants in the act of sex—someone whose consent and participation can be assumed and implied, rather than actively sought. Butler does not deny that pornography can be demeaning, as well as lead to the subordination of women; but they argue that pornographic depiction is a different kind of ‘speech act’ than, for instance, the burning of Ku Klux Clan insignia outside a Black family’s home, an event that US Supreme Court ruled was not unlawful as it represented the exercise of the right of free speech (R.A.V. vs. St Paul, referenced in Butler 2021: 21).

The distinction, for Butler, lies in the possibility for the subject to avoid, or evade, being constituted (Althusserian ‘interpellated’) by this kind of speech. Butler argues that pornography, despite its power, isn’t a totalizing discourse; in other words, it does not constitute the totality of desire or sexuality. By contrast, racism, and in particular symbolically salient acts such as burning a cross in front of someone’s house, do not really leave room for interpretation for those who are on the receiving end (see also Titley 2020; Delgado 1993). Racism in the US, in Butler’s position, is a total discourse.

But, one might counter, isn’t the difference between pornography as speech-act and the burning of crosses as speech-acts (and thus as both, equally, protectible by freedom of speech, as well as potentially harmful qua speech) primarily that between representation and reality? It would be easy to write off the critique of the performativity of pornography as ‘only images’ (in the ironic title of MacKinnon’s book, Only Words). Conversely, if there are ‘words that wound’—as the title of the edited volume by Matsuda et al. (1993) suggests—isn’t the perception of hurt (or injury) sufficient to regulate/prohibit this form of speech? And if one representation of violence (racist symbolism) can wound, then why is it any different to say that another (misogyny in pornography) can wound too?

Where these discussions ultimately converge, on the question how words (or speech) can wound (or act, in the broader sense), has clear implications for the introduction of the Higher Education (Free Speech) Act in the UK. According to the US Supreme Court, burning crosses is not in itself violence: it is a representation of violence (R.A.V. vs. St Paul, referenced in Butler 2021: 21). For those who think that free speech needs to be protected at universities, certain kinds of views—including those that are ‘gender-critical’—are speech-acts of the same kind. While they may be offensive, they are not harmful as such. In Butler’s terms, we would say that they do not act immediately; in Austin’s, they are perlocutionary, not illocutionary. From this perspective, banning speakers who present such views constitutes a violation of the freedom of speech, possibly also of academic freedom.

On the other hand, the justification for ‘no-platforming’ rests on two different kinds of arguments. One is about sovereignty: universities and their constituent units (including departments and student societies) should have the right to decide whom to invite (and, consequently, disinvite) as a speaker. The other justification also rests on the performativity of speech, insofar as it assumes certain kinds of speech can be harmful, in the sense in which Butler argues burning a cross outside a Black family’s home is harmful, and in which pornography is harmful in MacKinnon’s and Langton’s view. In this view, no-platforming is justified if, and when, it hurts (some) students.

Where these two arguments intersect is precisely the political, social, and institutional context of universities as sites of speech acts. If universities have (or ought to have) sovereignty over the kind of speech exercised within them, what does this say about the power that institutions grant speech-acts? Simultaneously, if speech-acts are only effective (or: can only act in certain ways) within universities, what does this say about the nature of speech-acts? In the following section, I consider these questions in turn.

Platforming Economy, Political Ontology

Social institutions are not expressions of eternal, immutable essences, but—much like gender categories—they are instituted and recreated through performances. Some of these performances are more formalized and ritualistic; ceremonies like matriculation, graduation, or the change of University Statutes or policies fall within this category (Boltanski 2019; Bacevic 2018). But institutions are also reproduced (and challenged) through small acts of repetition. In the context of the university, a ‘lecture’ is probably among the key such acts. The existence of a lecture is a form of iterative act that confirms the existence of the university as a social institution. In this sense, a lecture—especially if part of the regular curriculum, rather than a one-off, extracurricular event (which students, after all, can choose not to attend)—is a constitutive (speech-) act for the university itself. Therefore, it makes sense to consider how a lecture—in other words, platforming—occurs.

There are two main routes to speaking at a university: one is through an employment relation (permanent or fixed-term), by which individuals officially become part of the relevant community (Faculty, members, or staff), as was the case with Stock. The other is by being invited as a guest speaker: this can range from a one-off (giving one talk) to a longer engagement as a ‘visiting lecturer’—as was the case with Peterson. While formal employment relations are governed by statutes of the institution, as well as subject to relevant (national, international) labour laws, visiting appointments are usually less formal and less subject to scrutiny. In both cases, however, being ‘given a platform’ is an outcome of a process of selection, featuring a set of evaluations, calculations, and decisions, and featuring the transfer—or exchange—of economic, social, symbolic, and intellectual capital (Bacevic 2018).

UK’s Equalities Act explicitly forbids discrimination based on protected characteristics, including sex, gender, ethnicity/race and ability. Many universities today have policies or procedures designed to counter or at least draw attention to implicit bias, including monitoring the numbers of ethnic minority candidates, encouraging applications from underrepresented groups, and so on. This, however, does not mean that the process is meritocratic. Hiring committees use gendered assumptions about academic careers (Rivera 2017), and the very idea of ‘career trajectory’ can reflect and reproduce assumptions about social background and life choices (Hammarfelt et al. 2020; Hammann 2019; Angermueller 2017).

Biases based on categories that are not part of ‘protected characteristics’, furthermore, are more difficult to establish, and often tacitly accepted or at least tolerated. Historical examples could include Continental philosophers being more likely to get jobs in departments of English or literature than in philosophy departments in the US (Rorty 1982; Gross 2002); the clustering of heterodox economists outside ‘orthodox’ departments (Lawson 2013; Morgan 2015); and struggles over the positioning of ‘structuralists’ in humanities in the UK in the 1980s (Morgan and Baert 2020).

A special case, in UK law, concerns what is known as ‘protected beliefs’. Originally introduced to protect beliefs that fall short of the definition of religion, this concept has been applied to argue that beliefs about the ontology of gender should fall within this category, as in the cases of Maya Forstater and Kathleen Stock (Cowan and Morris 2022). From this perspective, not being hired (or not having one’s contract renewed) due to one’s beliefs about the nature of gender constitutes discrimination.

While invitations to speak or give a guest lecture or a visiting fellowship at a university or one of its units are usually not governed by formal criteria to the same extent as permanent appointments, the process of selection also involves a mixture of formal and informal criteria. The inviting institution or organization will usually consider a series of factors, from a potential speaker’s work to their status and standing. The latter is particularly the case when invitees are not academics or scientists themselves—for instance, when they are actors, politicians or civil servants, sportspeople, or activists. In this case, the formal criterion is usually whether the person in question is recognized as important, but also whether they have something interesting or important to say to the community in question. In other words, it is not sufficient to be a recognized scholar in the field or a ‘celebrity’/public personality to be given a platform: there is a ‘matching’, or exchange, of capitals between the platform (institution) and the speaker (Bacevic 2018).

As in the case of academic appointments, the invitation to give a lecture, talk, or spend time as a visiting scholar at a university entails the exchange of different forms of capital. While the capital is not always or primarily economic (academics normally give visiting lectures for free, though ‘celebrities’ often charge speaker fees), it is both symbolic and social 1997, 1984). Visiting lectures and talks provide an opportunity for establishing contacts, both in the formal and informal parts of the visit (e.g. by asking questions at a lecture, or participating in social activities after). This opportunity applies equally to the speaker and the institution, including participants at events. However, the relationship is somewhat more complex in the case of symbolic capital. On the one hand, high-prestige speakers certainly confer symbolic capital on the institution: we can think of the case of Dalai Lama giving a talk at the Central European University or Judith Butler giving a series of lectures at the University of Cambridge. But the opportunity to host a high-profile speaker requires an institution to already possess a certain amount of symbolic capital, or prestige. In-demand speakers will usually not accept invitations to speak ‘just anywhere’.

In this sense, visiting lectures could be seen as what is in the post-Bourdieusian sociology of conventions known as ‘tests’ (Boltanski and Thévenot 2006). They represent an opportunity for actors and institutions involved to establish their ‘worth’, that is, symbolic value on the market for intellectual products. Declining the invitation to give a talk could signal to the inviting party they are not considered sufficiently prestigious or important enough; similarly, rescinding an offer to give a talk conveys the message that the (dis)invitee’s social and symbolic value is not sufficient for—or has become a risk to—the institution. The reason usually has to do with the institution’s assessment that symbolic or political cost of hosting a speaker is higher than the benefit; for instance, if the cost includes the risk of repeated student protest or, in the worst case, boycott of the institution or its components.

What happens in instances of ‘no platforming’ is a revocation of the promised, or expected, transfer of symbolic, social, and (often) financial capital. But it is also an act by which the institution confirms its autonomy, that is, the right to exercise control over the distribution of such forms of capital. In this sense, it is an assertion of institutional sovereignty consistent with the pairing of academic freedom with university autonomy: it is the institution who grants the right to exercise free speech on its grounds.

The introduction of ‘free speech’ into legislation concerning academic freedom, then, is a reconfiguration of this balance of powers. The combination of academic freedom and university autonomy meant that universities had primacy (or sovereignty) over platforming. The combination of academic freedom and free speech, however, extends the power and responsibility for academic speech beyond academic institutions—transferring, at least some of it, to the state or its subsidiaries (e.g. independent regulators). But the reconfiguration of powers also has effects on (or in) constituencies that make universities. It is to these constituent units that I turn to next.

One element has to do with the role and the power of student unions. While UK universities have a long and vibrant history of student societies, since the mid-1960s, the power of unions as formal representative bodies of (all) students at universities grew. This was not the case only in the UK: the power to represent students as ‘one voice’ was a critical factor in the success of student movements at the end of the 1960s (Klemenčič 2014). These unions, not unlike their worker counterparts, were well organized, funded, and—as the strikes of the 1960s demonstrated—increasingly aligned with a Left political agenda. This also hints at the source of struggles for political control over unions: while governments of democratic states as well as university administrations often endorse ‘student democracy’, student unions have been both recruitment grounds for mainstream politics and, with the massification of higher education, increasingly important players in university governance (e.g. Bacevic 2015, 2010).

This is particularly visible in disagreements over ‘platforming’ within UK’s two oldest, most prestigious, and arguably most politically influential, universities—Oxford and Cambridge. In addition to representative student unions (Oxford Students Union and Cambridge Student Union), both have debating student societies: the Oxford Union and the Cambridge Union. Both the Cambridge and the Oxford Union are famous for having debating chambers that closely mimic those in the British Houses of Parliament, and style themselves as prep for debating under similar circumstances. Both debating societies have hosted famous speakers which have become a source of rift with the official, representative unions of respective universities. In the case of Cambridge, it was Jordan Peterson; in the case of Oxford, ‘controversial’ speakers have included David Irving, Katie Hopkins, and Steve Bannon.

One effect of the new legislation, then, will be that it will likely reduce the power of representative student unions, at the benefit of other kinds of student societies. While there is no space here to go into a detailed analysis of student political organizing, in the short run, this will probably minimize the influence of Left ideas on campus and student opinion more generally. In the longer run, the fragmentation of student politics means that tensions and disagreements that would have otherwise been brought to the surface will remain below the sphere of manifest, thus compounding the trend of privatization of political concerns. While ostensibly fostering plurality and minimizing conflicts over platforming, the disappearance of what Chantal Mouffe (e.g. Mouffe 2013) dubbed ‘agonistic politics’ from universities will certainly diminish the civic role of higher education.

This brings us to our final question: what does the introduction of free speech into legislation concerning academic freedom do for the institutions themselves? This raises the question of the performative effects of speech for universities.

Platforming Speech, Performing Violence

On a certain view, the purpose of higher education is to teach critical thinking, which involves challenging and potentially disrupting students’ entrenched views. While this argument had historically been associated with progressive pedagogies, more recently it has been appropriated by critics who claim that contemporary students have become ‘coddled’ and ‘spoiled’ by ‘trigger warnings’ and ‘safe spaces’ (Haidt and Lukianoff 2018; Furedi 2017). In this view, trans students listening to a gender-critical scholar would not have a significantly different experience than, say, Christian students in an evolutionary biologist’s lecture. To these critics, ‘no-platforming’ is an expression of intellectual immaturity, as it constitutes a refusal to countenance discomfort that is essential to learning.

My purpose here is not to engage with this argument, as much as to note that it rests on the belief in the pedagogical purpose of discomfort. In this view, being made to feel out of place is a natural, and necessary, part of education. Of course, who is more likely to feel ‘out of place’ is not unconnected to categories of gender, ethnicity/race, ability, and class. As e.g. Nirmal Puwar (2004) has argued, being made to feel ‘out of place’ is a fundamental part of how historically white, colonial, and sexist institutions reproduce themselves (see also Ahmed 2017, 2004). At the same time, the institutional weight of this fact is different from the example of religious students who have to listen to lecturers who accept evolutionary biology as the true account of the origin of species. Evolutionary biology forms the stock of ontological suppositions of Western scientific knowledge: there is, among scientific disciplines represented at institutions of knowledge production, a clear consensus that theory of evolution presents the best available account of the development of life on Earth (Vickers 2022). This, however, is not the case with gender-critical views: if anything, the prevailing position is that gender is a social phenomenon, and thus changeable (Ainsworth 2015; Pearce et al. 2020; Mackay 2021). In other words, there is no scientific knowledge or discipline that requires an a priori acceptance of gender-critical ontology.

This brings into sharper focus the assumptions concerning the performativity of speech in the context of demands for universities to platform certain kinds of speakers. On this assumption, a statement such as ‘trans women are in fact men’ is the same as ‘World War I started with the murder of Archduke Franz Ferdinand in Sarajevo’ or ‘Dune is a bad film’: it is open to disciplinary scrutiny—either as fact or as opinion—but it does not have effects beyond these. The idea that students hear the words ‘trans women are in fact men’ in the same vein in which they hear the words ‘Dune is a bad film’ assumes that speech has the same effects no matter where, and how, it is uttered.

It is conceivable that for some students—maybe even the majority—this is true: to them, the debate about gender identities may be, in the telling phrase, purely academic. But for at least a small portion of students—perhaps one, perhaps none in a generation—the topic will not be purely academic. These students may be in the process of examining their own gender identity, or know someone who is. Maybe they are trans. Maybe they have always been uncomfortable with the gender identity they were assigned at birth, and they are now encountering others who may have experienced the same. In the power setting of the classroom, trans students may feel that their right to exist is being challenged.

This brings a different set of performative effects and normative constraints concerning speech-acts at universities. As Austin recognized, institutions do not only platform (or provide the social and political space for) speech-acts; they also enable, and legitimize them. The ‘success’ of a speech-act that platforms gender-critical views therefore depends on its intended effects. If these are ‘only words’, the effects are limited to representation: in this sense, the institution has fulfilled its obligation to present a variety of opinions or views (leaving, for the purposes of this argument, aside the fact that no institution could aspire to present all possible views on any matter). But if we accept that these words can have other, perlocutionary effects (for instance, they may hurt, or harm) the success of the speech-act depends on the presence of others who are able, willing, or—absent these affective dispositions—obliged to sustain this kind of speech.

A university that platforms anti-trans views therefore does more than just abide by the regulations guaranteeing equal rights to platform or freedom of expression. It also creates conditions for a particular performative power of a speech-act, one that equates the position of power and authority with the right to enact discursive violence. While acknowledging Butler’s argument that no relation of power is absolute, and that thus it would be possible (and likely) for trans students in this situation to disrupt, challenge, or parody the discourse they encounter, this kind of institution therefore puts in front of its trans students a choice: they can either accept being subjected to casual acts of discursive violence, or leave. To be clear, there is no indication of discrimination: students are not being treated or evaluated differently. In other words, they are not being excluded from institutions; they are being included, but the condition of their participation is that they are able to accept listening to views that challenge a group’s identity and/or bodily autonomy. The institution is effectively saying: if you cannot listen to these views, you do not belong here.

This means that platforming gender-critical views has implications that go beyond the question of whose identities, freedoms, and rights are prioritized. Prioritizing the right to a platform assumes education works by creating discomfort. This, of course, is a well-worn trope of critical pedagogy: education is (at least) as much about discipline and control as it is about freedom. Yet, the conflation of free speech and academic freedom provides an efficient way to elide their difference. By assuming the pedagogical value of discursive violence, the institution fulfils both ends of what Adorno and Horkheimer (1947) framed as key components in the dialectic of the Enlightenment: it is both rational, and it is violent.

Conclusions

Political theorist Avishai Margalit introduced the concept of ‘decent society’ to argue that a good society is one that does not subject its members to intentional humiliation, and whose members do not humiliate one another (Margalit 1998). Of course, as Judith Shklar has recognized, it might be too optimistic to argue for a total eradication of cruelty (cf. Shklar 1982); however, in Margalit’s view, institutions of a decent society have an obligation to prevent the systematic humiliation of any of its members.

Introducing legislation that guarantees the right to a platform enables the creation of an institutional context for speech-acts that could subject certain populations to ritualized public humiliation. This, for instance, was the case of Rod Liddle’s speech at Durham University’s South College in 2021 (Wolfe-Robinson, 2021). Law as speech-act, in this context, is not performative as a rule, but as a constitutive force making certain forms of life more likely and/or more visible, while making institutions less conducive for other forms of life. This norming mechanism, of course, does not operate only by platforming anti-trans speech. Other examples can include casual racism, microaggressions, and both casual and systemic misogyny. In this sense, just like ‘no-platforming’ constitutes a claim on what a university is and should be, licensing certain kinds of speech is also a form of constitutive act, one that creates (or recreates) conditions of possibility. The role of universities as institutions, in this sense, becomes to normalize casual acts of discursive violence as part of ‘education’.

An institution whose role is to enable the performance of speech that knowingly produces discursive violence against at least some members of its intellectual community is an institution whose purpose has shifted further along the punitive end of the power of the state (Foucault 1980). This, in some ways, constitutes the logical end of the Enlightenment as Adorno and Horkheimer (1947) saw it: however, it certainly runs counter to the idea of academic freedom.

This brings us back to Isaiah Berlin’s distinction between positive and negative liberties. Platforming is a different kind of act from not denying a platform. While denying someone a platform because of their beliefs or views could rightly be considered discriminatory, there is no corresponding obligation to provide a platform to these kinds of views. The concept of academic freedom was designed to prevent the discrimination, persecution, or, simply, state interference with academics’ research. Conflating it with the concept of free speech gives grounds for its transformation into a positive liberty, which can be understood as the obligation to platform certain kinds of views. In contexts where these views constitute a direct denial of experience or identity of certain groups—for instance, trans students—their platforming could serve the purpose of excluding these groups from universities. Therefore, we are faced with a paradox: while not platforming gender-critical academics is discriminatory, platforming them is also discriminatory.

How to balance these situations? While there are different principles one could apply to this problem—for instance, privileging the protection of students versus the rights of academic staff (Simpson and Srinivasan 2018)—this article argues that the key achievement of the conflation of free speech and academic freedom is that it takes these kinds of decisions outside of the purview of universities, student unions, or individual academics, and transfers it into the hands of the state. The law, in this sense, becomes a performative utterance insofar as it creates conditions of possibility for this kind of transfer. In practice, of course, it is likely that the majority of these decisions will still be made on university, faculty, or departmental levels, in part so as to avoid or pre-empt ‘scandal’ (i.e. the issue getting into social media) and thus potential state intervention. This, paradoxically, re-creates the conditions which the concept of academic freedom was explicitly designed to counter—including state interventionism and self-censorship. In this sense, detaching ‘speech’ from the context in which it is uttered, as the blanket framing of ‘free speech’ is likely to do, will further undermine institutional boundaries and protections for academic freedom.