1 Introduction

We observe violations of academic freedom on many occasions. Speakers at university meetings are “de-platformed”, professors are victims of campaigns on campuses, in the media, and in social networks, “trigger warnings” and various forms of silencing are issued about certain courses, some academics are censored, ostracized, not promoted, or denied access to funding or publishing because of their views. Hate speech is everywhere. Political agitation in universities is by no means new, but it has become more visible since the advent, during the twentieth century, of mass universities: since the “campus wars” of the 1960s, universities have awakened from the dream of the ivory tower. Interrupting a lecture, bullying certain professors in the name of various causes having to do with race, gender, or religion, firing a rector who is suspected of harboring politically dangerous opinions: these are clear cases, but there are also a number of borderline cases. Moreover, there are huge differences in the historical, geographical, and institutional contexts of such threats to academic freedom. How to deal with such a diversity of cases? Is it possible to apply a unique definition? The most common conception of academic freedom is that it is a subspecies of the freedom of speech, applied to academic life, just as freedom of the press is a subspecies of the freedom of expression applied to the media. I wish to argue here that this conception is misguided, and that academic freedom involves a specific kind of freedom, relative to knowledge. In this cognitivist or knowledge-based conception, what is distinctive in terms of academia is that it is a space devoted to knowledge, and that this objective involves specific rights and duties.

2 Academic Freedom as Freedom of Speech

The idea that academic freedom is but a subcase of the freedom of speech is a historically recent development. In the Middle Ages, academic freedom was a special entitlement, granted by the Church authorities and by the State to the members of universities. Later, this right was granted by the state, often under the name of “academic exemptions”, the idea being that academics have, as a collective body, certain privileges specific to their status and to the organization of their institution (Beaud 2010, 2021). In this sense, academic freedom is a freedom of speech within the university, which holds between academic peers and which regulates their capacities of learning and of teaching. From this perspective, academic freedom is a professional privilege, comparable to those of lawyers or of merchants in their own sphere. In the contemporary world, this freedom belongs to a much larger sphere: it is commonly understood as the freedom of speech of academics as citizens within a democratic society, since universities are democratic institutions open to all and subject to the same rights and duties as the rest of the public space. Academic freedom, in this sense, is but an extension of the freedom of expression. In the United States, it is a simple consequence of the First Amendment; in many other countries, it is part of the declaration of human rights.Footnote 1 Within this broad view, academics have as much right to express themselves in public settings outside universities as they have to express themselves within the universities, and should not be prevented from doing so in their role as academics. The professor is not a special case. The student and the community member who attends a university meeting open to the general public enjoy the same freedom. Their right to expression is one and the same, because all participants, academics or not, are citizens alike. The only difference between the university and a political forum, and between an academic book and a popular book, is that the former are directed toward different audiences, one specialized, the other general, and they coexist in contemporary universities. Everyone on a campus has a right to his opinions, whether true or false, and an equal right to express them in public.Footnote 2

It is not difficult to see that this conception of academic freedom as freedom of opinion and expression runs into difficulties. Freedom of expression holds for opinions, whether true or not. But universities are not just spaces where one can express any opinion one likes. They are not forums, even if they can occasionally become platforms for forum-like events. Neither are academics free to engage in any kind of research whatsoever. They are not like the members of the Academy of Lagado, ridiculed by Swift in Gulliver’s Travels, where lunatics are allowed to inquire on any topic they like, from attempts to reconvert excrement into the food from which it comes, to extracting light from cucumbers. On the classical liberal model of liberty, freedom of speech is part of the “market of ideas”, where false opinions should not be banned, since they might, after a process of filtering and discussion, eventually lead to true ones. Unfortunately, it seems abundantly clear that, if an excess of false opinions are allowed to flood the market without due discipline, truth might not emerge in the end. Therefore, academic freedom cannot be just freedom of opinion.

Twentieth century defenders of the democratic conception of academic freedom attempted to formulate a tighter connection between the expression of opinion and the search for truth. In the United States, this initiative was led by philosophers belonging to the pragmatist tradition, like John Dewey, Arthur Lovejoy, later by Sydney Hook (Stone 2015). They were Darwinists, free thinkers, and democrats. Dewey is the author of one of the most quoted definitions of academic freedom:

In discussing the questions summed up in the phrase academic freedom, it is necessary to make a distinction between the university proper and those teaching bodies, called by whatever name, whose primary business is to inculcate a fixed set of ideas and facts. The former aims to discover and communicate truth and to make its recipients better judges of truth and more effective in applying it to the affairs of life. The latter have as their aim the perpetuation of a certain way of looking at things current among a given body of persons. Their purpose is to disciple rather than to discipline. […] The problem of freedom of inquiry and instruction clearly assumes different forms in these two types of institutions. (p. 1 in Dewey 1902)

Dewey insists on the fact that academic freedom is at the service of a specific goal—truth—and that this goal implies a certain kind of activity—inquiry—which is very different from the mere expression of opinion: it is, so to say, organized belief. So, in Dewey’s view, academic freedom is a controlled and systematic search for truth, according to rules and standards which are those of science and learning. This sets universities apart from institutions whose only purpose is teaching and which are at the service of other ends, be they religious, political, or economic. However, this noble aim has always been difficult to maintain within American universities, which are often private or religious institutions, and depend upon the endowments of rich alumni. In addition, Dewey has never conceived the university as separate from the public sphere. He conceives of it as a fundamentally democratic institution, not as a cloister removed from the debates of the society as a whole. Even though Dewey distinguishes academic freedom from the exercise of free speech, he still leaves open the question of the exact nature of their relations.

The democratic conception of academic freedom is in line with the liberal conception of freedom inspired by John Stuart Mill (Mill 2008). In this conception, (i) individuals are free to associate with one another, provided that no one else’s rights are violated, and, (ii) provided that there is no imminent danger of harm, individuals are free to express or criticize opinions as they like, whatever the nature of these opinions; finally, (iii) rights should be protected only to the extent that their protection does not violate other people’s freedom. Liberal freedom, as Isaiah Berlin and many others have insisted, is mostly negative: it prevents people from hindering others’ freedom (Berlin 1969; Pettit 1997). Its positive aspect rests upon the formula which Dewey took up from the founder of pragmatism, Charles Sanders Peirce: “Do not block the way of inquiry” (see Haack 2014). However, if this formula is taken to be the epitome of academic freedom, it nevertheless leaves a number of things unclear. First, is there room for a positive account of the freedom of research? The market-place of ideas is wide, but how do those ideas arise which are supposed to be generated and discussed within universities? How are they supposed to be assessed and validated? Second, how is the expression of these ideas supposed to protect their defenders against potential harm, and is it enough that they do no harm? How do we define harm done by ideas? Is it different from the harm done by speech? Should we distinguish potential harm in the long term from immediate danger in the short term? And for what kind of groups? It is clear that the wider the academic communities, the harder it will be to define the conditions of potential harm (Simpson and Srinivasan 2018; Couto 2020; Levy 2019). The no-harm requirement is vague, but also weak and potentially adverse to the very idea of leaving open the way of inquiry: if only harmless opinions or those which do not run the risk of offense are allowed to emerge and to stay, academic learning has little chance of progress. Freedom of speech is very demanding: it is supposed to allow any view, however wrong, to reach the public space. Understood in the strictest sense, freedom to express one’s opinion in the public sphere, of which universities are just a part, leads to the view that no attempt to prevent the expression of any opinion can be legitimate. However, the liberal conception of freedom can accept the idea that there are limits to this freedom in specific circumstances. This conception is not hostile to distinctions between influence, persuasion, and direct incitement, but it has little or nothing to say on the positive side of the sphere of freedom (Pettit 2018). Where does this freedom come from?

Moreover, the simple identification of academic freedom with freedom of speech may lead very quickly to conflicts, particularly in specific cases of inviting a speaker to give a university lecture, such as the following. We have, on the one hand, a lecturer who is a well-known holocaust denier and has been invited to give a talk, which may or may not bear on this topic, and, on the other hand, students who interrupt his lecture because he is well known for his views. Both parties are violating academic freedom: the one by professing an opinion which is, in most countries, banned from the public space; the other by preventing him from expressing his opinion.Footnote 3 Thus, within the perspective that academic freedom amounts to nothing more than freedom of expression, both are entitled to act as they do: the lecturer can claim that his academic freedom is violated, and the students can claim that they have a right to prevent him from delivering a speech on this issue. We understand that there is something wrong here, and indeed, what is wrong in the first place is that to deny the existence of the holocaust is to enunciate a falsehood, and moreover a proven one. But if one identifies academic freedom with freedom of expression, this consideration is inconsequential. Nevertheless, it ought to matter deeply within a university. If, in our definition of academic freedom, we want to give a space to truth, and not just to opinion, we must make this definition more precise: what are we to exclude and to allow, and who is entitled to exclude and to allow?

3 Academic Freedom as Freedom to Know

The basic difficulty created by the assimilation of academic freedom with freedom of expression emerges in those familiar situations when, during a university meeting or an academic encounter, a controversy arises, where a group of people express their opposition to a presented view, and when the moderator of the debate, or an academic administrator, asks the respective parties to have a “balanced” view on the topic (Bilgrami 2015). This is not intended only as a suggestion for respecting the time limits of the discussion or for being polite, nor even as a reminder that everyone has a right to speak. It is a recommendation for speakers to have “moderate” positions, and to adjust their views in order to make them compatible with those of their opponents. In other words, the requirement of balance bears not on the form or style of the debate, but on the content of the opinions expressed, so that each of them be equally considered as true. Therefore, “balancing” entails that one must find some middle ground between the two opinions, or that each be considered as equally entitled to be true. This is the perfect recipe for relativism: all opinions are equal because they are, in a sense, all true from their own perspective.

The mistake in this incitement to balance is obvious: it involves a confusion between the expression of conflicting views, which indeed implies that each of them has a right to be voiced, and the truth of these views, which implies that if one is true, the opposing one cannot be true. But, if universities are institutions devoted to the search for truth and the transmission of knowledge, they cannot treat all truths as equal. In this respect, the metaphor of the “market-place of ideas” is very misleading. It presupposes that all ideas, good or bad, weak or strong, true or false, are allowed to enter the market and to compete with others, and possibly to win. Although this process may hold (within well-known limits) in the field of politics, where all citizens have a right to free expression, the same process does not hold for scientific research and teaching, where only truths which have a certain pedigree have a right of entrance, and where false or weak views are ruled out from the start. Erroneous or strange views are indeed accepted, and they can persist, but the rule of the game is that they are not allowed to persist for long, because they must be backed by arguments, and must survive critical inquiry.

Academic freedom is very different from freedom of expression or speech. It is not, to take up Robert Post’s words, based on a “democratic competence”, but on a professional competence, which involves the production, promotion, and transmission of knowledge to various audiences belonging only in part to the public sphere (Post 2012). The professional competence that is required of academics is quite unlike the professional competence of physicians, of dentists, of architects, or even of journalists, who indeed need a certain kind of knowledge in order to perform their activities, but who are not, in general, required to improve knowledge in their respective fields (unless they also engage in scientific research). One expects from academics not only to exercise knowledge within a specific type of practice, and to transfer that knowledge, but also to produce and to create new knowledge. The model of knowledge which is in place within the academic sphere is not only that of the schools, including religious or theological schools, but is scientific knowledge. Scientific knowledge is not just any kind of knowledge. It is supposed to conform to certain standards: to be based on empirical evidence or proofs, to be objective and public, testable, and possibly falsifiable. Not all academics, professors, and students engaged in scientific work need actually to possess the relevant scientific knowledge, but they need to accept and promote its standards. They are supposed to acquire certain kinds of practices, rules, and habits. Their knowledge is controlled, through the degrees that they acquire and confer. Certainly doctors, dentists, architects, or policemen must also acquire professional competences, including through academic degrees, but they are not required to produce it, but rather to put it into practice. They must agree to certain deontological rules, just like academics, but these rules do not pertain to the production and enhancement of knowledge, unlike members of academia. Through their officers and administrations, academic institutions, too, are supposed to be devoted to these ideals and to rest upon a tradition that promotes said ideals. An important corollary is that academic freedom, so understood, is not only an individual, but a collective right and competence, which has to be implemented by members of the academic institution. The specific character of academic freedom comes from the fact that faculty and students are supposed to be preserved from the intrusion of external authorities and pressures, political or private. They are the gate-keepers of their knowledge.

Academic freedom in this sense is knowledge-based, and not opinion based. In public speech and expression of opinion, anyone, if they so choose, can express their opinions, including false or weird ones, and their freedom is supposed to be preserved from interference, in the sense of negative freedom. In contrast, knowledge-based academic freedom is primarily a positive freedom: it entitles academics to launch new research and to pursue it without external influences. It also entitles them to non-interference, insofar as they aim at furthering their research. The basis of the distinction between knowledge-based academic freedom and the freedom of opinion is the very difference between knowledge and opinion. Knowledge is not belief, and not even true belief. It is, in most views, justified true belief, or true belief with reason. A knowledge-based view of academic freedom rests on two tenets: the first is that knowledge is different in kind, and not only in degree, from opinion; the second that knowledge is the foundation of this freedom. The first tenet, which I cannot develop here, but which is central to recent epistemological theories (Williamson 2000), implies that knowledge is not just a species of true belief, but a distinctive epistemic status: to know is not to have better beliefs, but to be warranted and reliable in one’s beliefs. One may object that scientific beliefs can be overturned, contradicted, and are indeed often proved false. The point is not that science gives us absolute certainty or infallibility, but that it is at least safe and objective, in the sense that it cannot easily be proved wrong. It secures objective standards and regulates the very idea of a scientific enquiry. A knowledge-based view is in direct conflict with relativistic and post-modernist views of scientific knowledge, according to which there is no such thing as scientific knowledge or truth . In such views, which have dominated a number of recent discussions of academic freedom, the standards of truth and knowledge cannot serve as a basis for academic freedom: rather, the only foundation can be the democratic freedom of opinion and of expression (see in particular Rorty 1996; Fish 2014; and, for counterarguments, Boghossian 2006). This is wrong, and rests on a distorted conception of knowledge. Knowledge is not free, but constrained by the nature of things that we discover. Contrary to what the usual metaphor says, knowledge is not a construction. The second tenet of the cognitive conception of academic freedom is that the standards of knowledge guarantee a positive form of freedom as autonomy: research ought to be free in its aims, but also must be protected from competition with other, irrelevant, goals. This implies that the search for truth ought to be disinterested and not at the service of further objectives—those economic, political, or religious objectives in particular.Footnote 4 Academic research is incompatible with the intrusion of external aims, either through the funding of programs or through submission to specific authorities; or, at the very least, it ought to be protected against such intrusions.Footnote 5

The intrusions are, in fact, not only external. They come, most often, from within the universities, through their administrations. The expansion of universities has led, during the last part of the twentieth century, to a considerable rise of their administrative bodies, and to a progressive loss of their academic body’s decision-making capacity (Ginsberg 2011). Administrators—who are less and less often academics themselves—control not only the resources and financing of today’s universities, but also the orientation of learning and teaching. In many ways, their political and economic objectives clash with those of academics, whose elbow room for free inquiry is continually reduced. The capacity to protect a knowledge-based freedom for academic research and teaching is also constrained by the financial resources granted to higher education and research institutes, which vary considerably from one country and one institution to another. Threats to this freedom can come from many angles, including from academics themselves, and its protection is all the more necessary in today’s academic world. However, we need to understand what this sort of protection entails. It begins not only from the affirmation that scientific knowledge is possible, but also that it is intrinsically valuable: it is not produced for the sake of other aims. This does not mean that what universities produce is always and everywhere genuine knowledge—this is obviously not the case, and universities do not have the monopoly on the production of genuine knowledge—but at the very least they must assume knowledge as the standard and norm for their research.

4 Objections Answered and the Relation between the Two Kinds of Freedom

The knowledge-based conception of academic freedom is, in a sense, a traditional notion. It was asserted strongly, for instance, by academics who had endured the Nazi period in Germany. Thus, Karl Jaspers said famously: “The university is the corporate realization of man’s basic determination to know. Its most immediate aim is to discover what there is to be known and what becomes of us through knowledge.” (p. 2 in Jaspers (1946) 1959).Footnote 6 It is easy to foresee the objections which one might raise against this cognitivist conception of academic freedom. A first likely objection is the following: one of its consequences seems to be that there ought to exist a protected sphere, within which academics, and academics only, are free to inquire. Isn’t this a return to the Ivory Tower? Universities are not select clubs, attended and maintained by elite scientists, forming a kind of chivalric order. The name persists in Italian, where they were called baroni, and in Germany, where Jaspers talks of a “Geistesaristokratie”. This model may survive in various contemporary academies, but it cannot be the model of contemporary universities. A university is not only a place for research, but also for teaching and learning. Academic freedom is a privilege not only of the faculty, but also of students and of all those who aspire to a university degree. But this kind of objection is easily countered by the knowledge-based conception: there is no such thing as academic research if universities are not also centers for learning, which welcome students from all origins, provided they have the talents and merits that will allow them to participate in this common task. For this reason, in particular, research centers that host permanent researchers, such as institutes of advanced studies, centers of national scientific research, and other preserved spaces, are not good models for the knowledge-based conception defended here, because they sever pure research from teaching. Such a separation neglects the fact that these two actions cross-fertilize one another, as the founders of the Humboldtian model of universities saw quite well: Lehre und Forschung.

A second objection is that the sphere of knowledge can never be fully protected against the spheres of politics and economy; not only in its conditions, but also in its consequences. In order to lead academic research, one needs resources, in the form of university jobs, funding, and work space. And, we are told, knowledge is never neutral, including in the higher spheres of mathematics and physics, and even more so in the humanities and social sciences. Not only is knowledge rarely disinterested and pure, but it is also owned by some social groups and not accessible to others. This objection is a variant of the pragmatist critique of academic freedom: even if we could secure a proper sphere for knowledge, it would compete with the social values of democracy, such as justice, equality, and solidarity. The answer to this objection is that academic research and learning is, indeed, not a safe space preserved from the influences of the public and political world; nevertheless, this fact does not mean that one cannot distinguish scientific judgments from value judgments and political judgments, and try to act, in one’s academic decision, so as to prioritize the first over the second, and respect their differences, as Max Weber urged long ago (Weber (1919) 1958).

Indeed, as Robert Simpson has remarked (Simpson 2020), many defenders of academic freedom who do not want to erase completely its boundaries with free speech—which very often represents the voice of public and social values—accept a sort of compromise between the two: on the one hand, universities are protected spaces for knowledge reserved for academics, but on the other hand, they perform a public role, in organizing events and lectures for a wider public, where academics, as well as the public at large, are free to speak as citizens. So, there can be a protected zone in which universities are constrained by the professional requirements of knowledge, and another “free” zone in which they are not, and where free speech operates in the typical way. This is consistent with the practice of organizing a number of events and lectures on campuses, to which are invited political speakers who do not have particular university degrees. Another example is the tradition of awarding doctorates honori causa to political or literary figures: they are precisely intended to illustrate the coexistence of academic ideals and cultural or political ideals. However, this compromise solution is bound to meet its limits very soon, for the requirements of a knowledge-based academic freedom imply that speakers from the “free speech zone”Footnote 7 be criticizable from a scientific point of view, and in some cases not be allowed to speak if their claims do not meet some minimal standards of rigour. This would be the case, for instance, if a famous artist came to a campus to defend the view of creationism, or if a medical doctor known to be an impostor were allowed to give a talk in front of a large academic public. In such cases, representatives of the university would be allowed to disinvite these speakers. But this would violate, prima facie, their right to free speech, in the view that it can coexist or overlap with academic freedom. Alternatively, if the “free speech zone” were allowed to overlap with the “unfree” zone, political groups who want to oppose loudly or “disinvite” academic speakers who present views to which they are hostile, they would have a right to do so. And, indeed, this interpretation and application of free speech has been prevalent in many recent cases of “de-platforming” and other oppositions to speakers, works, or symbolic figures in the name of anti-racism, anti-sexism, or anti-colonialism. If free speech and academic freedom are coextensive, these actions are legal, and can be sanctioned only insofar as they fall within the range of circumstances which, according to the liberal view of academic freedom, either directly harm the freedom of others or may constitute a danger. However, as I remarked above, the boundaries of “harmful” actions are unclear: some speeches are harmless, others are genuinely harmful. What are the limits of no harm? If a physicist is leading research which is considered by some groups to lead to the creation of nuclear weapons, why can’t these groups prevent him from doing his research? If a group believes that an historian who leads research on decolonization can do harm to others, why can’t these groups prevent him from doing his research? Where to stop? In the end, only politically correct research will be left unhindered. The result might look very much like “Marxist” science in the former Soviet Union. Potentially, any group can claim that a certain kind of view is dangerous for some social group or other. What is the difference between direct and indirect impact? How to avoid the invasion of ideology into science?

Based on the above findings, it follows that the free-speech-based conception and the knowledge-based conception of academic freedom are not compatible: one or the other must prevail. If, on the one hand, the former has priority, the university is considered a part of the public space, and academic institutions are regulated by exactly the same rules as those of communication and expression within this wider space: even a protected zone of academic freedom is not protected at all, except by the rules of freedom of speech. This conception is the dominant one today in public universities: citizens have a right to know how public money is spent on research, and if, for instance, a laboratory leads research on chemicals which might have dangerous effects on the health of the population, citizens must be allowed to interfere and to prevent this kind of research from being done. Academics have no greater rights when speaking in an academic setting than do ordinary citizens, and their special expertise weighs no more, in their right to speak and to publish, than the rights which define freedom of expression and communication. If, on the other hand, freedom of speech is regulated by a knowledge-based academic freedom, speakers at university events are not allowed to speak or to communicate unless they have received the invitation to do so by members of academia, and the latter have a right to be protected in their activities of teaching and research. Academics not only have a right to protection, but they also have a duty to oppose attempts to undermine or to oppose explicitly the kind of competence that they represent. Indeed, much of what passes for knowledge is not knowledge, and much of what is presented as argument is not proper argument. Academic freedom in such cases does not mean that an elite group of experts must act as custodians of the Temple of Knowledge, but only that a set of individuals must be able to try to live up to its standards.

Difficulties, however, will inevitably arise when such a knowledge-based conception is applied. Areas of competence—and indeed of power—often conflict within the academic sphere. Conflict of disciplines and cases of what Kant called “the conflict of faculties” (Kant (1798) 1996) are permanent: the Faculty of Law can conflict with the Faculty of Letters and Humanities, and the Faculty of Sciences can conflict with Social Sciences Faculty. Academics often experience these conflicts very vividly, for instance when they are asked by their administration to move out from their offices and to occupy smaller ones, in order to make room for colleagues from an expanding rival discipline. Moreover, the boundaries of disciplines change, and what is recognized as a legitimate academic field evolves: one century ago, there were no departments of Political Science within universities, much less fields such as Gender Studies or Post-Colonial Studies, and for long some universities have included faculties of Theology. The agenda of each of these disciplines may vary, and it is to be expected that a department of Gender Studies is more attentive to issues about equality between sexes than, say, a department of Physics. Another important source of conflict of standards is the increased competition, in the fields of humanities and social sciences, between the intellectual production of universities and the ever-growing production of what has often been called the “second market” (Boudon 1990), a grey zone where the standards of academic writing often overlap, and sometimes conflict, with those of popular writing and journalism. In many cases, academics are asked to adjust their publications to the looser criteria of popular science. Even in the ever-growing system of evaluation of academic research, an important role is devoted to the “communication of the results of research” and to the “impact factor”. With so much at stake, is it possible to resist the pressures of the market?

For all these reasons, it seems impossible to give exact limits to the exercise of academic freedom. The boundaries of its application are perpetually moving, and subject to social and political pressures. But does it follow that we cannot specify the rights and duties which go along with it, and base these on the requirement of knowledge? The main problem is that academic freedom, unlike freedom of speech, has no legal grounds. Both are subject to limits, but it seems much easier to defend the limits of the latter than those of the former. In particular, in cases of de-platforming speakers on a campus, it is hard to send invitations for formal review to committees composed of academic experts who would be able to accept or to reject such invitations. But why would this practice be any harder than the evaluation of a paper for publication or a research project? Tolerance to bad science must have its limits. Another example of pressure on academic standards comes from the fact that the funding of research is today distributed by large agencies (such as ERC in Europe or, in the US, the National Endowment for Humanities), which impose their criteria—in particular that of “interdisciplinarity”—and favour certain topics over others. Such a redrawing of disciplinary boundaries and of styles of work changes the shape of research, and thus can be a threat to academic research. But it does not follow that academics should immediately yield when they find that their discipline is threatened.

Although the circumstances in which academic freedom as freedom to know need to be defended against numerous attempts to diminish or cancel this freedom, it is doubtful that the defense could take the form of specific sanctions or legal actions. When speakers are censored, when professors are fired, statues destroyed, conferences interrupted in the names of various causes by groups with a political agenda, these symbolic actions can be resisted, and rectors or presidents are allowed to call the police. But none of these actions can be opposed in the name of knowledge. It is impossible to oppose these actions when they happen, but they can be prevented in the long run, when a culture of learning is present in an institution. Unlike the freedom of expression, knowledge is not a measurable quantity (including by means of bibliometry). It requires much more than the freedom of expression. It requires a whole set of habits, of entrenched practices acquired during a long process of learning and exchanges. The only way to resist violations of academic freedom is through the forming of communities in which a set of standards of discussion and styles of learning are implemented.

This is by no means a plea in favour of a reduction of the freedom of speech in universities. It is indeed a necessary part of a university education that students learn the rules of democratic deliberation, and that all kind of subjects be openly discussed. But the point of giving priority to knowledge is to take freedom of speech as a consequence, and not as a prerequisite to academic freedom. Those who believe that, by guaranteeing freedom of speech and democratic institutions, academic freedom will follow—they put the cart before the horse.

5 Conclusion

Academic freedom, properly understood, requires in most cases what freedom of expression prohibits: it requires control of what one advances and publishes, and readiness to accept criticism. So, both freedoms stand in opposition to one another. Academic freedom is, in fact, not free at all. Its constraints are those of the requirements of knowledge and understanding. Freedom of speech does not entail academic freedom: the fact that one has a democratic right to express one’s opinion in no way involves a claim to exercise one’s freedom to know, in the sense of launching research and teaching one’s results. But academic freedom, in the knowledge-based sense, does entail freedom of speech. It does not entail it in the “democratic” sense that academics can express their views, however wrong or weak, but rather in the sense that free speech in an academic setting must be controlled by argument, proof, and inquiry.