1 Introduction

What kind of speech ought to be protected under a free speech principle is deeply disputed in legal literature. Claims as to why we should value free speech, and which harms are sufficient to override these reasons, are myriad. In order to illuminate such questions, legally-oriented theorists at times invoke philosophy of language directly to understand how speech functions. I operate in this tradition and aim to specifically bolster Kent Greenawalt’s view that ‘assertions of fact and value’ ([1], p. 43) have special relevance for the notion of a free exchange of ideas. I aim to develop this observation by offering an account of assertion that explains its relation to truth-discovery, and therefore provide an understanding of assertion that coheres with the truth-justification for free speech.

There are different ways of understanding assertion in the linguistic literature but I ultimately make the case that the dialectic approach is the best fit for understanding assertion from a free speech perspective. This approach focuses on the language game of ‘giving and asking for reasons’ ([2], p. 102) to believe things. I go into detail as to what this involves and argue that it is a better way of understanding assertion in a free speech context as such language games lay at the core of a free speech principle’s values. The fruits of this analysis are potentially plentiful, as I argue it can provide a content-neutral identification of a particular discursive social practice that resonates with the truth-justification in a sophisticated and consistent way. I therefore put forward a model for free speech with which one might ascertain a form of archetypically relevant speech to the pursuit of truth.

The progression of my argument is as follows. Section 2 introduces the relevant detail of the dialectic account and demonstrates its provisional utility for free speech. Section 3 introduces the necessary aspects of Greenawalt’s theory with regard to the prominence of assertive speech for free speech theory, and comments on the advantages of the dialectic formulation I apply. Section 4 contextualises my theory within variances of the truth-justification in free speech broadly and shows the wide coverage of my theory. Section 5 considers some potential alternative conceptions of assertion in the linguistic literature and argues against their utility in comparison to the dialectic account. I conclude by arguing that these considerations can help free speech theorists better understand the kind of speech which warrants only very careful regulation given a commitment to the truth-justification.

2 The Dialectic Account of Assertion

This article employs analyses of “assertion” offered by speech act theory to answer with greater clarity the question: what kind of speech is communicative in a way that makes it relevant for a truth-oriented justification for free speech? I rely on accounts of assertion which emphasise it as a ‘social practice’ ([3], p. 76), and which best explain the relationship between the language game of assertion and the discursive pursuit of truth. This, I shall show, beneficially accounts for assertion as a kind of activity that we participate in in society in myriad ways across uncountable contexts, and explicates the relationship of that practice to the pursuit of truth both at an individual and collective level. I also demonstrate that this supports existing formulations of the truth-justification. I begin in this section by introducing the key points of my application of the dialectic account before advancing my position in relation to free speech theory directly.

I draw on philosophers who can be taken to support such a dialectical understanding of assertion. One primary theorist in this regard is Lionel Shapiro who provides a recent and helpful refinement of different dialectical accounts of assertion [3]. The other main theorist I draw on is Michael Rescorla [2], who first introduced the term “dialectic” account of assertion and whose commentary remains central ([3], p. 79). I also include some key insights by both Robert Brandom [4] and John MacFarlane [5, 6]. All offer uniquely useful remarks for a free speech understanding of assertion. While there are points of contention between these theorists and they all formulate their own account of assertion, I will draw the preferable unifying interpretation where relevant and avoid differences beyond the remit of free speech considerations.

Part of the challenge of employing speech act theory to better understand legal theory is that there are many competing accounts of assertion. I advocate for the dialectical norm account of assertion in the free speech context, which explains assertion in terms of the language game of ‘giving and asking for reasons’ ([2], p. 102). As identified by Shapiro the dialectic account ‘can explain…connections between assertion and truth.’ ([3], p. 76). The dialectical account of assertion’s relation to the pursuit of truth is the primary factor which makes it so useful for free speech theory. I do not expect there to be a semantic or abstract answer to the question of what is an assertion outside of this remit and context; if there is an answer to such a question it would not necessarily be of interest to the law as a particular normative endeavour. The dialectic account of assertion has significant explanatory and justificatory power in identifying the kind of speech that should be well-protected due to its sufficiently wide reaching—yet also appropriately discerning—scope that is remarkably compatible with the degree of reach needed for the truth-justification for free speech.

An important first aspect of the dialectic account is that Rescorla rejects an approach which expects there to be a formally correct way to assert something, precluding any judgement as to ‘which assertions are permissible.’ ([2], p. 99). Instead he advocates for what he coins as a dialectic account, which ‘characterize assertion solely in terms of how it alters the speaker’s normative standing, not in terms of which assertions are permissible.’ ([2], p. 99). This means that—under a dialectical model—assertion is understood as speech that alters one’s normative/social obligations in a specific way. In support of a broad account of assertion in linguistics, Rescorla advocates for ‘non-restrictive assertoric norms.’ (My emphasis) ([2], p. 99). This initial technicality and its significance will become clearer.

Rescorla explains how attention to alterations of the speaker’s normative standing offers a stronger understanding of assertion to alternatives, which in turn I claim explains well the free speech utility of assertion as a kind of speech act:

‘The most promising candidates for non-restrictive assertoric norms reflect what I will call the dialectical model of assertion, which regards assertion as essentially a move within ‘the game of giving and asking for reasons.’ The intuitive idea here is that, by asserting a proposition, I commit myself to defending the proposition when faced with challenges and counter-arguments. I can cancel the dialectical commitment by retracting my assertion, but until then the commitment stands.’ ([2], pp. 99–100).

As a result, dialectic accounts are considered a member of ‘commitment accounts’ ([3], p. 75) of assertion in line with this ‘dialectical commitment’ ([2], p. 126) undertaken by a speaker. Commitment accounts are called such simply because they explain the speech act in terms of what the speaker “commits” to in making an utterance. Such a view of assertion is best understood through the scope of language games:

‘On this view, assertion is analogous to a game, in that it intrinsically involves certain rules. Someone who exhibits no sensitivity to the rules of a game does not grasp what it is to play the game, and someone who exhibits no sensitivity to assertion’s constitutive norms does not grasp what it is to assert a proposition.’ ([2], p. 99).

This situates assertion in a way that is very helpful for the truth-justification for free speech. To play the game of assertion under this view is to put forward proposition(s) and engage in rebuttals or questions, among other responses. To put aside the technicality initially, one can intuitively understand how this game is played in many different contexts through some examples. These basic examples, which I develop further at times throughout the paper, can illustrate this core point better than abstract explanation alone.

Firstly, in writing this article, I am making a claim (that free speech theory can be helpfully informed by attention to the dialectic version of “assertion”). This has altered my normative standing in the sense I have committed to that claim. If a friendly colleague over a weekly coffee catch-up says to me “I read your article the other day, while I liked its core argument I thought you were wrong to dismiss alternative accounts as they may cast light on some aspects of speech regulation you did not explore” then I am clearly under an obligation, normatively and socially, to respond to that in some way on the basis of my prior assertion. My commitment to what I have asserted remains in play, and I can subsequently engage with that commitment well or poorly. Other forms of challenge to my claim, such as responses by subsequent articles, fall into the same broad framework. Less formal instances than writing an academic article illustrate the same idea. For instance, a twitter post questioning the safety of mRNA vaccines might be responded to by interlocutors in the comments of the post itself or by friends in person later, among other engagements—one can understand the tweet as an act of assertion because of the normative obligation generated by the tweeter to defend (or retract) the proposition in at least some way upon engagement. Lastly, to show an informal example, a group of friends might be talking about something light, such as whether Marvel movies are a tired archetype or remain good entertainment. If one friend claims either it normatively alters the social situation in that it invites rebuttals or agreement, among other responses.

This understanding of assertion crucially therefore locates it in in connection with other speech acts. As such, Rescorla writes that:

‘(A)ssertion is individuated by its role within the game of giving and asking for reasons, which I will often refer to as ‘reasoned discourse.’…assertion can be fully characterized only through its normative relations to other speech acts, such as questioning and challenging.’ ([2], p. 102).

This understanding then locates the speech act of assertion as in some way contingent on other speech acts (questioning and challenging, or partial agreement, etc.). This is appropriate, subject to some forthcoming caveats, as it helps understand assertion as a linguistic activity that, in order to be performed at its best, relies on engaging with other kinds of speech acts. In raising a commitment to engage with questions/counter-assertions (or any form of challenge or agreement) to a proposition, an ideal asserter will typically engage with these subsequent related speech acts, which form a part of the overall practice of asserting something.

Situating assertion in relation to other speech acts therefore helps understand its role in ‘reasoned discourse’. ([2], p. 102, [3], p. 75). What counts as “reasonable” within “reasoned discourse” for free speech has the potential to be ideologically fraught/loaded, but the dialectic account offers a content-neutral and speaker-neutral definition (which I expand on in Sect. 4). Understood “dialectically”, reasoned discourse can be grasped coherently and usefully as what ‘moves’ ([3], p. 81) within the game of giving and asking for reasons are permissible to still be meaningfully playing that game. In other words, it can identify what kind of social practice accurately portrays, or meaningfully engages in, assertional practice. This raises the issue as to what moves are appropriate to still be meaningfully engaging in assertional practice. There is, however, quite widespread disagreement about what moves are appropriate within different dialectical accounts in speech act theory/linguistics ([2], p. 100, [6, 7]). Many of these surround how limited/wide the range of permissible moves should be for something to meet the minimum threshold of assertion. This is a contention that Rescorla takes with Brandom calling him too ‘restrictive’ ([2], p. 100), but this is resolved by Shapiro’s interpretation of Brandom, as I shall show. Shapiro develops both to ‘suggest an application for a liberalized dialectical norm account’ ([3], p. 76) which I rely on.

One can consider what would discharge the normative duty created by asserting something in some basic ways to better understand this point about permissible moves within assertional practice. If one, to use my earlier example, claims that mRNA vaccines are ineffective or unsafe, and they are challenged, they obviously need not—personally—demonstrate the claim by conducting a study in a personal laboratory. They could discharge this duty by referring to work they had read on the matter, etc. Similarly, if one makes a claim about the economy and inflation, and is challenged, referring to the work of an economist is patently an acceptable move in the social practice/language game of assertion. One could also draw on their own reasoning or things that they have observed. More combatively, one could attack the premise of the interlocutor’s view or their legitimacy to express the view themselves. One could retract the assertion and agree with the contestation. There are therefore a variety of ways one could respond (“moves” one could make in the game), all of varying persuasiveness to an interlocutor or audience. We would understand all of these (non-comprehensively) as engagement in a dialectic assertion game.

The application I make here therefore takes the view that imposing too much formality or strictness on what moves one may reasonably make when playing the game of assertion might be thought to be, in MacFarlane’s words, ‘over-generalizing from seminar-room assertions to assertions in general’ ([5], pp. 334–335). What is meant by this is that the real world social practice of assertion is easy to misunderstand by over-formalising expectations into a debate-like scenario. This understanding of assertion would expect a much more explicit and formal back and forth on premises than may take place than in everyday discourse (and everyday life is of course the natural context for free speech concerns). Formalised debates/seminar room discussions only offer one way in which a broader understanding of assertion would understand it—under a stricter/more restrictive view ‘(d)ialectical norm accounts may illuminate specialized “games of giving and asking for reason,” but perhaps not assertion’ ([3], p. 86). Variations in the strictness of dialectical account norms are therefore important for my purposes here. I deliberately employ theorists who envision a broad (non-restrictive) understanding of assertion to account for the uncountable ways discourse occurs in a way that pertains to truth-discovery. (Though, whether we call this broader understanding “assertion” or “specialised games of giving and asking for reasons” is immaterial for the substantive utility of this theory).

While observations made by Rescorla, Brandom, MacFarlane and Shapiro have variations in this respect, it would be a needless diversion to get bogged down in the minutia of such differences. The important point is that all can be taken to provide a broad account of assertion as developed most recently by Shapiro that is not overly formal or too ‘narrow’ ([3], p. 86). Shapiro explains well in his terms how roughly assertion can be performed in the wild, outside of ‘regimented’ ([3], p. 80) formal disputes such as seminar room discussions. Shapiro identifies the following broad point about all dialectical accounts of assertion:

‘Dialectical norm accounts explain the speech act of asserting in terms of its role in the “game of giving and asking for reasons” that its proponents regard as essential to linguistic practice…This role concerns how asserters are obliged to respond to challenges, and how assertions can be used to meet such obligations. ([3], p. 79).

As noted, in an ideal instance of the assertion language game, there will be a back and forth on the assertion(s) in question, as explained by Brandom whose account Shapiro favours overall ([3], p. 75), in the following way:

‘Responding to such a challenge consists in producing further assertions whose contents are appropriately inferentially related to the original one. Each justifying consists of further assertings, which may themselves be challenged and stand in need of further justification.’ ([4], p. 642).

This makes sense from a discursive truth-justification perspective, as such a form of exchange is a very important way in which truth-discovery happens both on an individual and collective level. But what level of effective engagement in this activity is sufficient is a crucial question for my purposes herein as I argue a broad understanding of assertion is necessary from a truth-justification perspective, while still being capable of discerning a distinct social activity.

Given that an asserter has—under the dialectical account—given a commitment to respond to a challenge (regardless of the quality of that subsequent engagement), an important question is whether one can simply not respond and still be taken to be engaging meaningfully in the language game of assertion. In other words, are there circumstances in which one can—as a valid move in assertional practice—neglect to respond to any interlocutor. MacFarlane poses the following query regarding an asserter who fails to respond to any challenges:

‘Suppose someone were to say: “You’ve given some very good reasons to doubt the truth of what I asserted. I have nothing to say in answer to your objections, yet I continue to stand by my claim.”… would we cease treating her as an asserter at all?’ ([5], p. 335).

Shapiro answers this by arguing that one need not necessarily perform any response at all to have meaningfully asserted something:

‘But why can’t the dialectical norm theorist regard such a speaker as having made an assertion? Brandom seems to have just this possibility in mind when he discusses “bare assertions”: cases in which the asserter undertakes a commitment yet “disavows” the responsibility involved, because they aren’t “prepared to shoulder the justificatory burden”.’ ([3], p. 87).

Shapiro utilises Brandom’s notion of ‘bare assertions’ ([4], p. 643) to explain this further: ‘Brandom doesn’t hold that neutralizing the challenge is required: speakers aren’t in general required to rebut challenges to their assertions on pain of having to retract, as they are in some kinds of regimented disputation.’ ([3], p. 80).

While meaningful assertional practice will often involve engaging with interlocutors, it is wrong to expect one to engage with every rebuttal (or move in the language game of “asking for” reasons) lest one not be playing that game at all. One can neglect to play that “move”, but still have normatively committed to do so (regardless of any failure to). It is this commitment, as opposed to actually doing so in every instance, that makes an utterance assertion: one might raise such a commitment and ineffectively discharge it. This becomes intuitively clear when we imagine a number of situations. To use the example of a colleague asking questions about the premise of my article over coffee, I might elect not to engage with their questions about the veracity of my claims by ignoring it, saying “pfft”, and/or rolling my eyes, but this would be an extremely poor discharge of my engagement with the social practice involved in assertion. It will likely lead my colleague to think I cannot defend against the rebuttal (not to mention that I am rude and not worth going for future coffees with). I will still have asserted something by virtue of having raised that commitment, though my engagement in the social practice of assertion will be poor. The very fact we can consider such neglect ineffective at all highlights the fact a commitment has been raised in making the claim. If no such commitment had been raised, it would not even make sense to think in terms of effective/ineffective engagement with that normative obligation. Not responding to a query can, where a reply is warranted, be understood as simply failing to meet the obligations raised by my assertion.

A further niche but illustrative instance of this is assertions made on social media, where asserters will often receive an impossible number of interlocutors in comments. These are however very important instances of public truth-seeking speech and any reasonable theory of free speech would include assertions made on social media as well-protected speech. Similarly, on a smaller scale the point remains valid—if one is at the pub and someone challenges the claim that the earth is round, and the original asserter goes “pfft” or ignores the challenge completely, then this is not providing reasons for the original assertion. In fact they are outright refusing to play their next move in the game of giving and asking for reasons. But one would, I think, still want to consider them as having asserted something—that the earth is round—and as participating in the language game of assertion/performed a speech act of assertion. It is not the obviousness of such a claim that makes it unnecessary to respond, for one could make the same case about something non-obvious, such as an assertion that lower tax rates yields better state income overall. One might not engage with an objection to this, but refusing to justify the claim/meet objections does not mean one is not relevantly asserting something, or engaging in the language game of giving or asking for reasons to believe something, it is just that they may not be doing so persuasively depending on their reasons for not meeting objections. Whether assertions are persuasive or not though is precisely not relevant from a regulatory perspective as the point is for the law not to pre-suppose what points are persuasiveas I contextualise further in Sect. 4.

While Shapiro considers Brandom’s point about bare assertions to be an acknowledgement that one can assert something without engaging with any rebuttals or questions, the terminology of this being a “bare” assertion or an “assertion proper” has no contingency here. The point is that we reasonably understand one to have asserted something if one fails to discharge their assertive obligations, bare or otherwise. As Shapiro goes on to say ‘assertional practice is compatible with there being cases of “bare assertion”’ ([3], p. 87) which is the important point for my application of the theory to the truth-justification. Assertion as a social practice can therefore take place with a range of—or no—responses and still be engaging in the game of giving and asking for reasons to believe things. It is simply a matter of contextual effectiveness of the assertion, and the assessment of that effectiveness by interlocutors is the very basis of the truth-justification, as I shall detail in Sect. 4. A poorly performed act of assertion can nonetheless be a significant (speech) act of assertion. This is an essential aspect of the dialectical account I advocate for, since one should be able to (from a free speech perspective) incompetently assert something. It is therefore the mere fact a commitment/obligation to engage with interlocutors (in a game of giving and asking for reasons to believe things) has been raised by an assertion that is significant for the dialectic account, regardless of how well that commitment/obligation is discharged.

Having argued for a non-restrictive account of assertion, there is little need from the perspective of legal utility to comprehensively analyse the linguistic particularities of all the moves possible within the language game of assertion that may be useful for a more abstract philosophical analysis (as explored further in the work of the theorists I apply here). Suffice to conclude that some assertions will be effectively responded to and other times the asserter will fail to respond effectively. To not respond at all is simply a sub-set of the effectiveness/ineffectiveness of an assertion. One may fail to properly discharge ones obligation to do so, or in the event no response is warranted rightfully disregard an objection, or not have the opportunity to respond to all challenges or questions. It is not the job for the law to attempt to anticipate which speech is asserting effectively, as the engagement in this process is exactly what is imagined by the truth-justification for free speech. However, what kind of speech is meaningfully contributing to truth-discovering discourse within a neutral framework is useful. Such a neutral framework is offered here.

Before embarking further into contextualising this claim in existing legal work, it is important to resolve a potential terminological issue. There are a variety of synonyms for assertion which might serve as well as “assertion” and illuminate its purpose as a descriptor for a certain kind of linguistic activity. Brandom, in advocating for a dialectic account, writes that ‘(o)ne of the original senses of 'assert' in its broader normative use is as meaning to defend, champion, or justify’ ([4], p. 641). I go with the term assertion due to its existing prominence in the literature, but the important point is not a commitment to a particular definition of, nor the term, “assertion”. Instead, I draw on an analysis of the conditions of a certain language game—that of giving and asking for reasons to believe something. Terminologically, “defending”, “championing” or “justifying” fits similarly into the framework of the game of giving reasons to believe something—the specific social practice indicated by these rough cognates is what is important. For my purposes herein, any speech that engages in the language game of giving and asking for reasons is interestingly “assertion” in a truth-justification sense. This way of understanding assertion is a disputed issue in speech act theory, but it is also well defended. Whether we call this linguistic social practice “assertion” is not material—engaging in a discourse of giving or asking for reasons to believe something is the key factor and we can call this, accurately enough, “assertion”.

3 Greenawalt’s “Assertions of Fact and Value”

The dialectical account introduced in the previous section is intended as an advancement of legal theoretical employment of “assertion”. In Greenawalt’s foundational legal and philosophical exploration of what speech ought to be regulated within a free speech principle he introduces an analytical distinction, the first part of which is a ‘situation-altering utterance’ ([1], p. 40). Such utterances are in his view ‘ways of doing things, not of asserting things’ ([1], p. 58). In contrast, he prioritises assertions as worthy of legal protection ([1], p. 43) [also referred to by Greenawalt as statements ([1], p. 60) and claims ([1], p. 40)]. In his analysis, situation-altering utterances should not seriously concern a principle of free speech and in contrast assertions constitute an archetypical instance of speech that should be protected. This is a central dichotomy to his theory—through which he seeks to understand what kinds of speech are regulable without offending a coherent principle of free speech. As he puts it, ‘(t)he central idea about situation-altering utterances…is that they actually change the social world in which we live.’ ([1], p. 58). In contrast, ‘claims of fact and value’ ([1], p. 40) are ‘communications that the justifications for free speech plainly cover’ ([1], p. 40).

Greenawalt’s main concern is with analysing purportedly regulable speech which involve ‘ways of doing things’ ([1], p. 58) and ought to be regulated. This regulation-focused theme tends to be consistent throughout the literature that applies speech act theory to free speech. The most significant area of this work concerns pornography and hate speech in particular, influenced by earlier applicators such as Judith Butler [8] and Catherine MacKinnon [9]. Both these areas of focus have been developed into a significant set of theories by authors such as Rae Langton [10], Mary Kate McGowan [11], Ishani Maitra [12] and Abigail Levin [13]. Regarding hate speech for instance McGowan argues through speech act theory that it ‘can enact norms that prescribe harmful practices’ ([11], p. 18). Regarding pornography, Langton makes the case that ‘(b)esides depicting and causing subordination…pornography is, in and of itself, a form of subordination.’ ([10], p. 26). Abigail Levin takes a distinct angle that ‘speech act theory is more fruitfully applied to the state’s speech act of rights deployment than it is to the actions of hate speakers and pornographers.’ ([13], p. 103). However, little work is spent on the counterpart notion of speech which asserts things as put forth by Greenawalt—the character of such speech, and what makes it ideally relevant for protection under a principle of free speech. I focus my developments on this aspect of Greenawalt’s theory accordingly. While Greenawalt’s work is concerned with free speech more broadly than just the truth-justification as a theoretical principle, I limit my considerations to only that aspect of his work.

As Greenawalt notes, not everything that counts as “speech” in a literal sense is relevant for a free speech principle ([1], p. 40). Speech is often regulated despite being, in a literal sense, speech. Part of the job of free speech law and theory is to identify what kinds of speech warrant legal protection. He considers that such ‘(c)lassification would indeed be simplest if all standard acts of communication, and no others, were covered by the justifications for freedom of speech’ but an ‘absence of fit between standard communications and the reasons for freedom of speech poses both a problem of terminology and a practical problem about the application of a principle that restrains interferences with speech.’ ([1], p. 41). Therefore, what counts as “speech” should be informed by the justifications that underpin such a principle. This article takes this to be an uncontroversial view and operates on the basis that speech which does align with the justifications for protecting speech should be reluctantly and carefully regulated.

I briefly introduce Greenawalt’s notion of situation-altering utterances, based on J. L. Austin’s theory of speech acts [14], solely in order to inform his understanding of the counterpart notion of “assertions of fact and value”. The intention of this article is to enhance our understanding of the latter, but to understand Greenawalt’s account of assertion, and a part of the benefit of the dialectical account I apply herein, it is necessary also to understand the basic thrust behind his counterpart notion of situation-altering utterances. It warrants noting that my enquiry is not specific to any particular legal system and Greenawalt’s work, in its relevant considerations here, shares this generality, being ‘about free speech both as a political principle generally and as a constitutional standard in the United States.’ ([1], p. 4). My focus is on free speech as a principle, and his concerns regarding speech regulation in the US are not considered herein. If my application is taken to be persuasive, then it is one that could be applied to any legal system that commits to free speech seriously.

Greenawalt’s work therefore focuses on how situation-altering utterances differ from assertions of fact and value. His ‘essential claim…is that utterances of these sorts are situation-altering and are outside the scope of a principle of free speech.’ ([1], p. 58). Situation-altering utterances, his argument goes, are regulable because they are ‘a means for changing the social context in which we live’ ([1], p. 57), which they do by altering the obligations of interlocutors: ‘the conventions of language and of ordinary social morality make certain utterances, such as promises, count as far as one's moral obligations are concerned.’ ([1], p. 58). His point can be demonstrated via some simple examples. For instance, with respect to agreements, he writes that ‘(w)hen two people have agreed to do something, each has undertaken an obligation toward the other to perform the task, an obligation that did not exist before the agreement was made.’ ([1], p. 63). In the same vein, he writes ‘(b)y such utterances, persons commit themselves to perform acts and/or impose on others a duty to act.’ ([15], p. 208). Major examples of such obligation-enacting speech are some kinds of threats ([1], p. 90) and orders ([1], p. 84)—if one is threatened to perform a task with a gun trained on them then this is, in a strong sense, altering ones obligations. If a mugger says “give me your money or I will kill you” his utterance is clearly extremely “obligation enacting” on the receiver. Alternatively, an order from a superior in the military (“fire!”) is imposing a severe obligation on the recipient of that order, due to the hierarchy of power involved and potential sanctions ([1], p. 58).

Rather than focus on what makes “assertions of fact and value” worth protecting, Greenawalt focuses in detail on what makes such situation-altering utterances regulable, as he ‘concentrate[s] more attention on communicative acts whose possible coverage within the borders of free speech is debatable, especially solicitations to crime and threats.’ ([1], p. 3). His identification of “assertions” is intended be a foil to his focus of situation-altering utterances and their ‘nonapplication’ ([1], p. 58) to free speech concerns. He goes into great detail on what makes speech acts such as threats, orders, agreements and promises “situation-altering” and what this means for their potential regulability—but these concerns are not the focus of this paper. I turn now instead to consider his counterpart, relatively underdeveloped, notion of assertions of fact and value—which he highlights in contrast to what he calls situation-altering utterances.

The more detailed notion of assertion found in speech act theory I draw on in this paper improves upon Greenawalt’s theory. Greenawalt focuses on speech which has the ‘dominant purpose…to accomplish something rather than to say something’ ([1], p. 40). In contrast, my aim here is to look at what it is, in more detail, to “say” something, in at least one important archetypical way and explicate how this is performed. This should in turn help resolve difficulties within Greenawalt’s theory in identifying the kind of statements/assertions that meaningfully engage in truth-discovery, and explain how such speech can be understood as situation-altering in a specific way that engages with the discovery of truth.

Greenawalt therefore primarily explores what kind of communications are not of concern to the free speech principle. He gives extensive consideration as to why some speech is not appropriate for protection by a principle of free speech. In contrast, he claims that assertions of fact and value (or statements/claims of fact and value, which he uses interchangeably throughout) are relevant for free speech protections, but as this is not his focus he does not explore what is meant by this beyond a few remarks:

‘Claims about general facts are critical for people's understanding of the world they inhabit, for their choices about how to live, and for their decisions on public issues. The truth-discovery justification applies strongly to general factual statements, and suppression of such statements would undermine independence of judgment and personal development.’ ([1], p. 43).

As examples of this he considers some basic examples such as: ‘factual propositions of a general sort, those asserted in statements such as “Physical objects have gravitational force,” “Rapid inflation causes social instability,” and “The human personality survives death.”’ ([1], p. 43). In addition to these claims of fact, he considers claims of value (evaluations) within the same framework:

‘A principle of free speech also reaches general and particular evaluations. General claims, like “Love is the greatest good” or “Capital punishment is unjustified,” and particular claims, like “You should not lie about this to Mary,” are bases on which people consider what situations and practices are desirable and how they should behave.’ ([1], p. 44).

In constructing this regulatory dichotomy between assertions and situation-altering utterances, Greenawalt therefore draws on the lessons of linguistics, but (rightly) does not expect such lessons to transpose directly to legal use:

‘I draw from efforts in the philosophy of language that have relevance to a principle of free speech, but the central analysis must lie in the domain of political and legal philosophy, and conclusions must be based on reasons that are significant for those fields.’ ([1], p. 59].

As he notes, ‘(c)onceptual analysis alone cannot yield a practical distinction between the question of coverage and ultimate protection’ ([1], p. 5). In my attempt to expand on his considerations, this is an approach I emulate. Philosophy of language and linguistics can help legal thinking about how language functions and what that means for regulatory questions, but what it says is not of automatic utility—its useful applicability is contingent on normative commitments within law. In this context, the prior presumption is that the discovery of truth through social discourse is one good reason to value free speech—and that therefore a strong descriptive understanding of a dominant way speech engages in this (through assertion) can assist our understanding of what should be protected. Philosophical or linguistic understandings of language can therefore only hope to inform free speech understandings contingently on the values which underpin it; as Greenawalt notes: ‘(l)anguage serves a variety of functions, only some of which are covered by the special reasons for freedom of speech.’ ([1], p. 339). The social function of asserting facts and values is illuminated further by speech act theory. The investigatory question herein is not which is right in linguistic terms, but what understanding is the best fit for a legal understanding presupposing a commitment to truth-discovery through discourse.

In illuminating situation-altering utterances Greenawalt therefore draws on linguistic theory, remarking that ‘categories developed in the philosophy of language, though not simply transposable into political principles, can illuminate the proper boundaries of free speech.’ ([1], p. 3). Though he modifies situation-altering utterances for legal use, he nonetheless bases their function on ideas developed within speech act theory—the idea that language is capable of ‘doing things’ ([1], p. 58) and which can ‘actually change the social world in which we live.’ ([1], p. 58). Accordingly, Greenawalt notes that ‘the point of many communications is neither to transmit information nor to assert values.’ ([1], p. 57). A separation between speech that asserts things and other kinds of speech is, this article contends, a valid one, but it is simply situation-altering in a way that warrants protection, given a prior commitment to the truth-justification. While my application of speech act theory here is not intended as a rebuttal of Greenawalt’s core thesis, as (a) his primary focus is situation-altering utterances, not assertions and (b) the idea that some speech is especially relevant as assertion is one I share; I offer a refinement of this view in exploring how assertion is (positively) significantly situation-altering.

The dialectic account can better explain the way in which speech “does” something specifically important for truth-discovery. The imposition of a normative obligation generated in asserting something, and the engagement with connected speech acts, such as challenging/questioning, is key to explaining assertion dialectically. Assertion is therefore better understood as a way in which one may perform a particular situation-altering utterance, but it is the way in which it does so that is exactly what makes it especially valuable as a speech act for free speech considerations.

Greenawalt does rightly consider the objection that ‘assertions have a significant performative aspect and also alter the world.’ ([1], p. 61). However he also considers that ‘(s)ituation-altering utterances…actually alter the normative world, shifting rights or obligations or both.’ ([1], p. 59). However, the dialectic account beneficially explains assertion by focusing on the commitments or obligations raised by asserting things. As put by Rescorla, detailed in the previous section, assertion is understood ‘in terms of how it alters the speaker’s normative standing’ ([2], p. 99). Assertion is significantly situation-altering, just in a way that makes it relevant for the truth-justification. The very fact the dialectic account purports a situation-altering version of assertion is what gives the theory its explanatory and justificatory power—as opposed to the isolation of a semantic statement—as it understands assertion in relation to other speakers/speech acts. Greenawalt briefly remarks that making an assertion impacts what may constitute an ‘apt response’ from an interlocutor ([1], p. 61) by which he means that making statements limits the relevant responses that can be made sensibly in reply. For example, if I say “black coffee is better than milky coffee”, there are some responses which do not make any sense in reply. However, the range of responses that make sense has no relation to the obligation generated on the asserter to engage in the language game of giving and asking for reasons to believe things as explained by the dialectic account.

While I argue therefore that the dialectic account can explain assertion as situation-altering, ultimately whether this makes them situation-altering by Greenawalt’s particular definition is not of primary significance. What is important instead is the contention that assertion alters normative standings in the game of giving and asking for reasons to believe things in the ways discussed in Sect. 2. I propose it is much clearer and simpler to accept that assertion is importantly situation-altering, and the way in which it is makes it relevant for a free speech principle. The essential point is to disassociate assertion from identifying a static statement and instead understand it as a practice that alters normative obligations, and permissible moves, on the speaker and interlocutors. This understanding highlights with much greater specificity a kind of social practice (assertion) that really matters for free speech.

Part of the problem for Greenawalt’s formulation then is that other kinds of either irrelevant or regulable speech (as far as a principle of free speech is concerned) are often also statements of fact ([1], p. 91). To begin with a basic example of this, Greenawalt explains in different situations why certain statements can be made exceptions. He invites the consideration of a gym assistant giving a gym member a locker code. Giving such a code is clearly a statement of fact (“the code is 8713”) but also has no obvious relevance to a principle of free speech. To explain this he considers that such a piece of communication is too ‘linked to narrow practical aims’ for which ‘the justifications for free speech are remote.’ ([1], p. 47). This seems like a stretch to account for the fact that statements can be used in everyday discourse to perform acts that have nothing to do with free speech. The fact such a justification need even be sought is due to the fact that “statements of fact” are too indeterminate a category of speech—it casts too wide a protective net. Considering the same example under the dialectic account, an assistant at the gym giving someone a locker code is simply not participating in a game of giving and asking for reasons to believe things. If a gym member, upon hearing the code, were to start challenging if the code really is 8713 or agreed that it was indeed the code, it would rightly prompt strange looks. This can be understood on the basis that no commitment to engage in this game has been taken up by the speaker (there may be other kinds of professional commitments raised, but this is beside the point). Further examples of statements that Greenawalt creates exceptions for are resolved by the same application of the dialectic account. For instance, Greenawalt accepts that some situation-altering utterances may involve statements of fact such as saying ‘I will’ at a marriage ceremony or an umpire asserting ‘out’ in a game of tennis ([1], p. 60). These are, by his account, situation-altering utterances that do not warrant protection from a principle of free speech, but are also statements ([1], p. 60). I propose that the dialectic version of assertion resolves these extremely well, since—in the same way as the gym locker example—there is no social convention of giving and asking for reasons to believe things taking place here. These therefore don’t need explanation as exceptions to assertional practice when understood through the more specific lens of the dialectic formulation.

While the above are examples of irrelevant speech (a principle of free speech simply does not care about these utterances) other kinds of often desirably regulable speech may operate via statements. Threats are perhaps the best example of this, as threats may involve all manner of statements but are as a general rule subject to a variety of legal prohibitions. Greenawalt notes threats can be split into ‘pure’ and ‘conditional’ threats ([1], p. 90). In essence, pure threats do not offer any alternative (“I’ll kill you”) and conditional threats do (“If you don’t do as I say, I’ll kill you”). Due to his commitment to statements of fact being important for free speech, Greenawalt deals with these somewhat uncomfortably. Regarding pure threats he says ‘(s)ince these threats involve particular assertions of fact, what one plans to do on some future occasion, they fall within the broad range of communications to which a broad principle of free speech applies.’ ([1], p. 91). This is in part potentially a fair statement, since there may be other ways to justify protecting some kinds of threats beyond the truth-justification. For example, Greenawalt considers that protecting speech which operates as an emotional outlet is a valid reason to protect speech ([1], p. 27), and some sort of threats may have this aim ([1], p. 91), particularly if the threat is not sufficiently sincere and expresses displeasure more than anything else. However, many kinds of threats will not be like this, and have no relevant connection to the reasons for valuing free speech. It is hard to reconcile any justification for free speech with a statement of fact such as “I’ll break your legs next time I see you here”. In any case, any kind of threat (even a tame one functioning as an emotional outlet) would be properly excluded from the justification from truth, despite being a statement of fact. Similar examples of speech that are unproblematically regulated, despite being often performed contextually via statements, are speech acts such as: orders (“I am your boss”), agreements (“I find the terms acceptable”) and bribes (“there is a lot of money in this briefcase”). These all involve statements of fact or value but typically have little to do with the reasons for valuing free speech, especially from the truth-justification perspective. Such speech acts would be distinguishable from “assertion” under my account and thus are coherently and consistently excluded from assertion as an archetypically truth-relevant speech act. In performing any of these other speech acts one is not participating in the social convention of giving and asking for reasons to believe things, or raising commitments to argue about the propositions, even though these speech acts may involve making statements.

It is clear then that not all statements are relevant for potential legal protection. There are many potential statements of fact/value that are not well described as assertion which my account here neatly excludes. However, all activity that participates in the game of giving and asking for reasons to believe things is at least relevant for the truth-justification, even if such speech will sometimes be regulated. Considering assertion dialectically therefore reveals a certain kind of communication that is important from a truth-justification perspective, as opposed to the communication of statements of facts/value wholesale, which is not sufficiently discriminatory against irrelevant speech. This promotes clear boundaries between other kinds of speech acts, that can be statements of fact, like threats or orders, but which are not participating in the convention of giving and asking for reasons offered by the dialectic account.

In addition to the apt discrimination of what kinds of statements/assertions/claims of fact and value matter for assertional practice, and the explanation of assertion as a form of situation-altering utterance, one abstract but significant additional way in which this is an enhancement of Greenawalt’s distinction is that it readily accommodates the idea that assertion is “doing something” as much as any other speech act—it instead locates the analysis into the kind of thing it is doing. We can see that assertion is simply one way of doing something, and that something is particularly relevant for free speech. This resolves a theoretical tension within his work ([1], p. 58) in that, for speech act theory, all speech is an action that is “doing something” in some way or another [14, 16].

While the dialectic theory of assertion I apply from philosophy of language describes, in general free speech theory, the relationship between the practice of assertion to truth-seeking for the truth-justification as a whole it also remedies Greenawalt’s theory for these reasons. This is especially significant given that Greenawalt specifically relates his distinction between statements and situation-altering utterances to the truth-justification himself ([1], p. 20) [though his project goes into other justifications in addition ([1], p. 27)]. These insights borrowed from linguistic theory into the nature of assertion can therefore be seen to fit into Greenawalt’s broad goal by informing how speech might meaningfully ‘say’ ([1], p. 40) something in a way that makes it very relevant from a free speech perspective. The relation of my formulation of assertion to the truth-justification broadly, beyond Greenawalt’s theory, is explained further in the next section.

4 The Truth-Justification for Free Speech

This section serves to provide context for my theory in broader free speech literature and to show that it is widely applicable across the variety of formulations of the truth-justification, even to theories which are sceptical of the truth-justification for free speech. There are a great variety of potential justifications for a legal system’s endorsement of some sort of principle of freedom of speech. To name a few of these, Frederick Schauer writes that ‘a vast array of theories has sought to explain the importance of freedom of speech in terms of embodying or fostering what others have variously described as self-realization, self-expression, liberty, dignity, individuality, or autonomy.’ ([17], pp. 233–234). However, as expressed by Schauer:

‘Alongside these justifications, and in important respects more longstanding, is the idea that freedom of speech serves as an effective mechanism for locating truth, for identifying and expunging falsity, and for increasing the stock of human knowledge.’ ([17], p. 235).

The kind of speech protected under the truth-justification has natural overlap with others justifications for free speech, particularly that of democratic self-government [18]/democratic citizenship [19]. For example, ‘nearly all speech restrictions that interfere with the search for truth also interfere with the right to “participate in the formation of public opinion.”’ ([20], p. 595). Assertional practice as identified in this paper is potentially important for a variety of justifications for free speech and the model I advocate for here could be useful in the context of other theories. The most relevant though, and the focus of this paper, is the truth-justification, which is also as put by Greenawalt, likely ‘(t)he most familiar argument for freedom of speech.’ ([1], p. 16). Other important theories, such as those based on democratic justifications [18, 19], autonomy [21, 22], dignity [23, 24] and others [25] come with a variety of other considerations which do not necessarily prioritise truth. The benefit of the model of assertion I put forward here is directly related to its ability to identify truth-relevant discourse, and it is therefore most at home in the context of the truth-justification. The relation of my theory herein to these other justifications is one that could be taken further, but each would require its own dedicated analysis. My conclusions are therefore specific to what the dialectic account can tell us about the truth-justification.

The justification from truth is ‘longstanding’ ([17], p. 235) and John Stuart Mill advances one of the earliest versions [26]. Mill considers liberty of speech to be valuable for a variety of reasons, such as the fact popular opinion may be wrong and the censored view right; that a view may be partially right; or a wrong opinion might sharpen correct ones even in being wrong:

‘If the opinion is right, they are deprived of the opportunity of exchanging error for truth: if wrong, they lose, what is almost as great a benefit, the clearer perception and livelier impression of truth, produced by its collision with error.’ ([26], p. 19).

Mill takes a strong approach to the truth-justification (one which has been the subject of considerable discussion and alterations, which I discuss shortly):

‘If all mankind minus one were of one opinion, and only one person were of the contrary opinion, mankind would be no more justified in silencing that one person, than he, if he had the power, would be justified in silencing mankind.’ ([26], p. 18).

Mill therefore does not restrict truth-discovery to those views which are true (epistemological problems of what “truth” means aside momentarily). Instead, he holds that even wrong opinions contribute to the greater understanding of truth ([26], p. 19). He further defends this idea by saying ‘(w)e can never be sure that the opinion we are endeavouring to stifle is a false opinion; and if we were sure, stifling it would be an evil still.’ ([26], p. 19). If some form of this argument is accepted contemporarily—which we shall see momentarily it is—then knowing what speech is relevant is an important step. The fact that any given opinion need not be correct to contribute to truth-discovery coheres with the model of reasoned discourse put forward by the dialectic account, as the whole point of discursive truth-discovery is not to presuppose that truth via pre-emptive regulation. The ability for interlocutors to assess the effectiveness of a given assertion—through engagement using other speech acts like challenging, questioning or agreeing—is therefore crucial to this view of free speech. The dialectic account provides a model to understand how ‘(a)ssertion invites normative assessment from diverse perspectives’ ([2], p. 98) to facilitate this view of the truth-justification.

One corollary, almost synonymous, notion to the truth-justification is the concept of the marketplace of ideas. Zoe Sherman, for instance, remarks that ‘‘(f)reedom of speech,’ as a phrase and as a concept, often keeps company with ‘the marketplace of ideas.’ ([27], p. 137). In connecting the marketplace of ideas with the truth-justification for free speech, she notes that ‘Frederick Schauer calls the idea that free speech is the best route to arrive at truth the ‘ruling theory’ of freedom of speech—not the sole theory, but the most widely recognized and cited.’ ([27], p. 137). There are variations on the marketplace of ideas [27, 28] but fundamental to them is:

‘(T)he idea that a regime of freedom of speech will enable the society within which such a principle is taken seriously to identify more true propositions, to reject more false propositions, and thus to facilitate the advance of knowledge within that society. The idea is sometimes discussed under the heading of the “search for truth,” and even more often as the “marketplace of ideas”’ ([17], p. 231).

Therefore little hangs on which term is used for my purposes herein. At the core of the idea is ‘the basic concept of freedom of speech as enabling a society to increase its level of knowledge, to facilitate its identification of truth, and to expose error has a wide and persistent currency.’ ([17], p. 231). The liberalised version of the dialectic theory of assertion—in the non-restrictive but discerning form I advocate for in Sect. 2—describes a form of social practice or language game which has an intimate connection with this idea.

The marketplace of ideas theory, and the truth-justification generally, is subject to extremely wide discussion over its potential variants, merits and problems, but the dialectic theory applies across these. Schauer writes that ‘(t)he search for truth/marketplace of ideas justification for a distinct principle of freedom of speech has for decades been subject to harsh criticism, but…the justification persists, both in judicial opinions' and in the academic commentary.’ ([17], p. 232). Others take a more vehement defence of the justification, notably Eugene Volokh, who writes ‘I think we should not dismiss the search-for-truth rationale-which is in practice similar to the marketplace-of-ideas rationale’ ([20], p. 595). The demonstration of the value of the truth-justification (or the marketplace of ideas theory) is a normative matter which this paper does not seek to demonstrate, except to note that the model of reasoned discourse offered by the dialectical account I apply fits such theories neatly, and can help identify instances of speech that relate to it.

A substantial body of critical work on the justification from truth derives from the idea that inequality within the marketplace of ideas renders it unfit for purpose in various ways [29–31], or that a free exchange of ideas may not promote truth [32]. Laura Nielsen, for instance, posits that the expectation on victims of offensive or harmful speech to ‘talk back’ ([31], p. 149) places ‘on less privileged members [of society] an unrealistic duty to respond’ ([31], p. 149). These forms of critique in general do not necessarily reject the positive relationship between truth and speech, but highlight a disparity in social power to use speech. One common suggestion within this line of critique involves advocating for a proactive state which provides additional mechanisms for particular views or vulnerable groups to express themselves. Katherine Gelber advances such an argument in detail ([33], p. 61) and has ‘developed the idea of a policy of speaking back, in which individuals who are the targets of hate speech are provided with the institutional, educational, and material support to enable them to speak back’ ([33], p. 51). Ioanna Tourkochoriti also advocates for a version of this, arguing that ‘soft paternalism towards changing hearts and minds is legitimate in order to prevent harm to others.’ ([34], p. 34).

A core premise of the truth-justification, as detailed above, is that speech should be protected for its truth-finding capacities. Tourkochoriti therefore acknowledges that ‘when discrimination results in concrete material harm, it is easier to make a case in favour of government intervention. In the area of thoughts and beliefs it is harder.’ ([34], p. 42). She considers the potential legal responses to this beyond criminal prohibition: ‘(t)he state has a broad array of tools that it can use in this area’ ([34], p. 47) and demonstrates that there are a variety of potential responses to this such as forms of public education: ‘(e)ducational methods broadly conceived can contribute towards preventing behaviours and legal enforcement should intervene when it is necessary to restore harm’ ([34], p. 42). Such theories of free speech add a layer of complexity for the value of the assertional model I advocate for here as they (by their very nature) do not support a model of speech regulation that defers to discourse between citizens in a content-neutral manner. That said, there is nothing fundamentally inconsistent between the model of assertion I put forward here and the view of a non-neutral/proactive state that takes a stance in issues and provides institutional support to particular views/vulnerable groups. Despite seeing the truth-justification in a different way to the typical liberal understanding (in a way that involves the state taking a side in what speech ought to be elevated) even these sceptical accounts involve a discursive understanding of “assertion” with which my theory is broadly compatible—it is just weighted by institutional involvement. Ultimately, as Greenawalt remarks about such perspectives they support ‘a particular understanding of freedom of speech rather than a rejection of the entire concept’ (my emphasis) ([1], p. 19).

There are many further sophisticated divisions within different understandings of the truth-justification but the account of assertion I offer should apply across these differences. While a comprehensive demonstration of this would be impossible here, in order to briefly illustrate this, there is the consequentialist justification from truth [as also discussed by Greenawalt ([1], p. 14)]—that free speech will produce overall societal good/collective truth—which is a different kind of truth-justification to the idea that it allows the individual to properly realise the truth. James Weinstein for instance argues against a consequentialist idea of truth-discovery, emphasising the deontological philosophical benefits of the pursuit of truth: ‘(a) more profound problem with characterizing the marketplace-of ideas rationale as a core free speech norm is that it justifies free speech in terms of the good it will produce for society as a whole, not as a true individual right.’ ([18], p. 502).

In contrast. Volokh understands it as ‘both as the right to uncover the truth for oneself and as the right to participate in the continuing development of human knowledge.’ ([20], p. 601). Whether one conceives it as valuable as an individual right or as a collective gain by a public contribution to truthful understandings of the world, my application of assertion applies to both, as the social practice of giving and asking for reasons, and the associated moves I attribute to that practice, have a strong connection to both. Another example of theoretical variance within the truth-justification is the distinction between the weak and strong justifications for truth:

‘This argument [the truth justification] can take a stronger, optimistic form or a weaker, sceptical form, or some combination of the two. The stronger claim is that in a free ‘marketplace of ideas’, true ideas and opinions will tend to find favour and eventually prevail over false ideas and opinions. The weaker, sceptical claim is that censorship will always be less likely to promote truth than free discussion because no authority can reliably identify wrong ideas and therefore suppression of ideas may simply lead to the suppression of true or at least partially true ideas.’ ([35], p. 390).

In either case, the role of assertive speech as I formulate it is core to both the strong and weak justifications, in the same way it is core to both the consequentialist and deontological understandings. Therefore even though the truth-justification is an umbrella term for a variety of different positions, my theory applies to these major deviations. I do not propose here to prove that those who value the justification from truth are right to do so, as the accounts above make compelling cases for different variations in detail and this section has shown that it is a widespread commitment. My aim is to assist in the development of a descriptive framework, informed by philosophy of language, on the presumption that at least some form of this long standing justification for free speech has merit to it. My account here refines the notion of assertion introduced by Greenawalt to better identify the kind of speech that can be rightfully described as engaging in a kind of reasoned discourse that contributes to truth-discovery in the broader literature.

While a detailed analysis of broader theories of truth is beyond the scope of this enquiry, given the centrality of truth as a concept to my argument these do rest in the background. For example, significant theories of this nature are (non-exhaustively but demonstrably) the correspondence [36], pragmatist [37] and coherence [38] theories. All of these are examples with a classical pedigree and extensive internal variance and entail different semantic, epistemic and metaphysical commitments [39]. It is therefore a potential advantage that the application I make here flies under the radar of this degree of abstraction and could be consistent with any such prior theoretical commitments as I shall briefly explain.

In useful summation of various theories of truth, including the above examples, Panu Raatikainen writes that ‘(t)here are many different kinds of things to which truth is ordinarily attributed: beliefs, judgments, (declarative) sentences, utterances, propositions, etc. These are now commonly called ‘truth-bearers’’ ([39], p. 218). The social practice of assertion as advocated for herein could possibly be considered in a loose way as involving such truth-bearers, though a particular relation to them is not necessary—whether assertional practice professes judgments/propositions/beliefs (etc.) is not material. As an example of this, while ‘(o)thers argue that the primary truth-bearer is a more abstract entity, a proposition or a ‘thought’ (in Frege’s sense)’ ([39], p. 218) the theory I purport here need not pre-suppose a view of truth that has specific epistemological commitments in any direction.

It might initially be thought that the social focus of the dialectic theory of assertion—seeing assertional practice as a dynamic social practice rather than isolated utterances—in relation to free speech fits best with pragmatist theories due to a shared emphasis on social and assertive practice [40]. However, Shapiro, in his relevant considerations of a dialectic account of assertion, similarly disassociates his particular theory of a dialectic account of assertion from these broader theories of pragmatism: ‘(n)or can I address the roles commitment accounts play in large-scale conceptions of language and thought, such as Peirce’s pragmatism and Brandom’s inferentialism.’ ([3], p. 76). Similarly, little hangs directly on these broader conceptual issues about truth in the context of my free speech application—an abstract view of truth which expected truth-bearers to pertain metaphysically, as typically the correspondence theory holds ([41], pp. 284–285), is equally provisionally compatible with my free speech arguments as is a pragmatist account of truth.

In a broad sense with respect to truth absolutism and relativism, and free speech this is a view shared by Greenawalt who notes that ‘the truth-discovery argument can survive a substantial dose of skepticism about objective truth.’ ([1], p. 17). Fundamentally it is important from a free speech perspective for people to be able to engage in assertional practice because the language game of giving and asking for reasons to believe things as detailed in Sect. 2 is at least relevant to any of these versions of truth. It is therefore possible to contrast two maximally different approaches to truth theory—‘pragmatism, of any stripe, will be against versions of the correspondence theory of truth’ ([41], p. 283)—and both are capable of cohering with my purported free speech-oriented dialectic understanding of truth.

An important remaining caveat is that that some scholarship has argued against the ideal of truth seeking on the basis that some speech is (in some contexts) sufficiently harmful to regulate anyway [42]. For example, arguments minimising holocaust are regulated in many countries [43]. Similarly, speech surrounding racial propaganda is variously considered justifiably regulable [44]. Another good example of this kind of speech, due to its almost universal regulation in legal systems that nonetheless have some sort of free speech commitment, is defamation. This is coherent with the application of my dialectic account. My theory herein is useful only with the assumption that, to at least some degree, the pursuit of truth is a valuable [if not totally overriding in every context ([35], p. 386)] justification. The argument herein merely operates on the assumption that the truth-justification is an important part of free speech law, and purports to provide a framework for an archetypical instance of speech which ought to fall into this principle. Therefore, if the framework offered here is acceptable, speech which fits into it should be regulated with care.

My argument is therefore not that speech that is assertive is always unregulable, but speech that is should be treated carefully by the law—that it should be restricted with caution—if that legal system values the truth-justification. As Caleb Yong writes, this is essentially a normative matter:

‘(E)ven if we accept the (not uncontroversial) claim that free discussion is the best way to discover truth, the strength of the free speech values protected by this argument will simply be contingent on the value attached to truth. As Schauer has pointed out, it is implausible to think that this value is absolute, or even that it has priority over other values and interests.’ ([35], p. 390).

Certainly the pursuit of truth is not an absolute nor perpetually overriding value with which to view speech regulation, and my account relies on no such absolutist conception. In his calibration of free speech values, Yong still retains it as a normative value in deciding what speech should be regulated:

‘Just as it is important to reach true judgments about which individual ways of life are good, it is important to reach true judgments about which collective ways of life are good, and which political and social arrangements are just. These are undoubtedly powerful values and interests. Nevertheless, their normative weight is presumably not absolute, nor should they always have strict priority over other social goals.’ ([35], p. 391).

The point then is not to say that any speech that violates the model of assertion I propose cannot ever be regulated—such a claim would be absurd—but simply to say that if it is accepted as a good way of understanding the relationship between speech and truth (and that truth is a value underpinning free speech) it should aid theoretical clarity on the issue and give pause when considering the regulation of speech which does do so. There are normatively meaningful examples of speech that fits my model that either could be, or in some cases are, regulated with potential justification.

Linguistic theory alone cannot resolve key issues in such a normative balancing act of values. While speech act theory can help identify truth-relevant speech in the way I argue, it doesn’t touch these value-oriented matters. Importantly, any such regulable speech is therefore not an exception to the utility of the dialectic account—which I argue applies comprehensively—as all my model proposes to do is identify speech that is relevant for the truth-justification and not identify speech that cannot be regulated. The normative balancing act of when to override the truth-justification is a different endeavour to the model I apply here. Therefore the regulation of some speech which does engage in the dialectical practice identified does not undermine the fact it still locates speech relevant for the truth-seeking justification.

5 Alternatives to the Dialectic Account

There are a number of alternative theories of assertion, some of which warrant explanation as to why I apply instead the dialectical model. One in particular—the liability account of assertion—may be thought to well explain the regulation of certain kinds of speech noted previously, such as defamation or racist propaganda. This is due to the fact the liability account focuses on how the assertional commitment a speaker makes alters their normative standing in such a way to make them liable for their assertion, requiring that ‘one takes responsibility as guarantor for the truth of the asserted proposition, assuming liability in the event that it’s revealed to be untrue’ ([3], p. 75). In contrast, while still a “commitment” account of assertion, the dialectical account holds that in order to assert something ‘one obligates oneself to respond in certain ways to appropriate challenges from fellow participants in reasoned discourse.’ ([3], p. 75). These are therefore distinct ways of envisioning assertion as the taking up of a social commitment. Peirce, a primary advocate of the liability conception of assertion [45], holds that:

‘(A)n assertion belongs to the class of phenomena like going before a notary and making an affidavit, executing a deed, signing a note, of which the essence is that one voluntarily puts oneself into a situation in which penalties will be incurred unless some proposition is true.’ ([45], p. 8.313).

Peirce considers an affidavit as an archetypical assertion due to this formulation—one fully commits to the truth of it, and expects liability if they are wrong. Peirce does not mean necessarily legal penalties though, as such liability could be conventional in other ways. For example, ‘(a)n act of assertion is a contract, the effect of which is that if what is asserted is not true, the assertor forfeits in a measure his reputation for veracity.’ ([46], p. 5). Shapiro explains this, drawing on Mitchell Green [47], putting it helpfully: ‘Green’s…claim that asserting is a way to stake one’s “credibility” on the status of a proposition.’ ([3], pp. 77).

While lacking any justificatory power in relation to the truth-justification, as it poses no relevant reason to warrant protecting speech, this may be thought to provide an interesting model for the regulation of some speech such as the above examples. Claiming that a restaurant’s food made one sick, for example, may be regulable due to the harm it caused the business if it turned out to be untrue, yet would nonetheless fall within a dialectical framework as the speaker would without a doubt be engaging in the social practice of giving and asking for reasons to believe things in trying to convince others that it was the spaghetti that made them ill. Further examples may include forms of hate speech which argue for the inferiority of a social group.

While the liability account may have prima facie persuasiveness in these kinds of context, its explanatory power is misleading. The truth-justification, as illustrated in Sect. 4, accommodates other values and it does not claim speech that seeks the truth can never be regulated, just that truth-relevant speech is of special relevance for protection and should thus be regulated carefully. The appeal of the liability account to explain exceptions to the truth-justification does not make it a good alternative to the dialectic, as such speech is much better explained as regulable in spite of its fit within the truth-seeking potential of the dialectic account I apply. In fact this is a significant advantage of the dialectical account—even when such speech is regulated, it promotes a cognisance of the balance of values involved in that process, promoting careful consideration of such regulation, but does not preclude regulation if desirable.

Having argued against the potential appeal of the liability account, one other kind of model of assertion warrants consideration: the various forms of the “correctness” approach, which are the main ‘rival’ ([3], p. 76) to commitment accounts ‘according to which asserting is the speech act constituted by a rule to the effect that the act is correct only if a specified condition obtains’ ([3], p. 76). I do not dwell extensively on correctness accounts, since they are all I argue susceptible to one characteristic that makes them incompatible for the truth-justification. While my explanation as to why these accounts are a poor fit are brief, it is worth including them as they may for some present viable alternatives to what I offer, and may be explored more fully from another perspective. It may be the case that other theorists wish to make a case for them and as I shall indicate for some existing theories there seems to be some overlap. However, I propose that the crucial incompatibility factor of correctness accounts of assertion for free speech is that ‘(c)orrectness rule accounts evaluate an assertion in terms of whether it’s categorically permissible’ ([3], p. 82) to make. This involves a focus on various factors other than social convention as captured by the dialectic account. What they all have in common instead is they focus is on whether the assertion itself is made justifiably.

To briefly introduce the key variants of this, Rescorla identifies a variety of possible (conflicting) ways of understanding assertion from a correctness account perspective. One prominent version is ‘The Knowledge Norm: One should assert only propositions that one knows’ ([2], p. 99). However, ‘(c)ountless other norms with the same structure are possible, including:

‘ The Honesty Norm: One should assert only propositions that one believes.

The Truth Norm: One should assert only true propositions.

The Warrant Norm: One should assert only propositions that one believes with ‘sufficient’ warrant.’ ([2], p. 99).

Whether an utterance is categorically permissible under such views therefore takes a variety of different forms. Though some of these have potentially more appeal to a free speech understanding than others, I reject all these for a principle of free speech for fundamentally the same reason: that ‘such norms [are] restrictive, since they restrict which propositions one may assert’ ([2], p. 99). Some of these correctness accounts may have some initial attraction to them from a free speech point of view, and one may be more inclined to make a case for one over another, however they are all inappropriate for the same basic reason: that they set conditional limits on what kind of things can be justifiably asserted. The problems are slightly different for each and I will very briefly demonstrate these for each in turn.

The truth norm can be dismissed the most readily as patently inapplicable to a free speech construction of assertion, since the whole point of the truth-justification for free speech, as shown in Sect. 4, is to reveal through discourse what is true. Pre-determining that truth is required to assess any statement by this correctness norm, and such a formulation goes against the whole basis of a free speech principle. The warrant norm might be understood as a softer version of the truth norm. Under such an account, one would not restrict assertion to only statements that are true, but it nonetheless substantively limits what counts as meaningful assertion since assessment need be made of whether one had sufficient warrant to assert what they did. This involves similar objections as the truth norm, to a lesser degree, in that it requires an evidential hurdle for a speaker to meet to justify their right to make the claim in some way (as a speaker, or evidentially). However, one of the main advantages of the truth-justification is that it protects unpopular, minority or wrong views, and is typically neutral from a content or speaker perspective. Both the truth and warrant formulations fly in the face of this fundamental feature of the truth-justification, whereas the dialectical account coheres with it.

As for the knowledge norm, ‘a categorically “correct” assertion is one that doesn’t violate the prohibition against asserting what one doesn’t know.’ ([3], pp. 82–83). But what it means to know something is philosophically fraught. [48, 49]. No matter how defined, the requisite standards of knowledge in this framework do not match the level of certainty needed to engage usefully in truth discovering speech, as they create barriers that do not provide an explanatory correlation between everyday assertion and the normative reasons for protecting free speech. Moreover, asserting what one already knows is only one small part of the overall utility of truth discovering speech—making assertions one is less than certain about to engage in discourse about the veracity of the claim is clearly useful from a truth-justification perspective for either the original asserter or participants of the dialogue.

Lastly, and perhaps most attractively of the four, the honesty norm has some appeal to it in relation to free speech theories which strongly favour the intent or mental state of a speaker (one such account being a free speech theory of “communicative action” favoured by Solum [50] and Wright [51]). Greenawalt also considers extensively, similarly, what a free speech principle should do with ‘insincere statements’ ([1], p. 48). For the purposes of this paper, suffice to say that legal investigation into whether an asserter really thinks what they are saying is problematic for a general theory of free speech, and does not represent the free speech norm of the truth-justification as shown in Sect. 4. Avoiding such an epistemologically problematic enquiry into a speaker’s mind is one of the advantages of seeing assertion in terms of the social practice I describe herein. Moreover, one should be able to engage in the language game of assertion without being sure or believing what they are saying is true. A big part of coming to learn the truth, on an individual and collective level, is by participating in discussions in which one may not believe what they are saying, either because they are engaging as devil’s advocate; have simply not made up their mind yet; accept some, but not all of their own premises; or are exaggerating to push the boundaries of a point. The honesty norm can be understood as incompatible with such a view on the truth-pertaining utility of discourse.

Some variation of these correctness theories may therefore be thought to fit into a theory of free speech (I do not conduct a comprehensive survey of this here), but this would be an atypical understanding of the truth-justification. They fit well, for example, with the following example offered tentatively by Schauer in which he considers whether public discourse is valuable from a truth perspective compared to the views of select people:

‘If the empirical claim of the marketplace of ideas account is understood, as it should be, as a comparative one, the question is not whether the marketplace of ideas is good at locating truth and advancing human knowledge, but, rather, whether it is better at doing so than a committee of experts, an agency of bureaucrats, or a hierarchy of powerful (and often self-interested) public officials.’ ([17], pp. 238–239).

Such an understanding would be well understood perhaps as a warrant norm—an expert might be considered to have sufficient warrant for their claims in a way not applicable to an ordinary citizen. Alternatively, this could be considered within the remit of the truth account, more strongly, in that such a committee could be considered to represent the truth. It is significant though for my argument that ultimately even Schauer’s more sceptical work concludes in favour of the traditional truth-justification:

‘I simply assume that there is at least some soundness in the comparative empirical claims that lie behind standard marketplace of ideas theory. That is, I will assume that there is some reason, in some contexts and on some subjects, to believe that something resembling an unrestricted clash of facts, ideas, and opinions will at times be more likely to tend toward the identification of truth, independently defined, then will some number of other approaches, institutions, and methods, especially those associated with government or other powerful and self-interested truth-determiners.’ ([17], p. 239).

The account I put forward represents better this ‘unrestricted clash of facts’ ([17], p. 239). If one embraced the notion of committee-led discovery of truth instead, then certainly it seems clear something like the warrant or another correctness norm governed account would be a suitable understanding of assertion for that view. However, as Schauer is one of the more critical voices of the truth-justification it is notable that he still commits to some idea of free speech that is not contingent on expertise or warrant, but general public discourse. Those who more strongly doubt the utility of public discourse in determining truth may find some form of the correctness account valuable in the ways mentioned above, or others.

In contrast, I have argued for a very different understanding in the dialectical account as a form of commitment account which identifies assertional practice by way of the commitments raised by a speaker without recourse to the justifiability of their assertion. Notably, Shapiro draws attention to one of the advantages of an alternate way of understanding assertion in commitment accounts: ‘(c)ommitment accounts should thus be attractive to those who doubt that an account of assertion can presuppose a notion of categorical correctness’ (my emphasis) ([3], p. 83). Commitment accounts—specifically the dialectic—are a more fruitful understanding due to their reliance on the idea that, instead of assessing the validity of an assertion, they rely on an understanding that a speaker “commits” to any assertion in an identifiable discursive practice. This leaves aside any concern of what makes an assertion appropriate or correct to make and instead emphasises the social convention that surrounds the linguistic practice of assertion. It is therefore an advantage for free speech utility that commitment accounts ‘reject any notion of ‘assertibility’ or ‘appropriate assertion’ ([2], p. 99). Take for instance again the example of a twitter post questioning the safety of mRNA technology. The tweeter may or may not assert effectively or persuasively, but this is beside the point—the language game of giving and asking for reasons to believe things is well underway. Whether the speaker/writer handles rebuttals well is another matter which is accommodated by dialectic accounts which prioritise the engagement in this practice/game instead of the correctness of their assertion. It seems implausible to this writer that speech should, as a general rule, be regulated based on how well one engages in assertion. Poorly defended assertions matter for the possibility to engage in truth-pertaining discourse in various ways. Indeed, from a speaker perspective, sometimes one only knows that one has asserted something indefensible by the effective interjections of interlocutors, or alternatively one may not be abreast of the best arguments to support an assertion (and therefore engage with responses ineffectively), but still may be right.

Where commitment accounts differ importantly from correctness accounts therefore is that they accept that one can be understood to assert something without it being justified to do so. In other words, they leave open the possibility of asserting something badly. This, on a fundamental level, is attractive from a free speech perspective as understood in the literature shown in Sect. 4—any strict requirement of assertion is antithetical to a notion of free speech which embodies discovery of truth through discourse. As articulated by Mill, even wrong views can discursively sharpen the truth ([26], p. 19). Commitment accounts focus instead on the conventional requirements in committing to something being the case in dialogue, and the justificatory burden that comes with such a commitment. Instead of asking whether one has sufficient authority/expertise or the knowledge to really “assert” something, or whether what is being stated is “correct”, or whether one truly “believes” it, it is possible to simply look at the language game being played: for the dialectic account, a loose exchange of reasons to believe a proposition or set of propositions.

6 Conclusion

This paper has sought to apply work from the field of philosophy of language to inform a key way in which speech can contribute to the discovery of truth, and thus be of significance for free speech theory. The idea that “assertion” holds special relation to the truth-justification is one found in Greenawalt’s early work, but the use of linguistics since in free speech theory has mostly focused on identifying regulable speech, rather than well-protected speech. This article has attempted to address this gap by bringing contemporary linguistic literature to bear on free speech theory. I have sought to apply theoretical advancements about assertion in speech act theory to the truth-justification broadly. There are multitudes of competing accounts of assertion in the linguistic literature, but I have here argued for the application of a broad interpretation of the dialectic account due to its attention to the kind of language game that underpins truth-relevant assertional practice—giving and asking for reasons to believe things. I have argued that this offers a sufficiently open, but appropriately circumscribed, understanding of assertion from a truth-discovery perspective.

The application of assertion to free speech theory herein therefore develops the technical dichotomy introduced by Greenawalt between assertions and situation-altering utterances, and locates this understanding in the broader free speech literature on the truth-justification. Firstly, specifically for Greenawalt’s work, his focus on isolated assertive statements over-simplifies the fact assertion is a commitment based and participant-oriented exercise of reasoned discourse which is reliant on other relative speech acts—such as questioning, challenging or agreeing. This better emphasises assertion as a social practice that is relational to uncountable varieties of discourse that cannot be formalised propositionally, and with a focus on what kind of engagement in that practice accurately reflects meaningful assertion. This holds greater discriminatory power in identifying speech that is relevant for the truth-justification by excluding speech acts (that may be performed via statements of fact) such as threats, orders, agreements and bribes. Secondly, in contrast to Greenwalt’s view, assertion is significantly situation-altering, but simply in a way that positively explains its relevance for the truth-justification. The terminology of “situation-altering” aside, this promotes an understanding of assertional practice that focuses on the normative social obligations generated by a claim or statement. This additionally resolves the problematic idea that assertion is not meaningfully performative compared to other kinds of speech acts. The refined version of assertion advocated for herein embraces assertions particular propensity to do something like any other speech act, it is just that what it does (as a speech act) is especially relevant for the truth-justification.

I have argued also that the theory of assertion I apply fits best with the truth-justification in general terms, across variances in free speech theory, providing a content-neutral framework that can identify a key kind of speech—and the associated language game of that speech—that contributes to the societal and individual pursuit of truth. An understanding of the limits of a descriptive linguistic framework—that it can identify when speech is potentially truth-relevant, but not when it should be regulated or not—keeps normative legal analysis in its correct primacy of place, but assists that analysis by enhancing an understanding of what kind of speech is meaningfully truth-seeking in at least one crucial way. While the primary goal of this paper has been to help inform legal theory by applying advancements in linguistics, it is also my hope that the demonstration of the practical utility of a widely applicable understanding of assertion might provide some additional evidence for such an understanding in linguistics—particularly in support of non-restrictive understandings of the dialectic account.