1 Introduction

Norms can play two kinds of essential roles in arguments. There are rules, inferential principles, and the like, that can be seen as or reconstructed as articulating our argumentative exchanges. There are also norms that are the conclusion of so-called practical arguments. My interest lies on the latter.Footnote 1 In a very general form, norms are the conclusions of practical reasoning, where what is at stake is a decision on what to do. In particular, my attention focuses on deliberative dialogues, where the participants make a joint effort to reach an agreement on the best course of action (Walton and Krabbe 1995).

In modern democracies, deliberation under certain conditions of equity, freedom of speech, inclusion, etc. is taken to be the best way to allow the citizenry to jointly make decisions that can be considered as politically legitimate. Some of these decisions deal with the formulation of norms (in the form of laws, rules, directives, guides of action, etc.) that the participants aim to enforce and endorse, recognizing themselves as the makers of the norm. Nevertheless, there are also many cases in which the agents affected by the norm have not participated in the decision-making process. In such cases, it is essential for democracy that they can know the reasons that have supported the final decision. For only in this way will their compliance with the norm be based on a rational consent. A deliberative setting for the adoption of practical norms of collective action makes of the compliance with those norms a rational decision and not a passive or imposed acceptance. This is why clarifying the communicative process through which a norm can be agreed upon and endorsed is relevant and worth pursuing. This paper aims to contribute to this clarification from the perspective of speech act theory and the theory of argumentation.

Some examples of the type of argumentative dialogues I am considering here would be the following: the members of a parliament debating and amending new traffic legislation; a group of neighbors deciding whether to increase their community quota in order to have some improvements carried out in their building; and the members of a university department discussing and agreeing on the criteria for the selection of new faculty. Generically, thus, the type of practical reasoning I am here concerned with is a deliberative dialogue carried out by collective agents and where the expected outcome is a norm that should guide action.

A point of departure in this work will be the assumption that the statement of a norm of action of the type here considered can be the conclusion of an argumentative dialogue. But notice that a concluded norm can be understood differently, the interpretation depending on which illocution is considered to have been performed. Therefore, assuming that the statement of a norm can be the conclusion of argumentation, I will ask what particular type of illocution is the conclusion of practical argumentation, when this conclusion is a norm. This is the question I propose to address.

Usually, in argumentative dialogues and particularly in deliberation, the norms at issue are translated into claims of the form, “We ought to do A”, or “Action A ought to be done”. In this way, in concluding a norm of action, the participants conclude the statement of a norm, as something different from concluding an imperative. However, the concluded norm can be interpreted as an exercitive in Austin’s (1975/1962) terminology (a speech act which would roughly correspond both to directives and declaratives within Searle’s 1969 taxonomy; more on that below), whenever the conclusion binds or affects other agents who are not the ones that have made the decision. It is possible that the claim also is taken to be a hybrid speech act with the force of an exercitive and a commissive, whenever it is the discussants themselves who become bound by the conclusion. An alternative interpretation would see the concluding illocution as belonging to the family of assertive speech acts (see Green 2020; Sbisà 2020). My aim is to give support to this last view.

In what follows, I will be endorsing a normative approach to speech acts, as it has been developed by Marina Sbisà (2002, 2006, 2020) and taken into account or further elaborated by other authors (Witek 2015, 2019; see also Heal 2013, Caponetto 2021). According to this approach, speech acts can be characterized by saying how they change the normative stances of the interlocutors, namely, their obligations, commitments and responsibilities, as well as their rights, entitlements and authorizations, and other similar stances. Moreover, these changes are effected due to social acceptance (possibly tacit) or, in more informal settings, due to the interactants’ agreement that those changes have taken place. My suggestion is to extend and apply this framework to study the kind of illocution a concluded norm can be said to be.

Before this, however, it will be in order to clarify some of the notions that are going to be discussed.

2 Verdictives and Exercitives Withing the Austinian Framework

The normative approach to speech acts that is here considered has also been characterized as conventional (as opposed to intentional) and Austinian. It can be seen as a completion and expansion of Austin’s (1975/1962) original framework. According to Austin, verdictives are characterized by the giving of a verdict or judgement; they consist of “the delivering of a finding official or unofficial, upon evidence or reasons as to value or fact” (p. 153). In an interactional account, this type of speech act presupposes that the speaker is epistemically competent in relation to its subject matter. Moreover, verdictives change the normative positions of the participants in the exchange. They commit the speaker to giving justification or support for the verdictive whenever requested and assign the interlocutors certain rights and entitlements. If they accept the verdictive, they become entitled to issue other speech acts that follow from the former; in other case, they acquire a right to raise doubts and objections and a legitimate expectation to receive an answer.

An exercitive is “the giving of a decision in favour of or against a certain course of action, or advocacy of it. It is a decision that something is to be so, as distinct from a judgement that it is so: it is advocacy that it should be so, as opposed to an estimate that it is so (…)” (1975/1962, p. 155). From an interactional perspective, exercitives are characterised for transferring or assigning an obligation or responsibility on the addressee, and this by virtue of the influence, power, or authority accorded to the speaker. Thus, exercitives presuppose that the speaker has the corresponding authority or authoritativeness. Moreover, the interlocutors become bound by the corresponding responsibilities or obligations.

It is worth noticing that Austin considered the dimensions of evaluation corresponding to verdictives and exercitives to be different. Verdictives have connections with truth and falsity, in that they have to have certain correspondence to the facts, broadly understood (Austin 1979/1961, pp. 117 and ff.) This correspondence depends on the speech situation, the situation of the world, and the pertinent evaluative criteria in the given circumstances. Therefore, certain subtypes of verdictive speech act, like estimates and assessments, could better be evaluated according to their being right or fair (see Sbisà 2020). In contrast, exercitives are capable of being evaluated, in general, according to their correctness or incorrectness (their fairness or unfairness, but also their legality or illegality, appropriateness or lack thereof, etc.). Austin tries to make this point clear by means of the example of an umpire calling a certain movement ‘Out’, ‘Three strikes’ or ‘Four balls’ (1975/1962, p. 153). Here, the verdict can be assessed as true or false to the facts. In that speech situation, however, the umpire has the required authority to issue an exercitive speech act that introduces certain changes in the game. Therefore, in such a situation the umpire’s speech act should be seen an exercitive.

Searle’s influential contribution to speech act theory is sometimes presented as a direct continuation and completion of Austin’s original views. Nevertheless, there are significant differences between both developments and, as I see it, Searle’s version of the theory has certain difficulties in order for it to give a satisfactory account of the specific normativity that articulates interaction in speech. To see why, it is worth noticing that for Searle, what he terms illocutionary effect is achieved merely by the hearer grasping the speaker’s intention to produce an utterance with a certain content and force. He writes, “In the case of illocutionary acts we succeed in doing what we are trying to do by getting our audience to recognize what we are trying to do. But the ‘effect’ on the hearer is not a belief or response, it consists simply in the hearer understanding the utterance of the speaker” (Searle 1969, p. 47). In this way, the conventional effect characteristic of illocutionary interaction is blurred.

A second reason to take some distance with Searle’s version of speech act theory is its defense that speech act types are constituted by a set of necessary and (jointly) sufficient conditions. The idea that there are groups of conditions that suffice to produce a certain speech act seems excessively rigid and subject to counterexamples. For instance, it becomes difficult to argue that an insincere promise is a promise after all, that as such creates an obligation and entitles to legitimate expectative and, in case of non-compliance, to a right to hold the speaker accountable.

For these reasons, the interactional framework here endorsed is not based on Searle’s notion of the illocutionary. Notwithstanding this, Searle’s taxonomy has been highly influential and widely adopted. From an extensional point of view, it can be said that, broadly speaking, Austin’s exercitives comprise Searle’s directives and declarations. In the same broad sense, Austin’s verdictives would mainly fall under the category of Searle’s assertives; but, due to the dimensions of assessment noticed above, some Austinian verdictives should be classified as Searlean declaratives. As Sbisà has noticed, Austin’s classification was not aimed at yielding an exhaustive division into non-overlapping classes. It offers an approach to complex procedures where a speech act can have a hybrid character and be a prototypical member of a class while at the same time present some traits of another (2013a, p. 34). This is why, resorting also to Searle’s taxonomy (without endorsing the theoretical views that motivate it) can be of help in specifying the type of exercitive that is a particular norm of action.

3 Practical Argumentation, Practical Reasoning, Deliberation

A statement is a linguistic unit, a declarative phrase that pragmatically can be seen as a linguistic action by means of which a speech act is performed. In practical argumentation, the conclusion is (usually and primarily) regarded as the statement of a norm, saying that a certain action A ought to be carried out, or that A is advisable, allowed, the best option, etc. to be carried out. In many cases, the obligation, allowance, etc. is attributed to some particular person or institution (I, you, we, the neighborhood community, the Government, the UN, etc.). Here, I will contend that what the conclusion of practical argumentative dialogues states is first and foremost a practical judgement concerning what to do.

Argumentation, following Toulmin et al. (1984), will be understood here as “the whole activity of making claims, challenging them, backing them up by producing reasons, criticizing those reasons, rebutting those criticisms, and so on” (p. 14). In some cases, the activity of reasoning and that of arguing are interpreted as the two sides of the same coin; reasoning would be a cognitive, inferential process that a person performs in thought, whereas argumentation would make linguistically explicit such process. In practical reasoning, thus, the purported conclusion would be either an action (Aristotle, Anscombe 1963) or an intention to act (Lewiński 2021a, b) on the part of the reasoner.

This notion is not as clear-cut as it seems, however. For reasoning designates also the process of drawing inferences, in other words, the transition that leads from a set of propositions to another proposition that these propositions can be taken to support (Corredor 2020a, b). Although this kind of passage is usually reconstructed in the form of rules (formal or informal), what is minimally required is, as Audi observes, “a sense of some relation of support” between the former (usually called premises) and the latter (called the conclusion) (2004, p. 126). This reconstruction is not to be equated to real argumentative dialogues, where the interlocutors can contribute in many different ways (making claims, challenging them, expressing doubts and concerns, raising objections and rebuttals, etc.) in the course of the dialogue. I take it that arguing is a communicative, interactional activity, whereas reasoning can be understood as the reconstruction of an abstract process, as conducted either in speech or thought.

The basic scheme of practical reasoning arguments (as the abstracted product of both practical reasoning and arguing) is usually taken to comprise: (i) a first premise stating that agent A has a goal/value G, (ii) a second premise stating that carrying out action B is a means to realize G, and (iii) a conclusion stating that, therefore, A should bring about action B. This basic scheme is instrumental and has been shown to be faulty in several respects. As Macagno and Walton (2018) urge us to keep in mind, practical arguments are taken to be grounded on argumentative inferences from goals and values to a choice and a recommendation to act; as such, they presuppose “the determination of what is good or better” (p. 520). This, in their view, is insufficient to capture “how the basic, instrumental scheme is locked in together with supplementary evaluative and classificatory schemes” to form a more complex structure (p. 521). Additionally, Audi observes that the second premise in the basic scheme “may indicate not only instrumental means but constitutive means”; the latter are “the kind essential in the end itself” (2004, p. 128). Both considerations, namely: first, that a basic instrumental scheme is in need of completion concerning the determination of which ends are good or better; and second, that certain means can be underscored as constitutive and essential to the end itself, should warn us against a quick reduction of the practical to the instrumental.

Finally, the notion of deliberation is also differently understood among scholars. Whereas for some authors it is possible to deliberate internally in thought by considering pro and con reasons in relation to a course of action, I will follow here Walton and Krabbe (1995) in characterizing deliberation as a communicative activity where two or more interlocutors try to reach a common solution to a practical problem and do so by means of exchanging reasons and assessing those reasons. As such, the goal of the dialogue is to find the best available course of action. Moreover, the interlocutors’ relationship is collaborative and aimed at coordinating plans and actions. These features guarantee that deliberative dialogues are a species of practical argumentation. Whereas the participants in argumentative dialogues oriented to solve practical problems do not always have a collaborative attitude but focus on their individual goals and interests, deliberative dialogues are to be seen as a joint effort to reach a common conclusion. This makes of deliberation a main communicative means of social coordination.

A further concern is whether the ends themselves are also capable of rational justification, whenever these ends are put into question. Strong conceptions of practical rationality (as exemplified by Habermas 1996/1992; see also Ihnen Jory 2020) contend that practical ends (pragmatic, ethical and moral) can be seen as justified whenever all those affected by their introduction could agree in a rational discourse to its introduction.Footnote 2 A rational discourse is an argumentative dialogue of a deliberative kind that complies with certain demanding requirements on the logical, dialectical and rhetorical level (possibly together with other requirements that are social, political, and economic). These requirements should guarantee that the structure of the dialogue “rules out all external or internal coercion other than the force of the better argument and thereby also neutralizes all motives other than that of the cooperative search for truth.” (Habermas 1990/1985, pp. 88–89; see also Habermas 1984/1981, pp. 25–26; 1996/1992, pp. 4, 323). As some commentators have highlighted, this normative theory of practical rationality can serve as a counterfactual model to critically assess real practices. Moreover, the model shows that also ends themselves, when problematized, can become the subject of the argumentative dialogue.

All the above leaves open the question posed at the beginning of this paper, namely, what type of illocution is the conclusion of practical argumentation. We already mentioned several different proposals available in the literature, according to which the conclusion of practical arguments would be the action itself (Anscombe following Aristotle), an exercitive (like a prescription, allowance, recommendation, invitation, etc.), and a verdictive (stating a practical judgement). The first of these proposals is nowadays of only historical interest. The second and third ones are going to be considered and discussed here.

4 Exercitives as the Conclusion of Practical Arguments Establishing Norms of Action

The interpretation according to which the conclusion of practical reasoning and argumentation should (in the most general case) be seen as an exercitive speech act has a long tradition in argumentation studies. According to Hitchcock (2002), this tradition goes back to Aristotle, for whom the conclusion of practical reasoning is a decision to perform a certain action, which immediately brings about the action itself (Nicomachean Ethics). In this vein, Hitchcock contrasts practical reasoning with reasoning about what to believe or, as he prefers to call it, epistemic reasoning.Footnote 3 He notes that even if, in practical reasoning, the conclusion is a policy decision or the like, this “is not the sort of thing that can have a truth-value” (2002, p. 248). In his light, the “imperative conclusions” of practical reasoning should not be assimilated to those of epistemic reasoning “by recasting those imperative conclusions as indicative ‘ought’ statements”. There are two reasons for not doing so. First, it is doubtful that ‘ought’ statements have truth values. Second, it is possible to affirm an ‘ought’ statement and make the opposite policy decision, without this move being a contradiction. For these two reasons, Hitchcock concludes that ‘ought’ statements of the kind here considered “are not the same as policy decisions”. In Hitchcock’s view, both reasons lend support to the view that the conclusions of practical reasoning cannot be assimilated to ‘ought’ statements; instead, they communicate imperative decisions whose character is irreducible to that of statements of facts.

Similar views seem to be held by Fairclough and Fairclough (2012), for whom the conclusion of practical arguments is a “claim for action” of the form: “Agent (presumably) should do A”. However, notice that this claim can be interpreted differently, either as an exercitive (directing action or advocating for it), or as a verdictive belonging to the family of assertives. The same type of underdetermination of the illocutionary force of conclusions in practical arguments is to be find in Macagno and Walton’s (2018). They propose a complex, modular approach to practical argumentation, showing how in deliberation the basic instrumental scheme that goes from a goal and a means to an action must be supplemented with other argumentation schemes of evaluative and classificatory type. The conclusions of those types of arguments can take one of the following forms: “Therefore, A should bring about action B” (in instrumental practical reasoning), “Therefore, B should not/should (practically speaking) be brought about” (in arguments from consequences), and “Therefore, B must be carried out” (in arguments from rules). As in the previous case, whether those conclusions should be understood as appeals, requirements, orders and the like, or they are statements of the corresponding norms and as such belong to the family of assertives remains underspecified. An additional difficulty is that this framework does not seem built in the terminology of speech acts, but aims to consider abstract agents and schemes apt to be implemented in automatic machines.

In an illuminating paper, Lewiński (2021a) argues that the conclusions of practical argumentation are what he terms action-inducing speech acts, a broad class that comprises commissives, directives and their hybrids (the latter including proposals and offers). This new category, modelled after Searle’s (1969) taxonomy, is to be distinguished from assertives (which Lewiński equates to representatives), expressives and declarations. What distinguishes action-inducing speech acts is that “their point is to get an agent (whether ‘I’, ‘you’, or ‘we’) to perform an action that will bring the world into a state captured in the intentional content of the speech act” (Lewiński 2021a, p. 445). Still, the elements in the category can be further sub-distinguished depending on who is the primary agent (the speaker, the hearer, or both) and their illocutionary strength.

I sympathize with Hitchcock’s view that a concluded norm is not to be equated to a factual statement, as such capable of truth or falsity. Moreover, it seems to me that the more general view that the conclusions of practical arguments are claims to action or, in Lewiński’s original reformulation, action-inducing speech acts is well justified. For one thing, the view answers to the agents’ own perspective. Whenever a norm is agreed upon in a deliberative setting or it is institutionally enacted after being debated and approved, it is the norm as such which has an exercitive force (and in some cases, also a commissive one) on those affected by the norm.Footnote 4 This very perspective seems to have been endorsed by the scholars who see the conclusion of the argumentative dialogue as itself belonging to the family of imperatives, directives, and the like. Nevertheless, I think that there are good reasons to resist such endorsement.

5 Assertions as the Conclusions of a Critical Discussion

The second, alternative view to the aforementioned one is famously represented by van Eemeren and Grootendorst (1984). Within the framework of pragma-dialectics, argumentation is conceived as a complex speech act to which different types of speech acts can contribute in the different stages of resolution of a difference of opinion. In particular, the authors contend that the attitudes and points of view of the participants need not refer only to assertive speech acts that can be true or false; instead, these attitudes and points of view may also be related to other speech acts whose acceptability is under discussion. Notwithstanding this, they write, “In fully externalized discussion the expressed opinions and the argumentation must always in our view consist of elementary illocutions belonging to the class of assertives” (van Eemeren and Grootendorst 1984, p. 98). Even if, in the process of the discussion, other speech acts may play a role (namely, that of commitments, declaratives, questions, requests, etc.), speech acts that convey points of view belong to the family of assertives.

To exemplify their contention, the authors propose the following utterance: (1) “Let’s take an umbrella, or do you want to get wet?”. Here, the first part of the utterance, “Let’s take an umbrella”, is the conclusion of a practical argument, where the reason that lends support to it is presented by means of a rhetorical indirect question, “or do you want to get wet?”. According to van Eemeren and Grootendorst’s analysis, the speaker recognizes the possibility of a dispute arising about their proposal of taking an umbrella. The dispute would concern the question of whether the proposal, namely, the interlocutors’ taking an umbrella, was a good one. Therefore, and despite its grammatical form, the expressed opinion at issue could have taken the alternative wording, “It is advisable to take an umbrella”, which, according to the authors, is an illocutionary act belonging to the category of assertives.

It is worth noticing that van Eemeren and Grootendorst’s reconstruction of the underlying argument performed by (1) entails that, for them, the conclusion is not a mere expressed opinion in the form of a proposal. The concluding speech act makes explicit the point of view that the speaker believes to be in need of support. Therefore, it has not the illocutionary force of a proposal, but that of an assertive speech act.

As I see it, van Eemeren and Grootendorst’s intuition is right. Even if the concluding speech act might (possibly, should) be seen as an exercitive if uttered alone, and it will become an exercitive (a piece of advice) whenever the interactants accept it as such, in the argumentative context of example (1) the speech act is making explicit a judgement susceptible to be questioned and discussed.

To a certain extent, van Eemeren and Grootendorst’s paraphrase could be analyzed not as a type of assertion, but as an expositive speech act in Austin’s (1975/1962) original terminology. Expositives are speech acts in which an explicit performative verb shows “how the ‘statement’ is to be fitted into the context of conversation, interlocution, dialogue, or in general of exposition” (p. 85). Within the group of expositive speech acts, Austin includes “I argue (or urge) that…” and “I conclude (or infer) that…”. These speech acts have an “orthogonal” or hybrid character, in that they usually perform at the same time another illocutionary act (see Sbisà 2020). In the particular case of van Eemeren and Grootendorst’s views, their refusal to endorse an alternative analysis equating the conclusion to a directive speech act strongly suggests that they conceive the concluded point of view as a verdictive. According to Austin, verdictives essentially consist of “giving a finding as to something -fact, or value- which is for different reasons hard to be certain about” (1975/1962, p. 151). These speech acts are performed “upon evidence or reasons” (p. 153) and need not be final but can take the form of an estimate, reckoning, or appraisal. Austin emphasizes that verdictives should be seen as judicial acts, acts of judging, as something different from legislative or executive acts which are both exercitives.

This is the suggestion that I propose to follow.

6 Verdictives as the Conclusion of Practical Arguments

As a speech act, the conclusion of a practical argument can be subjected to doubts and objections; it can be criticized and argued against, which opens the way to a new argumentative exchange. This is characteristic of assertive speech acts, particularly of verdictives, where the speaker is committed to the correctness of their words. In the case of an exercitive, in contrast, the responsibilities and authorizations, obligations and rights, etc. that the speech act conventionally brings about are of a very different kind. This second type of normative positions is paradigmatically related to obligations and rights, commitments and authorizations, etc. concerning certain actions. The conventional, normative effects of performing an exercitive speech act are not the same as those of performing a verdictive.

For the sake of illustration, suppose that the members of a UN organism are responsible for the adoption of some measures confronting the threat of climate change on a global scale. They have thoroughly examined all the scientific evidence available to them and discussed possible actions and their consequences. Finally, they have concluded that the energy transition from fossil fuels to renewables cannot be required from all country parties in equal measure. Thus, their conclusion takes the form,

Example 1

“The UN should not require that all States parties achieve the energy transition from fossil fuels to renewables in equal measure at the same time”.

Now, as long as this statement is pronounced in the discussion setting, it still can be subjected to doubts and objections, and more evidence or better detailed reasons in favor of it can be demanded and provided. Depending on who carries the burden of proof, the proponent has certain dialectical obligations (e.g. provide further evidence or reasons, answer doubts and objections, etc.); their interlocutors, in their turn, have correlative dialectical rights (e.g. to request further evidence or reasons, raise doubts and objections, etc.) These normative stances correspond to a verdictive speech act. Once the statement of the norm is concluded upon, it is capable of assessment according to its fairness, appropriateness, applicability and the like, in the way practical judgements are.

As soon as the statement is approved as a new normative guideline, however, it will bind the UN in the way in which exercitive speech acts do. The UN policies and actions will have to conform to it lest the international institution incurs inconsistency with its own norms and is prone to criticism. The normative guideline institutes an obligation for the subject of the norm, namely, the UN, bringing about certain political entitlements and rights for the recipient countries. For example, they hare authorized (possibly, obliged) to enact domestic policies in agreement with the guideline. These normative stances, brought about by the statement of the norm as an exercitive speech act, are not dialectical in nature and differ from those characterizing verdictives.

This is not to say that exercitives cannot be argumentatively examined and criticized. There are many different dimensions of evaluation for them, to borrow Austin’s expression. My hypothesis is that, whenever a norm of action is the conclusion of an argument, it should be analyzed as the statement of a norm and thus as a verdictive. If the context is appropriate, and the interactants are sincerely (or institutionally) committed to their argumentative exchange and its conclusion, then this verdictive speech act motivates and institutes a new illocution with the force of an exercitive. The interactants’ recognition and acceptance that the new illocution has been performed lends the norm its exercitive force. Thus, the same words perform two different illocutions.

It seems possible that these illocutions take place successively, or else in just one and the same act of speech in one span of time.Footnote 5 There may be situations in which the concluding verdictive and the exercitive are performed in succession, in two different spans of time. For instance, some of the participants could say aloud, “We have agreed that we should do A. Therefore, now we should do A!”, with unanimous acquiescence of the rest of deliberators. But it also seems possible that, in certain contexts (for example, institutional or highly regulated contexts, like a parliament), there are rules stating in advance that the conclusion reached (a verdictive) will immediately become enacted. In such contexts, the recognition accorded by the deliberators to the exercitive force of the agreed norm is, so to say, anticipated by their endorsement of the deliberation rules.

It can be objected that the difference between the concluding verdictive and the enacted exercitive, if correct, is just a conceptual one, not operative from the interactants’ perspective. To this, I am prepared to assent; moreover, this fact nicely accounts for the theoreticians’ preferred interpretation. But the distinction may be significant, whenever the acceptability and evaluation of the norm is at issue. As the example shows, the statement of the norm qua verdictive can be assessed according to different criteria of fairness, practical applicability, factual consequences, etc. The approved normative guideline qua exercitive, being capable of criticism as well, is to be assessed, e.g. by examining whether the established procedure to enact it has been followed correctly and completely and, once this is accepted, taking into account whether it is or it is not fulfilled, or has some consequential effects that were not foreseen before.

My contention is that the same act of speech (to borrow Green's 2021 term) can perform at least two different illocutions, whenever it occurs in a deliberative dialogue. I agree that putting forward a norm is, primarily and out of a deliberative setting, an exercitive speech act, in line with an invitation. Nevertheless, once it takes the form of a proposed normative statement in a deliberative dialogue (in the form: “We ought to do A”, “A should be carried out”, and the like) and is subjected to critical consideration (pro and con reasons), the act of speech is a new illocution, a verdictive: a judgement that things should be so, to paraphrase Austin. It is this judgement that constitutes the object of the deliberation. Analogously, once this act of judgement (or a revision of it) is reached as the conclusion of the deliberation qua concluded judgement, and to the extent that it has been agreed upon by the participants, a new illocution takes place, namely, the decision to endorse “We ought to do A” (and the like). Making a decision, endorsing a position, accepting an outcome, etc. are exercitive speech acts.

This proposal is subject to objections, some of which are to be considered in the next section. Before that, it is worth noticing that the idea that the same act of speech, the same uttered words can perform different illocutionary acts is by no means foreign to speech act theory (see e.g. Sbisà 2013a, b; also Lewiński 2021b). My proposal is to see the illocution as determined depending on whether its performance takes place within or as outcome of the argumentative dialogue.

7 Some Difficulties

To the above proposal, it can be objected that it is not clear how the transition from one speech act to another is to take place.Footnote 6 The answer to this question is given by the normative approach I am endorsing. It is the interactants’ agreement that certain changes have taken place in their normative stances (obligations and rights, commitments and entitlements, etc.), or the social acceptance of such changes that makes effective the performance of the illocution. In the particular case of Example 1, the UN discussants’ making the decision to approve and adopt the normative guideline agreed upon is what institutes the new normative illocution, an exercitive. From a structural point of view the performed exercitive, namely, the decision to undertake the norm is effected by the discussants’ final agreement on the concluding verdictive.

Still, it might be questioned why the act of making a decision is to be seen as a “conventional effect” in Austin’s terminology, and not as, say, a perlocutionary effect of the deliberative agreement. Austin distinguished an illocutionary act’s consequential effects of a causal nature from a response or sequel invited by convention. He writes, “If this response is accorded, or the sequel implemented, that requires a second act by the speaker or another person; and it is a commonplace of the consequence language that this cannot be included under the initial stretch of action.” (1975/1962, p. 117). This means that for Austin the giving of a response, even if it has been invited by convention, should be seen as the performance of a second act. In the particular case here considered of a deliberative setting where an agreement is reached on a practical norm, the speech act of making a decision on adopting the agreed-upon norm is a new speech act, different from the act of concluding the norm. In my view, the latter, concluding the norm, is a verdictive, whereas the former, making the decision to adopt it, is an exercitive speech act. Moreover, the exercitive is performed in virtue of the discussants’ agreement to do so. According to the normative approach to speech acts I endorse, the effects of speech acts are conventional in that they are brought about in virtue of the interactants’ recognition (or social acceptance) that these effects have been brought about (Sbisà 2002, 2006). Remember that the kind of effects that are being considered here are changes in the normative stances of the interactants (their obligations and rights, commitments and authorizations, etc.)

A second objection, which is related to the first one, is the following. There may well be cases in which the participants in the deliberation (and possibly other interactants) agree on making a decision in favor of the adoption of a norm, but this is not due to the argumentative dialogue itself; instead, the decision is motivated or forced upon them by some factual evidence.Footnote 7 For example, let’s imagine a community of neighbors deciding whether they should contribute some extra money to their quota in order to fix and improve the old plumbing network, which shows signs of being in bad condition. The practical statement to be discussed and agreed upon could have the form,

Example 2

We have to contribute extra funds to fix and improve the plumbing network.

Before they have reached an agreement, one of the shared pipelines starts leaking profusely and all the neighbors have to contribute as they had discussed. In such a situation, it is not an agreement reached after deliberation but the new adverse circumstance that forces them to accept the exercitive, “We have to contribute extra funds to fix and improve the plumbing network”.

It must be acknowledged that in this situation it is no longer possible to say that the exercitive, the making of a decision on what ought to be done, has been performed in virtue of a previous agreement on the judgement that the interactants ought to do so. The imagined situation of Example 2 is intended to make clear that the decision has been imposed upon the interactants by external facts. Nevertheless, in such a case there is no doubt that two different speech acts have been performed, the first one being the act of proposing a norm and the second one consisting of a decision. In both cases, the interactants’ agreement brings about the corresponding changes in their normative positions.

A potential third difficulty concerns the scope of the contention that one and the same statement of a norm can be of use in the performance of at least two different speech acts, the second one with the same content as the first speech act but a different illocutionary force. The framework here adopted is limited to practical argumentation in which the adoption of a norm is at issue, focusing in particular on deliberative dialogues. Yet, it might be objected that Example 2 does not seem like a case in which a norm is adopted. Other borderline or doubtful cases could be those in which a proposal is put forward and an agreement on it immediately reached, without further argumentative discussion. Also, the case of a tyrant just proclaiming that something should be done might be seen as unexplained.Footnote 8

In relation to Example 2, I think it safe to interpret it as a situation in which a norm is in fact agreed upon and instituted. In normal conditions, it might be expected that the decision made will be recorded on the proceedings of the meeting and will become in force, thus binding all the community members and making them accountable if they fail to comply with the norm. In those cases where an immediate agreement is reached, without deliberation or a concurring fact that imposes a decision, there are at least two different speech acts in play; namely, the proposal put forward and the exercitive agreed upon. For example, a group of youngsters are deciding what to do on Saturday evening; one of them suggests, “Let’s go to the cinema?”, and the rest of the group answers with enthusiasm, “Let’s go to the cinema!”. Here, no explicit reason has been needed. The first speech act has been a proposal or suggestion, and the last one an exercitive. Finally, the case of a tyrant dictating a norm should be seen as a case in which the speech act has bare exercitive force; to that extent, it is not put forward to be critically considered nor intended to reach agreement, but its only aim is to become in force and be complied with. It must be acknowledged that the last two types of case do not belong to the ones considered here, where -as remarked above- the focus is on argumentative dialogues aimed to reach an agreement on the adoption of a practical norm.

The previous discussion tacitly relies on a pending task: that of clarifying how arguing should be analyzed from a speech-act theoretic perspective, and what types of illocution do play a role in it. The following section is an attempt to suggest a response.

8 The Speech Act of Arguing

In exemplifying the distinction locutionary-illocutionary-perlocutionary, Austin writes, “We can similarly distinguish the locutionary act ‘he said that…’ from the illocutionary act ‘he argued that…’ and the perlocutionary act ‘he convinced me that…’” (1975/1962, p. 102) It is possible to take someone to be arguing without judging whether they are convincing their addressees. Moreover, convincing or persuading are perlocutionary effects intrinsically associated with the act of arguing, in contrast with other perlocutionary consequences. Thus, he writes, “The perlocutionary act may be either the achievement of a perlocutionary object (convince, persuade) or the production of a perlocutionary sequel.” (p. 118) Furthermore, arguing belongs to the category of actions that only can be achieved in saying something, as is the case of state, inform, estimate, consider, etc.

Austin classifies arguing among expositives. These speech acts help to clarify how speech fits into the course of an argument or conversation and contribute to organize the discourse; moreover, expositives have a hybrid or orthogonal nature that usually allows us to classify them into some other group, and possibly in more than one. The same applies to the act of concluding, as his examples suggest (1975/1962, p. 85):

“I argue (or urge) that there is no backside to the moon”

“I conclude (or infer) that there is no backside to the moon”

Both examples show that arguing is closely related to exercitives like urging, which exert some kind of powers, rights or influence, whereas the act of concluding is equated to that of inferring. Moreover, the parallelism in the examples strongly suggests that Austin understands arguing as the act of adducing a reason, in line with other contemporary theoretical treatments (notably, pragma-dialectics; see van Eemeren and Grootendorst 1984, 2004). It is not clear, however, how the connection between the acts of adducing a reason and of concluding from this reason should be established. Even if expositives help to clarify the organization of discourse, the precise articulation between arguing (adducing a reason) and concluding is not dealt with by Austin.

As I see it, both adducing a reason and concluding from this reason should be treated as verdictives in Austin’s original terminology. Undoubtedly, in most cases of practical argumentation, a speaker’s adducing a reason in support of a claim aims to convince or persuade their interlocutors to accept the claim; for this reason, the speech act can be seen as an act of exerting influence, in the same sense in which advocating is so. But it is important not to forget that adducing and concluding are two interrelated acts, in that one cannot make sense without presupposing the other. Moreover, I take it that their joint intrinsic function is that of achieving justification. To the extent that both speech acts are felicitous, the adduced reason is to be seen as offering justificatory support in favor of the concluded claim. This is not to say that in all cases of arguing, justification is achieved. But it seems to me not out of order to say that justifying a claim is a conventional effect of the act of adducing, in that the very concept of adducing a reason does not make sense without this internal relation of epistemic support with a corresponding conclusion.

Dealing with argumentation in the terminology of speech acts is frequent among argumentation scholars. Nevertheless, their theoretical analyses usually are not sufficiently overarching and systematic. There are, to my knowledge, two theoretical models that have systematically analyzed argumentation in the terminology of speech act theory, namely, pragma-dialectics and the linguistic-normative model set forth by Bermejo-Luque (2011). In pragma-dialectics, argumentation is a speech act complex consisting of a constellation of single speech acts. For Bermejo-Luque, arguing is a second-order speech act, which articulates the acts of adducing (a constative) and of concluding (from the adduced reasons), plus an implicit constative that asserts the inferential link between these two first-order acts. Here, I am not endorsing these authors’ models, although I am indebted to their work.

In previous work (see Corredor 2020a, 2021), among the conditions of felicitous performance of of the speech act of arguing the following have been suggested. Adducing a reason is a verdictive speech act. According to the normative approach I endorse, (i) verdictives presuppose that the speaker occupies a certain epistemic position or that they are epistemically competent in relation to the subject matter of what is adduced; (ii) the speech act commits the speaker to giving justification or support for it whenever this is asked for by the interlocutors; and in the particular case of adducing, I suggest adding a third condition, namely, (iii) that the speech act also entitles the interlocutors to critically examine the adduced reason, for example, its acceptability and its relevance for the claim at issue.

These conditions are put forward as necessary, but not sufficient, for the felicitous performance of an act of adducing. In certain contexts, other conditions may be appropriate or in order. For example, in judicial contexts certain presumptions might alter the distribution of the burden of proof or assign restricted rights to the interlocutors. But notice that these cases require a specific explicit regulation and some justification for it, which I take to be a tacit indirect recognition of the conditions here proposed. In a similar vein, the adduced reason might be subjected to additional constrictions of admissibility. Notwithstanding this, I take it that the above stated conditions will be in force as recognized by the interlocutors in (what can be taken to be) the general case.

In regard to the act of concluding, conditions (i)–(iii) put forward above for the act of adducing are in need of some modifications. Both precondition (i) that presupposes the epistemic competence of the speaker and condition (ii) that commits the arguer to provide justification are redeemed in the very act of arguing. But condition (iii) should be elaborated upon to yield a new condition (iii)′ that grants the interlocutors certain dialectical rights and assigns them certain dialectical obligations. For instance, they become entitled to raise doubts and objections and to oppose rebuttals and counter-arguments. Moreover, whenever they accept a conclusion, they become entitled to take it as a justificatory reason providing support for other claims.

In the particular case of a deliberative dialogue, once the concluded norm qua verdictive has been agreed upon by the participants, their endorsement of the conclusion becomes an exercitive speech act (just as making a decision and assuming a position are). Moreover, whenever the discussants also become bound by the norm, their illocution will have a hybrid character in that its force will also be that of a commissive. Remember that, in Austin’s (1975/1962) original terminology, exercitives are “the exercising of powers, rights, or influence” (p. 151); but they also are “the giving of a decision in favour of or against a certain course of action, or advocacy of it.” (p. 155). Commissives “commit the speaker to a certain course of action” (p. 157). According to Sbisà’s (2002, 2006) interpretation of Austin, exercitives (i) presuppose some degree of authority or authoritativeness on the part of the speaker and (ii) assign rights or obligations to or from the addressee. But notice that in exercitives, the authority is practical and not epistemic, as is the case in verdictives; moreover, the kind of rights and obligations involved are not dialectical but social and political in nature. In the particular case of a norm of action, provided that the speaker or institution has the required authority, its enaction creates an obligation to be fulfilled by the addressees of the norm. In concluding a norm, however, as something different from enacting it, the normative changes introduced are of a different, distinctly dialectical type, as argued before.

9 Discussion. The Illocutionary and the Perlocutionary in the Speech Act of Arguing

A possible objection to the proposal put forward in the previous section would be that condition (i) in cases of adducing does not appropriately distinguish between preconditions and conditions of felicitous performance. In answer to this, we may consider that, as Sbisà has remarked, descriptions of the illocutionary effects of Austin’s verdictives “appear to require the introduction of the epistemic dimension.” (Sbisà 2020, p. 9). Moreover, this epistemic dimension takes the form of requiring from the speaker a certain epistemic competence, to wit: “what entitles her to perform a verdictive or, more specifically, make a certain assertion is the possession of some relevant knowledge.” (Ibid.) As I see it, the speaker’s possessing knowledge can be considered a conventional condition of felicitous performance inasmuch as it is required and, as the case may be, acknowledged by the interactants, who will be entitled to hold the speaker accountable if the condition fails to be fulfilled. A fortiori, the same condition applies to speech acts of adducing. This is so because it is the interlocutors in the particular argumentative dialogue who assign the speaker such epistemic competence; therefore, they are also entitled to hold her responsible for it. Notice that in general argumentation is justificatory because the reasons given are taken to be true or otherwise well-established, or at least better established than the purported conclusion; for only then can these reasons give support to it.

Another potential objection concerns the divide between the illocutionary and the perlocutionary. It might be questioned whether the interlocutors’ normative positions, as brought about by the arguer’s act of adducing, should be seen as perlocutionary and not illocutionary effects. This objection throws doubt on the formulation of conditions (iii) and (iii)’. In my view, however, there is a clear conceptual distinction between the interlocutors’ acquiring certain dialectical rights and obligations, on one side, and on the other the responses they might give (like raising doubts and objections, asking for justification, manifesting agreement, etc.) In one of the passages quoted above (Austin 1975/1962, p. 118), it was made explicit that perlocutionary effects could be a perlocutionary object (convince, persuade) or the production of a perlocutionary sequel. After both the acts of adducing a reason and of concluding (from this reason) have been performed, the exchange that might follow is part of a perlocutionary sequel in Austin’s terminology. But notice that what is at issue in conditions (iii) and (iii)′ are the normative stances that the interlocutors recognize and assign each other, or that have been socially or institutionally assigned to them (i.e., what they are obliged or allowed to say and do). It is in virtue of this social and/or interpersonal recognition that the interlocutors have the normative positions they have in the argumentative dialogue. To that extent, their acquired dialectical obligations and rights, commitments and allowances, etc. can be seen as a conventional effect of the speech acts performed in the dialogue and, as such, an illocutionary one.

A final concern is related to the following remark. The proposed account seems to draw a dividing line between the conditions of felicitous performance of the interrelated speech acts of adducing and concluding, on the one hand, and on the other the criteria of assessment of both acts constituting a good argument. Austin, recovering his old distinction between constative and performative utterances (which he already had abandoned at this particular point), he writes, “We may feel that there is here a a dimension in which we judge, assess, or appraise the constative utterance (granting as a preliminary that it is felicitous) which does not arise with non-constative or performative utterances”, and this dimension concerns the “correspondence with the facts” (1975/1962, p. 140). Putting aside the difficulty of dealing with a notion of correspondence that might fit in Austin’s framework,Footnote 9 the passage makes clear that assessment presupposes a felicitous performance of the speech act at issue, in our case the interrelated acts of adducing and concluding.

Moreover and regarding acts of arguing, Austin writes, “Or again there is a parallel between inferring and arguing soundly or validly and stating truly. It is not just a question of whether he did argue or infer but also of whether he had a right to, and did he succeed.” (1975/1962, p. 141). It is worth remembering that, according to our previous discussion, Austin understood the act of arguing as that of adducing; and he explicitly correlated the acts of concluding and inferring. Soundness or validity, as assessment criteria, are to be predicated of arguments seen as the product of complete acts of adducing a reason and concluding (inferring) from this reason. This evidences that Austin is implicitly presupposing the interrelation between both acts. As for the assessment criteria themselves, a formal account takes it that an argument is sound only if it is a deductively valid argument with true premises (Hitchcock 1999). Alternatively, a more informal approach only requires acceptable premises (acceptable reasons) and some form of informal validity that lends plausibility to the conclusion; informal validity, in its turn, can be understood as e.g. relevance and sufficiency of the premises to draw the conclusion (Johnson and Blair 1977). Considering Austin’s historic and academic context, it is the first notion of soundness that he could have been referring to. If this interpretation is right, then an assessment of soundness or validity becomes possible only “granting as a preliminary” that the speech acts of arguing (adducing) and concluding under scrutiny have been felicitously performed. Therefore, soundness and validity are not conditions of felicitous performance, but criteria of assessment related to the “correspondence with the facts” of the argument.

The other two criteria that Austin mentions, whether the arguer “had a right to” and “did he succeed” are of a different kind. Our previous reading strongly suggests that the success of arguing (adducing) and concluding has to be equated with convincing or persuading the addresses, which as already seen is a perlocutionary object. In my view, however, whether the speaker had a right to argue has to be seen as a normative property belonging to the conventional conditions of felicitous performance. Such a right depends on the interlocutors recognizing it (in informal settings), or on a social and institutional recognition (in more formal ones). Moreover, only in this way does it make sense that in certain formal and institutional contexts where the speaker lacked such right, the speech act might result in a misfire.

At this point, it seems advisable to briefly recapitulate. My suggestion has been that when a norm is the conclusion of an argumentative dialogue, in particular in deliberative settings, it has to be seen as the statement of a proposed norm, i.e. of a practical judgement and, as such, as a verdictive speech act in Austin’s (1975/1962) original terminology. Whenever the interlocutors reach an agreement on the adoption of the norm, this decision has to be seen as a new speech act with the force of an exercitive. The same utterance is used to perform two different speech acts in the two stages of the argumentative exchange, a verdictive and an exercitive. Of course, the first proposal submitted to discussion in a deliberative setting does not always need to have the same wording of the conclusion reached. But at some stage in the discussion, the final formulation has to be put forward in the form of a practical judgement. My point is more general in that it concerns the illocutions that perform the acts of adducing and concluding.

In a previous section we had already noticed that the idea here suggested, namely, that the statement of a norm can be of use in the performance of at least two different speech acts, the second one with the same content as the first but a different illocutionary force, by no means is foreign to speech act theory. More generally, the tenet that an act of speech can perform different illocutionary acts has been particularly fruitful for the analyses of some pragmatic phenomena of special interest in current debates.Footnote 10 In the case here discussed of concluded norms in practical argumentation, after the norm is stated in the form of a verdictive speech act, it is in virtue of the agreement reached by the interactants (or otherwise due to the institutional and social context) that a second speech act with exercitive force is performed by means of the same utterance. It is an open field of research worth exploring under which conditions an utterance can perform more than one illocution.

10 Conclusion

There is in the specialized literature a disagreement as to what type of illocution the conclusion of a practical argument should be. For some scholars in the tradition of Aristotle, such a conclusion is peculiarly related to action, in the same way in which certain speech acts (characterized as imperatives, Searlean directives or Austinian exercitives, but also as commissives and other hybrid types) are. Alternatively, other scholars take it that the conclusion of practical argumentation is a speech act of the assertive family. I have followed suit with the latter in trying to lend plausibility to the suggestion that the statement of a norm subjected to critical examination in a deliberative dialogue is a verdictive. Once the norm is agreed upon, the same statement becomes a new speech act of a different type, namely, an exercitive (with possibly a hybrid commissive nature, whenever those that reach the decision on the norm are the same bound by it). In order to give support to this suggestion, I have taken into account the different normative interactions that both types of illocution institute among the interactants, and this in virtue of the conditions of felicitous performance of the acts of adducing and concluding.