Beliefs, Commitments, and Ad Baculum Arguments

Abstract

Typically, an ad baculum argument is one where an arguer threatens a respondent in order to induce them to adopt a standpoint. It is a fallacy, a common account goes, because the power to impose a standpoint is irrelevant to its truth or acceptability. However, fallacies, if they are to be anything, ought at a minimum to be persuasive, and it is hard to see how an ad baculum might persuade. Employing an ad baculum just underscores how terrible someone’s reasons are. Despite this, cases of fallacious ad baculum arguments seem to exist, and this is a fact that requires some explanation. This paper offers an account where the real target of an ad baculum is an audience downstream from the initial ad baculum exchange. This means that the ad baculum consists of misrepresenting the quality of evidence by means of the forced adoption of a particular standpoint.

1. Introduction

Suppose you and I are having an argument about whether p. You hold that p; I hold that not p. You adduce a series of arguments, so do I. Our disagreement is not shallow, so we get nowhere, but, because I am persistent, I say: “if you continue with this p business I’m going to punch you in the face.” Were I a more subtle person, I might have achieved the same result by suggesting that continuing to assert that p might occasion the revelation of some embarrassing personal fact about you: “it truly would be a shame if the Dean found out that you’ve been skipping your office hours” (or something like that). If you have had some elementary logic or critical thinking classes, you will likely recognize that my arguments are ad baculum—appeals to force. In a standard picture, a fallacious ad baculum argument occurs when one arguer uses a threat to induce an interlocutor to draw a conclusion (Van de Vate 1975, p. 43; Walton 2014, p. 296). They are fallacious, so the story often goes, because the threat, or the power to make good on a threat, is irrelevant to the truth, advisability or acceptability of the conclusion. Sticks-and-stones objections notwithstanding, violence does often seem like an option in argument, especially if we include threats of things like doxing, which is the practice of revealing someone’s personal information as a means of intimidation—presumably so that people more comfortable with the physical ad baculum show up at the target’s house.

On reflection, however, maybe this is ridiculous, or so it seems to many scholars of argument. Would it not, after all, be blindingly obvious even to the least competent of arguers that no amount of force is going to make someone believe something or even think that the threat is good grounds for the truth of some proposition? To start with, beliefs, by most plausible accounts, are involuntary and do not respond to commands, bribes, or other inducements (Cohen 1995; Woods 1998, p. 496). Besides, the threat of force, even if it were subtle as my implicature above, would seem to call attention to the lack of rational grounds for the arguer’s conclusion. Since fallacies are supposed, in some sense, to appear to be stronger arguments than they are (Hamblin 1970), it would be odd to have a fallacy where deception is not part of the picture. Indeed, there cannot be any, because if the threat is going to have any force as a threat, it needs to be recognized by its target for what it is. The problem, in other words, is not that the ad baculum is particularly uncommon, difficult to employ or difficult to detect. The problem is that it seems that what makes something a threat is conceptually and psychologically incompatible with it not being taken as a threat.

To appreciate this problem, it would be useful to be clear on what a threat is. As a speech act, a threat is akin to promising in that it commits the speaker to a course of action. “I’ll take you to lunch when the semester is over” is a promise. “I’m going to take your lunch” is a threat. They both specify things that the speaker will do, so theyare commissives. The threats of interest to ad baculum arguments also involve conditions, but in a particular way. “I’ll take you to lunch when I finish my paper” is a conditional promise; “When I get home, you’ll be sorry” is a conditional threat. Ad baculums, however, are also meant to direct the action of a target relative to the propositional content of the threat, and so they are what Blanco Salguerio calls directive–commissive speech acts (Blanco Salguerio 2010, p. 218). In addition to committing the speaker to a course of action (e.g., “I am going to get you”), threats of this type put the condition of the commissive on the target (e.g., “if you keep talking trash about me”). From all of this, it is pretty clear that, for a threat to make any sense as a threat, a few things need to be true: the target needs to understand (a) what the condition of fulfillment is (to stop talking trash in this case); (b) that the action is within their power to accomplish; and (c) that doing this action is the promised action of the one doing the threatening. It is fairly clear how threats work in the context of action, how they work in arguments is another matter.

One finds what I have above called “the standard picture” of the ad baculum in many popular critical thinking or introduction to logic texts. We will see one (Hurley and Watson 2018) in more detail below. By contrast, the comparatively extensive scholarly literature on the subject has essentially declared the standard picture to be dead; even the most ontologically generous accounts of the ad baculum barely leave it standing (Brinton 1992; Van de Vate 1975; Walton 2000). Many deny that the standard picture is any kind of argumentative scheme at all (or that, if it is a scheme, there is nothing logically wrong with it) (e.g., Wreen 1989). For pragma-dialectics, the ad baculum breaks a rule of comportment rather than any kind of logical rule (van Eemeren and Grootendorst 2004). More recently, Budzynska and Witek (2014) have made a more fruitful suggestion: we ought to look beyond the inference–schematic features of the ad baculum to appreciate it as a “complex rhetorical technique”, where the ethotic components of the person making the ad baculum argument are central to understanding how it works.

In this paper, I too plan to look beyond the inference–schematic features of the ad baculum to its ethotic features. My focus will differ from that of Budzynska and Witek. While they focus on the ethos of the speaker, I will instead focus on the ethos of the addressee of the ad baculum in an argumentative exchange. More specifically, given that something like ad baculum arguing is a common enough occurrence to merit investigation (i.e., people do use threats in arguments), my question is what makes it effective and therefore attractive as a move in argument contexts—i.e., where people are trying to get others to adopt standpoints (rather than, say, carry out actions). I will study examples of two main accounts: an informal–logical scheme account and Walton and Krabbe’s pragmatic account of the ad baculum as an instance of dialogue shifting (Walton and Krabbe 1995). The shortcomings of each approach point us in a different direction. Namely, the ad baculum is a curious case of intersection between two perspectives on the ultimate contents of arguments—commitments and beliefs. The ad baculum argument is a case where negotiated commitments can effectively be converted into reasons to believe something in an argument. We might call this argument laundering, because illegitimate reasons are effectively rendered clean to a clueless third party. Crucial to this is the understanding that the audience for an argument may not necessarily be present for the actual ad baculum exchange. In fact, in the particular ad baculum cases I shall be describing, they cannot be. So, two parties, A and B, have a dialogical exchange, where A induces B to refrain from speaking about p. For some further observer, C, at a later time, B’s not speaking about p appears to be evidence that p is not worth hearing. The ad baculum, therefore, is not exhausted in a dialogical exchange (as Walton had argued) between A and B, with A, for example, shifting the dialogue from argument to negotiation. Rather, the ad baculum requires A to threaten B, and B to succumb to the threat by corrupting the quality of the evidence of that exchange downstream for C, who does not realize that the conclusions of the dialogue between A and B were negotiated. In short: in employing the ad baculum, you don’t get someone to believe a conclusion, you get them to adopt it, so that other people believe it. My argument relies on the often-overlooked distinction (at least in argumentation studies) between beliefs and commitments (Paglieri and Castelfranchi 2006; Godden 2012, 2015). You can get someone to adopt a commitment through force because you adopt commitments voluntarily. Beliefs, by contrast, are involuntary, and mainly responsive to evidence. Understanding the difference between beliefs and commitments, then, shows how threats can be both understood and be successful in argumentative contexts.

2. The Ad Baculum as an Argument

I want to start with a well-known account of the ad baculum, Hurley and (new this edition) Watson’s widely used Concise Introduction to Logic, as of 2018, in its 13th edition. It is noteworthy that this account of the ad baculum, which has changed little across its many editions, has already been subjected to detailed (and scathing) critique in the scholarly literature (Wreen 1989).1 I do not mean to duplicate that work here. I think rather that there is something of value in Hurley and Watson’s attempt that the critique has missed. Besides, I think that, given what I take to be the general failure of the standard informal logic approach to the ad baculum, this account is as good as any other.

Hurley and Watson begin their discussion of the fallacies of relevance with the argumentum ad baculum, or the “appeal to the ‘stick’”. This fallacy, they write,

occurs whenever an arguer presents a conclusion to another person and tells that person either implicitly or explicitly that some harm will come to him or her if he or she does not accept the conclusion.

(Hurley and Watson 2018, p. 129)

People familiar with the Hurley and Watson text know that it views informal fallacies as commonly deceptive argument schemes where the premises do not provide adequate support for the conclusion. This means that they focus their analysis on the schematic features of the argument, how the premises fail or succeed in supporting the conclusion. This, I think, is part of the reason their account of the ad baculum is a mess. In the next section, we shall look at an alternative to this.

As Hurley and Watson describe the ad baculum, the reasons given for the conclusion are an implicit or explicit threat of some harm.2 This brings us back to our key question: how can we force or threaten someone into accepting a conclusion? Their answer to this is somewhat surprising. First, the examples.

Child to playmate: Sesame Street is the best show on TV, and if you don’t believe me, I’m going to call my big brother over here and he’s going to beat you up.

Lobbyist to senator: Senator Casey, of course you support our bill to reduce inheritance taxes. After all, you wouldn’t want the press to find out about all the contributions you receive from the Ku Klux Klan.

(Hurley and Watson 2018, p. 129.)

Before adding other comments, it ought to be said that these examples are not completely ridiculous. In my day, kids used to fight over which kind of food tasted best (or worst in some cases). Kids are not all that great at argument, but adults do not necessarily fare much better. I was once threatened (certainly only rhetorically I hope) in an argument over who liberated Italy in World War II (I held that it was the combined allied armies; my interlocutor held that it was the Italians by themselves). Beyond this, I have heard slightly more subtle threats employed at faculty meetings. So, it is at least plausible that people threaten each other this way. The question, however, is what they want to achieve. One can see that there is something of a confusion in this account as to whether one forces another to believe or to accept, or, as in the last case, merely to do something. The conclusion of the second example is, or at least seems to be, an action: Senator Casey is meant to do something by expressing support for some bill—probably voting or speaking in favor of it in the Senate Chamber. In contrast, the conclusion of the first example is that the playmate ought to believe something. These are certainly different things. They are also different from accepting something, which was mentioned in the introduction (more on this presently).

Now, let us turn to Hurley and Watson’s analysis. They write:

The appeal to force fallacy usually accomplishes its purpose by psychologically impeding the reader or listener from acknowledging a missing premise that, if acknowledged, would be seen to be false or at least questionable.

(Hurley and Watson 2018, p. 129)

This is a puzzling claim. Their general analysis of a fallacy of relevance is that the premises are psychologically but not logically relevant to the conclusion (p. 129). Take, for comparison, the ad hominem fallacy. On a standard account, what makes an ad hominem fallacious is that the personal failings or bad character of some arguer is not relevant to the non-character-dependent arguments they make. We think it is relevant, in other words, when it is not. In the case of the ad baculum, however, the appeal to force impedes recognition of the questionable premise, rather than seeming or appearing relevant when it is not. Hurley and Watson, somewhat admirably, hereby offer an attempt to explain why someone would be duped by an ad baculum: they are duped because they are afraid. If they were not afraid, one might imagine, then they would notice how bad the argument is. Fallacies, if they are going to be fallacies, ought to be deceptive. I am going to argue a bit later that the ad baculum is indeed deceptive. It is just not deceptive here in the dialogue where it occurs. Additionally, it is not even supposed to be. It would not work if it were so. Let us look at the reconstruction:

If my brother forces you to admit that Sesame Street is the best show on TV, then Sesame Street is in fact the best show.

If I succeed in threatening you, then you support the bill to reduce inheritance taxes.

(p. 130)

The reconstruction, they argue, makes it clear that the premise—which is the force—does not imply the conclusion. It is worth noting that the reconstruction is different in a few important ways from the original arguments. The original arguments had two different conclusions. In the first case, it was the belief that Sesame Street was the best show; in the second case, it was the willingness to support the inheritance tax bill.3 The term “support” is notoriously vague in this context. It can mean that one merely favors something (e.g., “I support the new President”) or that they will engage in certain actions (e.g., “I support my local animal shelter”). The ambiguity of the term makes this somewhat maddening. The first version of the Sesame Street case has the kid believing that it is the best show as the conclusion. I think, in other words, that Hurley and Watson have missed what is interesting about their cases. Part of the reason for this is their informal logic approach: the failure needs to be captured in a scheme, and the scheme is a sorry method for capturing this sort of failure. Another reason that they miss what is interesting about their cases is that they are unclear as to whether they mean to explain why ad baculum arguments are persuasive as a psychological matter, or how they fail as a logical matter. A successful account of the fallacy ought to explain why these failures match up—something that fails logically happens at the same time to psychologically succeed.

The main reason, however, is that they did not ask themselves what the cases are about. I mean, why would someone threaten someone in order that they believe something? That is psychologically impossible (or at least very difficult) and, as an argumentative matter, pointless. The Sesame Street bully, we have to imagine, must know this fact about beliefs and so have some other scope in mind. Given the limited nature of the example, it is hard to see what it might be. However, one thing that would make sense is that the addressee’s acceptance of the proposition has some kind of value. Even if the argument seems at one glance to be senseless, it is clearly the case that the Sesame Street bully is trying to communicate something. More than this, since threats are escalations, he is presumably trying to communicate something important. The question is whether this particular kind of communicative act has anything to do with arguments. Again, a standard critique of examples such as these (e.g., Wreen) is that they are perfectly fine (well, logically fine, but not necessarily morally so) means–ends pieces of reasoning; to try to shoehorn this into argumentation theory fails to grasp what threats really achieve.

I think, however, that there is something argumentative about threats. Our interest, after all, in having our conclusions accepted by others is not limited to those with whom we directly interact. Argument is a great way to spread the word. What is curious about the ad baculum, as the present cases might have shown on a more subtle analysis, is that we can convert non-reasons (in this case threats, though the same strategy works for bribes as well) into reasons. Call it argument laundering, because, like money illegitimately earned, a multistep process of erasing its criminal origin makes it appear legitimate.4 This is clearly the case with Senator Casey. Given the Senator’s standing in the community (or given our polarized politics, his standing in certain communities), there is epistemic value in his action of supporting (or not supporting) something. In other words, the fact that the esteemed Senator Casey deems some bill worthy of his support is a fact that an onlooker might take into account when thinking about what to think about it. The ad baculum employed against Senator Casey uses his ethotic standing on an issue to increase the credibility of the target proposition. The same could be said of the Sesame Street case. Perhaps the addressee’s support for the superiority of Sesame Street has some significance to Sesame Street skeptics on the playground. Knowing (or rather believing) that the addressee is a supporter might make the difference in their deciding which children’s show to spend their valuable and limited screen time on. They, after all, will not know about the forced nature of the commitment, so little Bobby expressing that he likes Sesame Street seems like reasonable ethotic evidence.

3. The Ad Baculum in Dialogue Theory

The standard account of the ad baculum, or so I have called it, fails for a lot of reasons. The additional reason I have given here is that it does not offer a plausible explanation for why anyone would try it in the first place. I propose that this argumentative purpose is to make their view more acceptable in the minds of another audience. However, as we have seen, it is difficult to see this in a schematic account such as the one offered by Hurley and Watson. We get a better view when we examine how threats work in the context of actual conversations. Walton and Krabbe’s (1995) concept of dialogue shift offers an interesting way to represent what is going on dialectically and pragmatically with the ad baculum. On their view, dialogues are normative models for conversation, i.e., models for how conversations ought to go, as defined by their purposes. Fallacies occur within this normative framework when participants illegitimately (i.e., without permission or warning) shift from one kind of dialogue to another (Walton and Krabbe 1995, p. 2). Crucially, dialogue shifting also explains the success of the fallacy deployment. Because fallacious moves are sometimes legitimate, an interlocutor may be duped into taking them to be valid when they are not. They think they are having one conversation when they are really having another. Critically, in its fallacious use, the shift is covert, unilateral, or not agreed upon. The heart of the deception, for dialogue theory, is to shift the context of the argument in a way that the interlocutor does not notice. The deception of another dialogue participant,which is a central part of the traditional account, therefore retains its place in this dialogical account. The interesting thing about the ad baculum, I shall argue here, is the way it involves the blending and confusion of dialogue purposes over time and space. So, roughly, A and B have one kind of dialogue, then B and C have another, on the basis of the original dialogue.

An enlightening comparison case to the ad baculum is what Walton and Krabbe call the “Fallacy of Bargaining”. This happens when one attempts “to replace an offer for an argument” (p. 104). In a very general sense, the fallacy of bargaining occurs when a critical discussion illicitly slips into a negotiation. Given Walton and Krabbe’s broad dialogical approach, the most direct way for this to occur is when one participant in a critical discussion demands of the other that they meet halfway, or compromise, on some standpoint. Walton and Krabbe’s example is as hilarious as it is improbable:

A doctor recommends that her patient quit both smoking and drinking, giving medical reasons for the recommendation.

Patient: O.K., I’ll quit smoking, as long as you allow a glass of wine once in a while.

(p. 104)

The patient’s offer to split the difference constitutes a dialogue shift. Their approach misunderstands or twists the purpose of a critical discussion, where the aim of each participant is to persuade the other of the truth or correctness of their position. In this case, the “medical reasons” signal that a critical discussion is at hand. The aim of the negotiation, in contrast, is to make a deal, and making concessions, such as the occasional glass of wine, is an example of that. Such moves would not be out of place in a negotiation over the price of something, such as medical care (at least in the United States). So, this seems like a clear instance in which an argument is suitable in one context but not in another.

While dialogue shifts by participants of dialogues within dialogues are certainly common, as we have seen, they are easily detectable by minimally competent participants. Such shifts are indeed often comically obvious, especially in this case. Imagine another medical case. This time, doctors Abel and Brava disagree about the correct diagnosis of a patient’s condition. Dr. Abel maintains that it has a bacterial origin; Dr. Brava argues that it is neurological. They cannot just split the difference—they cannot agree that it is a bacterial form of cancer, if such a thing exists. The nature of the discussion precludes that. Now, imagine Dr. Abel offering Dr. Brava a new set of golf clubs to go with her bacterial account. Or perhaps, arguing that the bacteriology department has fallen on hard times, and it would be a real boost to their spirits to have a new patient. Such offers are certainly shifts, and they might indeed even be unilateral and unagreed upon, but they could hardly go unnoticed.

Now, let us see how this dialogue shift account might work in the case of an ad baculum. Take the following (fictional) exchange between Frank Forthright, chief of compliance department at the Globex Corporation, and Assistant Divisional Chief, Mr. Malafide.

Forthright: Mr. Malafide, I’d like to show you some of my lab results. As you can see from the chart, there is a high presence of estradiol—known to cause deformities in frogs—in our plant’s waste water discharge. I have concluded that we are to blame for the recent deformities discovered in the frog population.

Mr. Malafide: I don’t agree with your reasoning, Frank. Left out of your analysis are the deep cuts we will have to make to this department if we have to comply with the law. Further, you’re not considering the financial hardship your family will face should this information get leaked to the public.

Malafide’s attempted ad baculum is, like all ad baculums of this sort, a strategy to change what ought to be an epistemic question about the cause of frog deformity into a practical discussion of Forthright’s future at Globex Corporation. What is crucial is that the success of this strategy relies on Forthright’s recognition of the changing context. If he did not notice the shift, a clueless Forthright might puzzle over how Malafide means to offer a meaningful objection and fail to see that Malafide intends to coerce him to take a particular course of action. Far from being an unannounced shift in dialogue, with the intent to fool its victim into taking the bait, the ad baculum is a patently obvious shift in dialogue.

Another feature of the exchange that reveals the obviousness of the offer-for-an-argument ploy is the actual target of the offer. As the challenge for the informal logic account above was how to explain how threats might be evidence of the fallacious kind in an argument (they cannot), here, the challenge is to explain how someone might be tricked or duped into negotiating a propositional commitment. The surprising answer, though one not appreciated by Walton and Krabbe, is that they can (even though, again, given the pragmatics of threats and promises, this is unlikely). Because, central to the conception of dialogue at issue here is the concept of commitment. Dialogues concern the commitments (or standpoints) of the participants. Commitments are not psychological entities and are freely adopted and abandoned, though they have logical properties, such as closure (Hamblin 1970, p. 264; Walton and Krabbe 1995, p. 21). According to Hamblin, once I have (voluntarily) incurred a commitment, I also acquire the commitments that are logically related to it (p. 264). Walton and Krabbe also remark that “[e]ach commitment implies many others”, which they call subcommitments (Walton and Krabbe 1995, p. 43). To cite their example, if one is committed to cooking dinner, then one is committed to cooking food (because all instances of dinner are instances of food, presumably) (p. 42). In terms of propositional commitment specifically, for example, if one is committed to the conjunction p & q, then one is also committed to p and q separately (by conjunction elimination); similarly, if one is committed to a universal proposition, such all ducks are birds, then one is committed to each individual duck being a bird (p. 43) (by subalternation). We can put this another way and say that commitments are closed under entailment. If I am committed to p, and p implies q, then I am committed to q.5 Without going too far into the weeds, there are several other interesting ways one can acquire commitments (and therefore subcommitments). These include the presuppositions required to ask an intelligible question (pp. 32–33). This would also include, though Walton and Krabbe do not note this, implicatures. If I ask where someone is from, I am bound by the implication that the answer is not obvious or is noteworthy. I also incur commitments by my affiliation with a well-organized group (e.g., political parties, unions, religious denominations, and so forth). Again, commitments are not psychological and so can be held corporately (p. 34). Commitments are central to the concept of a dialogue theory in part for this very reason. You can move them like pieces in a game. Most significantly for my purposes here, you can adopt commitments that you do not believe. In other words, the notion that commitments can be traded freely in an argument is a feature, not a bug, of dialogue theory.

By contrast, no party to a dialogue about beliefs is able to negotiate them, because beliefs are involuntarily held. While it is true that beliefs can be occasioned by deviant causal chains, where they are brought about indirectly, in general, beliefs track reasons, or at least the appearance of reasons. In trading an offer for an argument, one is not tracking reasons, and this is plainly obvious. Mr. Malafide knows that Forthright is not going to change his mind. It does not matter anyway, because all Mr. Malafide needs is that Forthright change his public commitments. This would mean that Forthright ceases to claim that the frogs have been poisoned by Globex and, among other things, that he stops using this claim as a premise in other reasoning (Cohen 1995, p. 4).6 The upshot of this is that, while Malafide cannot directly cause Forthright to change his belief with an inducement or threat (or a bribe—why did not he try that?), he can more directly control the beliefs of others further down the conversational chain. In the present case, it is likely that people who become aware of the Globex corporation’s malfeasance will respond accordingly. However, if Malafide deprives them of the opportunity to respond to the evidence, then he has effectively, though indirectly, affected their beliefs. In effect, Malafide’s ad baculum is not directed at Forthright so much as it is at other potential participants in their extended dialogue. It is directed at them by excluding them. Crucially, they are not observers or witnesses, as the ad baculum would then prove equally ineffective.

4. Conclusions

In this paper, I have approached what I take to be an underappreciated problem with the ad baculum fallacy, namely the fact that you cannot force an interlocutor to believe something. The persistence of ad baculum cases, however, cries out for some kind of explanation. I have argued that the ad baculum stratagem should be understood in light of the broader purposes of argument. This means that the ad baculum actually involves three parties: someone (A) who invokes a threat in the context of an argument; a respondent (B) to whom the threat is directed; and an audience (C), not present at the exchange between A and B. While the respondent can never take force as inducements to believe p, they can take them as reasons to commit pragmatically to p. Additionally, it is not at all difficult or puzzling to see how B can accede to the threats in the first place, though it is difficult to imagine that they could be oblivious to them as threats. Their commitment to p may then be seen by their audience as evidence for p—the fact that Senator Casey supports the bill—or perhaps merely does not oppose it—is ethotic evidence to a certain audience that the bill is worthy of consideration.

This result tracks some recent work in fallacy theory. Consider, for example, the straw man fallacy, where one distorts the standpoint of another in order more easily to refute it. Here, too we have what seems to be an impossible argument strategy on an informal–logical or dialogue scheme account, since few would be convinced by a distorted version of their own view (Aikin and Casey 2022a; Stevens 2021). Like the ad baculum, the straw man may serve merely to make argument, with the straw manner being unpleasant, and so drive the target from the discursive space. However, there is more to straw manning (and the ad baculum) than this strategy of annoyance. In a paper on the empirical effectiveness of straw man arguments, Bizer et al. (2009) studied downstream audiences rather than parties to a debate. This is because the straw man is most often deployed as being for the benefit of an onlooking audience, and so it is there that we must look for its effectiveness (de Saussure 2018). While the ad baculum must be deployed in the second person, the real target of the ad baculum is a third-person, downstream audience whose total evidence will be constrained by the forced adoption of a commitment.

A second consequence is that force, threat and/or sanction are not operative features in this scheme. This three-party scheme works just as well with bribery or arguments ad carotam, as Bermejo-Luque has suggested (Bermejo-Luque 2008). The key fact this that someone can voluntarily trade their commitment to some proposition to avoid harm (in the case of the ad baculum) or for some gain (in the case of the ad carotam). This is why Walton and Krabbe assimilated the ad baculum arguments to “the fallacy of trading an offer for an argument”. Both of these schemes work by converting an illegitimately adopted commitment into evidence for some third party.

The expanded conception of the ad baculum has another consequence. The standard account of the ad baculum, as I have called it, focuses on the irrelevance of the threat of force to the truth or acceptability of some claim. This is certainly true in a very general sense, but the broader meta-argumentative goal of the ad baculum is to misrepresent the dialectical state of play. It means to give the impression that a certain standpoint has more (or fewer) true adherents than it actually does. A broader view of the ad baculum shows the inadequacy of the two approaches we have discussed to represent it. As Wreen has shown, if you consider the ad baculum as an argument scheme, you may miss what is attractive about it as a fallacy (see also Budzynska and Witek 2014). It will always turn up as a perfectly reasonable means–ends piece of reasoning. Walton’s dialogue approach fares no better at capturing it, because the real target of the ad baculum is not a party to the dialogue.

In the end, argument analysis in informal logic suffers oftentimes from what we might call methodological individualism. Even dialogue theory, with its deeper reach into the structure and purpose of persuasive communication, stops its analysis at the end of individual argument encounters. Argumentative exchanges are not necessarily exhausted when the addressee has received the message. To explain the effect of the ad baculum, it would have to construct a new encounter, where the defect is the insincere commitment—the ad hominem does just this. However, that does not do justice to the strategy of the one who employs the ad baculum. I think our arguments are fundamentally meant to outlive our encounters. Additionally, I do not think that this is unique to the ad baculum. Scott Aikin and Robert Talisse have made a similar case with regard to other fallacies (such as the ad hominem), arguing that even the dialogical model, itself an expansion, incompletely represents the dialectical situation (Aikin and Talisse 2019). With the straw man, for example, the purpose is to misrepresent the quality of an addressee’s argument to an onlooking audience. In contrast to the ad baculum, however, it is most effective when the addressee is absent and so not able to defend themselves. For the ad baculum, a critical part of the strategy, however, is that the target audience is not there.

This expanded view of argumentation has one further feature, which I note in closing. Arguments can involve a complex interplay of different ways of conceiving of the fundamental objects of argumentation. As we have seen, pragmatic theories of argumentation, such as dialogue theory, tend to view arguments in terms of pragmatic commitments or speech acts. Informal logic, by contrast, typically understands arguments to concern propositional attitudes, or beliefs. I have argued here that neither of these approaches alone does justice to argumentative cases of the ad baculum fallacy. To explain how such a fallacy works, one needs to view argument from both perspectives at once.