No Place for Compromise: Resisting the Shift to Negotiation

No place for compromise: Resisting the shift to negotiation David Godden and John Casey* David Godden Philosophy Department Michigan State University East Lansing, Michigan USA 48824 Email: dgodden@msu.edu www.davidgodden.ca John Casey Philosophy Program Northeastern Illinois University 5500 N. St. Louis Ave, Chicago, IL USA 60625 Email: j-casey1@neiu.edu Godden, D. and Casey, J. (2020). No place for compromise: Resisting the shift to negotiation. Argumentation: An International Journal on Reasoning http://dx.doi.org/10.1007/s10503-020-09517-z This is a post-peer-review, pre-copyedit version of an article published in Argumentation. The final publication is available at link.springer.com http://link.springer.com/article/10.1007/s10503-020-09517-z ABSTRACT: In a series of recent papers beginning with their “Splitting a difference of opinion: The shift to negotiation” (Argumentation, 2018, 32: 329-350) Jan Albert van Laar and Erik Krabbe claim that it is sometimes reasonable (i.e., rationally permissible) to shift from a critical discussion to a negotiation in order to settle a difference of opinion. They argue that their proposal avoids the fallacies of bargaining (substituting offers for arguments) and middle ground (mistaking a compromise for a resolution). Against this permissive policy for shifting to negotiation, we argue that the motivating reasons for such shifts typically fail, and that the permissive policy avoids neither fallacy while structurally incentivizing two types of strategic maneuvering that constitute rational and argumentative hazards: argumentative overcharge and abandonment of discussion. KEYWORDS: abandonment of commitment; abandonment of discussion; argumentative overcharge; critical discussion; dialogue shift; discursive norms; fallacy of bargaining; fallacy of middle ground; negotiation dialogue; principle of retrospective evaluation; rational resolution; * Acknowledgements: We consider this joint work; our names are listed in reverse alphabetical order. Earlier versions of this paper were presented by David Godden to the 11th International Conference of the Ontario Society for the Study of Argumentation (OSSA), “Argument, Objectivity, and Bias,” May 18–21, 2016 (which appears in the proceedings (Godden 2016) as a commentary on van Laar and Krabbe (2016a) and to which they replied in their (2016b)), and to the Center for Research in Reasoning, Argumentation, and Rhetoric at the University of Windsor, November 2, 2018. In addition to the audiences at those talks, we offer our thanks especially to Erik Krabbe, Jan Albert van Laar, Harvey Siegel, and Douglas Walton for their incisive and constructive comments on earlier drafts of the paper. 1 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION 1 Introduction Suppose we disagree about the rational acceptability of a standpoint—e.g., the reasonableness of a truth claim, the probative value of putative reasons or evidence for a claim, or the rational merits of a policy proposal. There are many things we might do when reaching such an impasse. At times it is prudent to ignore disagreement. If we are confident that our position is correct and the reasons informing the incompatible view neither defeat nor outweigh our own, there is little impetus, practical or epistemic, to reconsider or refute objections or alternative standpoints that we have already taken account of. So, if the success of our respective endeavors does not depend on the cooperation or agreement of the other, we might just walk away and go on about our business. The practical value of getting on with things might outweigh the cost of having one (more) disagreeable soul in the universe. If we must settle the matter, there remain several courses of action available: we might throw dice, draw straws, or consult a Ouija board. Suppose that we resist agreementgenerating strategies like these on the grounds that they are, at best, only accidentally related to producing a correct, rationally defensible result. We might then argue. That is, we might engage in a critical discussion in order to establish the reasonableness of the standpoints at issue (van Eemeren and Grootendorst 1984; 2004). Alternatively, we might negotiate. That is, we might make a series of “trade-offs”—or conditionalized demandconcession pairs of the forms, “If you grant p, I will concede q,” “If you agree to r, I will agree to s,” “If you will do X, I will do Y,” etc.—until we have reached a mutually agreeable result. This paper concerns the reasonableness of shifting to negotiation in order to achieve agreement when argumentation has failed to yield a resolution. In a series of recent articles (2018a, 2018b, 2019a, 2019b; van Laar 2019) beginning with “Splitting a difference of opinion: The shift to negotiation” (Argumentation, 2018a)1 Jan Albert van Laar and Erik Krabbe study an important type of dialogue shift, namely that from a critical discussion about a policy proposal (where critical and justificatory argumentation is exchanged about the rational acceptability of action or policy options proposed by various parties) to a negotiation dialogue (where agreement is sought via a series of compromises, or trade-offs, on the part of each side in a disagreement). Minimally, van Laar and Krabbe may be read as claiming that it is sometimes reasonable—i.e., rationally permissible—to shift from a critical discussion (a persuasion dialogue) to a negotiation in order to close a difference of opinion.2 [W]ithin a persuasion dialogue, participants may dispose of and express good reasons for shifting towards a negotiation dialogue in which they cooperate to settle their 1 All citations of van Laar and Krabbe will be from this (2018a) paper unless otherwise indicated. For present purposes, we may treat ‘persuasion dialogue’ and ‘critical discussion’ as roughly synonymous. Properly understood, a critical discussion is an argumentative dialogue conforming to, and governed by, the normative procedural system devised by pragma-dialecticians (van Eemeren and Grootendorst 2004). While a critical discussion has the procedural form of a persuasion dialogue, it may be used for other, nonpersuasive, dialectical purposes, e.g., inquiry (by allowing discussants to take on assigned or hypothetical, rather than actual, opening commitments), or deliberation (by changing the kind of standpoint being discussed). 2 2 DAVID GODDEN AND JOHN CASEY difference of opinion by way of compromise. Splitting a difference of opinion may be a sensible idea, and the shift towards negotiation can be warranted … . (348) Let’s call this a permissive policy for shifting to negotiation. Much of van Laar and Krabbe’s recent work on argumentation and negotiation may be read as offering a groundbreaking account of the role of argumentation within negotiation (2018b, 2019b) and the operation of argumentation about (e.g., criticizing and justifying) particular negotiated compromises (2019a). We do not take issue with these important aspects of their work. Rather, our argument specifically concerns their permissive policy, and the propriety and justifiability of dialogue shifts from critical discussions to negotiations. In making their case for “splitting a difference of opinion,” van Laar and Krabbe argue that the fallacies of bargaining—“substituting offers for arguments” (340; see our §2.5)— and middle ground—mistaking a compromise for a resolution (346; see our §2.7)—are avoided by their proposal, thereby making it unproblematic. We disagree. Van Laar and Krabbe contend that, because they distinguish compromises from resolutions, their proposal avoids the fallacy of middle ground. Yet, we find that their proposal treats negotiated compromises as discursive proxies for resolutions in some subsequent talk and action—specifically in contexts of implementation—allowing the negotiated compromise to do the same (extra-)discursive work as the (unachieved) resolution would have done. Moreover, seemingly because they take the discussants in a shifted-to negotiation to have avoided the fallacy of middle ground, van Laar and Krabbe do not apply the principle of retrospective evaluation (Walton and Krabbe 1995: 102-103; see our §2.4) to their proposal when making their case for it. Yet, we argue that their proposal does not satisfy the principle of retrospective evaluation. Shifted-to negotiations do not proceed according to the discursive standards of a critical discussion and nor are they capable of satisfying its goals, properly understood. Further, by separating the discursive activity of negotiating from critically discussing, van Laar and Krabbe claim that their proposal avoids the fallacy of bargaining. Yet, we argue that the permissive shifting policy permits the substitution of a preference-tracking procedure, consisting of sequences of trade-off-responsive moves, for a reason-tracking procedure, consisting of sequences of reasons-responsive moves, and thereby permits the substitution of offers for arguments. Additionally, the fallaciousness of bargaining is partly explained by the fact that, in shifting to negotiation, critical discussants abandon their propositional commitments together with the rational and discursive responsibilities attending to them, thereby committing the fallacy of abandonment of discussion / commitment (Walton and Krabbe 1995: 110; see our §2.6). Again, seemingly because they take discussants following their policy to have avoided the fallacy of middle ground, van Laar and Krabbe never discuss the fallacy of abandonment of commitment and seem to take it to also have been avoided. Yet, we find that following their policy can involve a fallacious abandonment of commitment, particularly the obligation to defend one’s viewpoints and to retract them should that defense be unsuccessful. Finally, we find that van Laar and Krabbe’s permissive policy for shifting to negotiation structurally incentivizes two types of strategic maneuvering that constitute rational and argumentative hazards: argumentative overcharge and abandonment of discussion. Combined, we fear that these consequences of adopting a permissive policy for shifting from critical discussion to negotiation—abandoning reason- and truth-tracking procedures for 3 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION preference-tracking procedures, permitting critical discussants to abandon their rational and discursive responsibilities, and incentivizing the rationally hazardous behavior of argumentative overcharging—contribute to the subversion of the very activity of rational argumentation and the goods delivered therefrom. In outline, the paper proceeds as follows. We begin, in §2, by introducing the theoretical framework, relevant concepts, and fallacies that inform the discussion to follow. Next, in §3, we briefly summarize van Laar and Krabbe’s case for their permissive policy for shifting to negotiation. In §4 we identify, locate, and circumscribe what we take to be the locus of the normative problems with the permissive policy. There we argue that van Laar and Krabbe’s principal reason (see §3.2) fails to properly motivate any dialogue shift from an unresolved critical discussion. Our main critical arguments are offered in §5, where we claim that following the permissive policy commits the fallacy of middle ground, fails to satisfy the principle of retrospective evaluation, and commits the fallacies of bargaining and abandonment of discussion / commitment. In §6 we consider the rational and argumentative hazards of argumentative overcharge and abandonment of discussion that we take to be structurally incentivized by a permissive policy for shifting to negotiation, and consider the practice of overcharging, common in the North American legal system, as a real example. Our conclusions are offered in §7. 2 Preliminary considerations 2.1 Types of dialogue and dialectical norms In the dialectical normative framework adopted by van Laar and Krabbe, a situated occurrence of argumentative conversation may be classified as a dialogue of one type or another according to characteristic features like its initial situation, dialogue goal, participant aims, etc. (Walton and Krabbe 1995: 65ff.; Walton 1998). For example, van Laar and Krabbe characterize persuasion dialogue as: the kind of dialogue that starts from a difference of opinion … in which the participants try to convince one another of their standpoint by appealing to reasons that the other side is willing to concede [i.e., accept], doing so in a cooperative endeavor to resolve their difference of opinion on the merits of both sides. (332) By contrast, a negotiation dialogue is described as: a dialogue that starts from a difference of interests or from a difference of opinion … in which the participants aim at getting as much as possible what they want, doing so in a cooperative attempt to strike a deal. (332) Dialectical theories of argumentative norms then seek to leverage the overall goal of a type of dialogue as a source of normativity when (i) setting out permissible, obligatory, and prohibited proceedings (e.g., conversational moves) that can occur in them and when (ii) qualitatively characterizing the end states of dialogues of that type (i.e., evaluating the results of the dialogue). Thus, in dialectical theories, rational norms are articulated as 4 DAVID GODDEN AND JOHN CASEY procedural norms for the proper (i.e., reasonable) conduct of a discourse. 2.2 Resolving versus settling disagreements Following van Eemeren and Grootendorst (2004: 58) (cf. van Laar and Krabbe 332f., 342f.) we may distinguish resolving a difference of opinion from merely settling a dispute. Resolution is the goal of critical discussion. A dispute is settled when, by mutual consent, the difference of opinion has in one way or another been ended—for example, by taking a vote or by the intervention of an outside party who acts as a judge or arbitrator. … A difference of opinion is only resolved if a joint conclusion is reached on the acceptability of the standpoints at issue on the basis of a regulated and unimpaired exchange of arguments and criticism. (van Eemeren and Grootendorst 2004: 58) While ‘settled’ names a descriptive property of end states of disagreements, ‘resolved’ is a success term. Resolution is the normative end state of a successful critical discussion where either (i) a proponent retracts a standpoint because they failed to successfully defend it from a respondent’s criticisms, or (ii) a respondent retracts their critical doubts, and thereby accepts a standpoint, on the basis of the argumentation offered by its proponent in response to the critical reactions of the respondent (van Eemeren and Grootendorst 2004: 133). The aim of critical argumentation—indeed a motivating reason to engage in it, rather than some other dispute settlement procedure—is to achieve resolution, not merely agreement. That a disagreement has been resolved, rather than merely settled, entails that the rational acceptability of the standpoint under discussion has been properly determined as a result of successfully engaging in a process of critical argumentation. 2.3 Negotiation and the nature of compromise While resolution is the goal of critical discussion, compromise is the goal of negotiation. That is, ‘compromise’ (in its technical sense) names an inherently normative classification of the end state of successful negotiations. Building from Weinstock’s (2013: 539) definition, van Laar and Krabbe define ‘compromise’ as follows: a compromise is a kind of agreement that concerns an arrangement that for each competing party is inferior to their initial positions yet which they have reason to accept in the social setting at hand. (343; cf. 2019a: 94-95) The discursive means by which a compromise (in the technical sense) is achieved, and by which negotiation is conducted, is via each discussant making compromises (in the ordinary sense)—that is, aspects of discussants’ initial, preferred positions are “traded-off” or “bargained away” in order to reach some mutually agreeable end state. For clarity, following van Laar and Krabbe, we may call these ordinary compromises ‘concessions.’ And, we will distinguish concession, understood as a trade-off-responsive move, from retraction and acceptance understood as reasons-responsive moves. Making concessions, then, is partly constitutive of the activity of negotiating and necessary for achieving 5 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION compromise. In van Laar and Krabbe’s words: “No compromise is without mutual concessions” (343). Van Laar and Krabbe (342f.; cf. 2018b 551f.) proceed to identify five characteristic features of compromises, some of which bear on our subsequent discussion. (i) Compromises are not resolutions. Compromises involve mutual concessions, and are different from resolutions because “Strictly speaking, a compromise does not end the disagreement” (343; quoting Benjamin 1990: 7). (ii) Compromises are the result of free choices of discussants, rather than being imposed or coerced (343-344). (iii) Compromises result from quid pro quo exchanges and thereby involve a kind of “commodification” of one’s opinions. (iv) Compromises involve both action commitments and propositional commitments on the part of the discussants. Central among the latter is that an action or policy proposal reached through compromise in negotiation “merits acceptance and implementation,” “all things considered” (344; emphasis added). More specifically, the justification of a compromise requires satisfying the condition of eligibility—that the compromise agreement is preferable to continued disagreement—and optimality—that no better compromise is available (344-345; cf. 2019a: 100f.).3 Lastly, (v) while compromises do not resolve first-order disagreements, they may “yield” resolutions to related, secondorder disagreements, e.g., a “second-order disagreement … about what method to select for dealing with the irresolvable first-order issue,” and a “second-order disagreement … about precisely what compromise solution to select for dealing with the first-order problem” (345).4 2.4 Dialogue shifts: Permissible and fallacious A dialogue shift occurs when, during the course of a conversation, the kind of discussion changes from one type of dialogue to another (Walton and Krabbe 1995: 100ff.; Walton 1998: 198ff.).5 Douglas Walton and Erik Krabbe have developed two criteria for judging the permissibility of a dialogue shift. The first criterion applies the conditions of transparency—that that the shift is “announced … in an appropriate way” such that the transition is not “concealed by one party, so that it functions as a tactic of deception against the other party”—and intersubjective validity—that “it [the shift] is properly agreed upon 3 In (2019a: 100f.) van Laar and Krabbe set forth their two-premise second scheme for the Defense of a Negotiated Compromise wherein, roughly, the progress premise asserts that the eligibility condition is satisfied while the optimality premise asserts that the optimality condition is satisfied. 4 We note, in passing, that a compromise settlement of a first-order disagreement does not “yield” a resolution of a second-order disagreement as to the proper method to select for dealing with a(n irresolvable) first-order disagreement; rather the compromise, indeed the negotiation itself, presupposes such a resolution. Relatedly, while negotiation may result in a compromise that in turn supplies part of the content of a resolution to a second-order argument about what compromise solution to select for agreement in settling a first-order disagreement, it cannot “yield,” or “bring about” (2109a: 94), such a resolution. Rather that resolution must be the result of second-order argumentation, which at some point cites the compromise of the first-order negotiation, together with reasons that negotiation is a legitimate first-order dispute-settlement mechanism in this circumstance. In order to avoid the fallacies of middle ground and bargaining, van Laar and Krabbe are committed to saying not only that compromises are not resolutions, but that bargaining cannot produce, result in, “yield,” or “bring about” resolutions (see our §3.3). 5 See Chris Tindale’s (1997) “Fallacies, blunders, and dialogue shifts: Walton’s contributions to the fallacy debate” for a survey of these topics. 6 DAVID GODDEN AND JOHN CASEY by the participants” (Walton 1998: 216; cf. Walton and Krabbe 1995: 115). Van Laar and Krabbe (335f.) stipulate that the dialectical shifts endorsed by their permissive policy satisfy both of these conditions by being properly announced and agreed upon. If satisfying the conditions of transparency and intersubjective validity were alone sufficient to warrant a dialogue shift, then trial by ordeal or trial by combat might remain permissible alternatives to trial by judge and jury, so long as the participants understood and agreed. Yet, recognizing that the outcomes of ordeal and combat have only an accidental, if not entirely biased and superstition-based, connection to the aims of justice and truth central to any judicial process, it must be acknowledged that these ends are not served by ordeal or combat. As such, a further condition for permissible dialogue shifting is required. The second criterion for permissible dialogue shifting, which van Laar and Krabbe do not discuss, applies a principle that, following Walton (1998: 214), we will call the principle of retrospective evaluation. To judge whether a shift is licit or illicit in a particular case of argumentation, we first have to pin down the original context of dialogue. Then we have to identify the new context, and decide whether the shift is licit or illicit by looking backwards and judging by the goals and standards of the original context. Is the new dialogue supporting those old goals, or at least allowing forward movement on their fulfillment, or is it blocking them? (Walton 1998: 201 emphasis added, cf. 214; cf. Walton 1992: 139; Walton and Krabbe 1995: 102-103) The principle of retrospective evaluation seems designed, in part, to ensure that the end state of the new dialogue can, in subsequent discussion, properly occupy a discursive place and role equivalent to that which the conclusion of the old dialogue would have occupied (had it been achieved).6 Since, as we already noted, dialectical theories derive norms of reasonableness partly from the collective aim of the activity type (e.g., the kind of argumentative dialogue) in which discussants are engaged, changes in the type of activity will result in concomitant changes to the norms of reasonableness that properly apply to discussants and their conversational moves. The principle of retrospective evaluation acts as an evaluative signpost and regulative guardrail for such discursive lane-changes. First, it prescribes that the end state of the new dialogue should be judged according to how well it satisfies the goals of the original dialogue according to the standards of that original dialogue. Relatedly, moves in the new, shifted-to dialogue are to be evaluated according whether, and the extent to which, they adhere to those standards and contribute to attaining the goals of the original, shifted-from dialogue. To the extent that they do, commitments gained and undertaken, and retractions made, in the new dialogue will have the same rational, discursive status and role as commitments and retractions in the old dialogue. 6 By “subsequent discussion” we mean to include both conversation that follows the initial dialogue (say a deliberation that follows a persuasion dialogue that either included an imbedded inquiry or that shifted to an inquiry) and conversation that returns to initial dialogue (say, a return to a persuasion dialogue following an imbedded, or shifted-to inquiry). 7 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION Importantly, the principle of retrospective evaluation presupposes the propriety of the initial dialogue. In cases where it is either illegitimate or redundant, the principle of retrospective evaluation is misplaced. This observation makes visible a class of trivially acceptable shifts whereby discussants somehow “find their way” to a proper dispute settling mechanism. For example, consider a conversation that begins eristically but changes to a critical discussion. Here, we might suppose that the dialogue ought to have, or could just as well have, started as a critical discussion. In such cases, concern does not arise from the dialogue shift, but from the initial dialogue itself which, at best, would be redundant or a false start, and at worst could disorient or even derail the conversation such that it does not lead to an appropriate dispute settlement or means thereto. Both the fallacy of bargaining and the fallacy of middle ground may be understood as violations of the principle of retrospective evaluation. And the fallaciousness of each may be explained by reference to the fallacy of abandonment of position (or commitment). 2.5 The fallacy of bargaining Since, on dialectical accounts, argumentative norms derive, in part, from the goal of a dialogue, and since the goals and standards of a critical discussion are different from those of a negotiation dialogue, dialogue shifts can be the site of—indeed, can constitute— fallacious argumentation. To the extent that the goals or standards of dialogues of different types do not align, the normativity at work in a dialogue of one type may be “unfit” for dialogues of other types. This helps to explain the principle of retrospective evaluation, which seems designed to detect and prohibit normative mismatches of just this kind. Shifts from critical discussions to negotiations are specifically susceptible to the fallacy of bargaining, which may roughly be characterized as “the fallacy of substituting an offer for an argument” (Walton and Krabbe 1995: 104). As an example of this kind of fallacious discursive maneuvering, Walton and Krabbe consider the ad baculum (appeal to threat or force) as a move in argumentative discourse. (Here, we might understand the threat as an “offer” of a negative consequence—of a disincentive rather than an incentive.) During a negotiation type of dialogue, threats and appeals to force or sanction are characteristic. However, if the context is supposed to be that of a critical discussion, the same kind of argumentation which was appropriate enough in the negotiation context can become highly fallacious. (Walton and Krabbe 1995: 109; cf. Walton 1992: 141) Importantly, van Laar and Krabbe explicitly permit moves of this kind, which they call “pressuring,” in negotiation (343-344; 2019b). Walton and Krabbe cite two principles in explaining the fallaciousness of appeals to threat in critical discussions. First, threat appeals can be used to suppress an interlocutor’s conversational contributions, thereby violating the freedom rule (van Eemeren and Grootendorst 2004: 136). In this context [i.e., a critical discussion], the use of the ad baculum argument is always highly suspicious and tends to be at odds with the discussion, because it is a way of trying to force closing off the free expression of one’s point of view that is necessary 8 DAVID GODDEN AND JOHN CASEY for critical discussion. (Walton and Krabbe 1995: 109; cf. Walton 1992: 141) Discursive contributions that result in violations of the freedom rule do not adhere to the standards of a critical discussion. They further derail such discussions by obstructing, if not preventing, the attainment of the dialogic goal. Thus, permitting such moves in a shifted-to dialogue would fail to satisfy the principle of retrospective evaluation. Walton and Krabbe (1995: 110) offer a second explanation of the fallaciousness of bargaining in terms of its being a special case of the fallacy of abandonment of discussion, to which we now turn. 2.6 The fallacy of abandonment of discussion (or commitment) Substituting offers for arguments is an instance of a more general fallacious move by which one abandons one’s discursive commitments, thereby violating the obligation to defend rule (van Eemeren and Grootendorst 2004: 138f.). The fallacy of bargaining … together with other fallacies that substitute something else for argument, is a special case of the fallacy of abandonment of discussion (i.e., of persuasion dialogue). What these fallacies have in common is that whoever commits one of them reneges on his original propositional commitments (burden of proof, etc.) in persuasion dialogue by moving to another type of discourse. (Walton and Krabbe 1995: 110; emphasis changed) In advancing standpoints, critical discussants incur obligations to defend those standpoints from expressions of relevant doubt and criticism. A central dialectical standard by which the rational acceptability of a standpoint is measured is the extent to which its proponents succeeds in defending it from critical doubt and objection. Generically, this standard may be called the proponent’s burden of proof. Operationalizing this standard gives rise to the obligation to defend. And discharging one’s obligation to defend demonstrates that one criterion for this standard has been satisfied. Yet when critical discussants switch to negotiation they thereby fail to discharge their obligation to defend. By “commodifying” their standpoints and subsequently engaging in trade-offs of components of the viewpoint to which they are committed and take themselves to be rationally entitled for elements of a view to which they have no commitment and no claim to rational entitlement, critical discussants abandon their obligations to defend their commitments. And doubly so: in making a concession in a shifted-to negotiation a critical discussant (i) abandons some claim that they are obliged to defend in exchange for (ii) adopting some other claim as a commitment which, because of how they arrived at it, they are not in a position to defend—i.e., which they are not entitled to accept. When judged against the standards of the original critical discussion, such discursive moves patently fail to measure up. Moreover since the goal of a critical discussion is not merely to generate agreement, but to do so on the basis of critical argument about the rational merits of a standpoint, the goals of a critical discussion are similarly unfulfilled by shifting to negotiation. Because of this, it is not merely particular discursive commitments that are abandoned, but indeed the entire critical discussion itself. The fallaciousness of 9 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION abandonment of discussion (commitment) may thereby be explained in terms of the principle of retrospective evaluation. 2.7 Fallacy of middle ground The second fallacy discussed by van Laar and Krabbe is the fallacy of middle ground: “We label the mistake of presenting a compromise as a regular resolution, either to oneself or to others, the Fallacy of Middle Ground” (342). As should be clear from our preceding discussion, to treat a compromise as a resolution clearly fails to satisfy the principle of retrospective evaluation. The standards for reasonable (e.g., permissible, obligatory) acceptance and retraction do not align with the standards for reasonable concession; they cite different features of the attendant claims. Relatedly, the goal of a critical discussion cannot be achieved independently of satisfying those standards, since the final acceptability of a standpoint is to be determined on the basis of the critical argumentation transacted by the discussants. Put simply, if compromises are only achieved via a sequence of negotiated concessions (understood as a trade-off-responsive moves), while resolutions are only achieved via sequences of argued retractions and acceptances (understood as reasonsresponsive moves), then one cannot negotiate one’s way to a resolution. 3 Summarizing the case for splitting a difference of opinion In making their case for the permissibility of splitting a difference of opinion, van Laar and Krabbe (331ff.) address two questions concerning shifts from persuasion to negotiation dialogues: What legitimate reasons might there be for such shifts? And, under what conditions would such shifts be permissible? In answering these questions, they seek to formulate a permissive policy framework for shifting to negotiation that is sensitive to, and avoids, the fallacies of bargaining and middle ground. We begin with their answer to the second question. 3.1 Preliminary discursive conditions for permissible shifting In making their case for the permissibility of shifting to negotiation, van Laar and Krabbe place some fence posts around the discursive circumstances in which they claim that their policy may properly be applied. The first limits the focus7—though not the scope8—of the kinds of standpoints that are conducive to shifts to negotiation: namely, disagreements about action prescriptions or policy proposals. “Our account of splitting differences of opinion deals only with situations in which the participants have to make a practical decision” (341). The second marks a dialogue state which, they claim, “most naturally gives rise to a shift towards negotiation” (334). The circumstance is one where a critical discussion has become deadlocked because, despite each party having made “serious 7 They write: “In this paper, we are particularly interested in dialogues about an action proposal” (331). They write: “That a disputed proposition is factual or theoretical does not and should not keep parties from adopting a compromise” (341). 8 10 DAVID GODDEN AND JOHN CASEY attempts to convince the other,” neither party retracts any element of their opening position (335).9 Call this participant stance giving no quarter, and define it as: Giving no quarter The party (i) maintains its own, original standpoint (in the case of a mixed dispute), and (ii) maintains its critical stance towards the standpoint of the other party. (334; our paraphrase) When each party gives no quarter in a critical discussion van Laar and Krabbe claim: When it has become obvious that a persuasion dialogue only leads to a stalemate, it is less opportune to start another persuasion dialogue and perhaps time to look for other ways to make progress. Here the negotiation dialogue is an obvious candidate. (334) 3.2 Reasons to shift In situations of no quarter, van Laar and Krabbe claim that participants may have good reasons motivating them to quit the persuasion dialogue as fruitless. The principal reason for shifting away from persuasion dialogue … is the expectation that the persuasion dialogue’s common goal of resolving differences of opinion is not realizable for the issue under discussion and that consensus is therefore out of reach. (336) Call this van Laar and Krabbe’s principal reason for shifting, and note that, to the extent it is a reason at all, it is only a reason for shifting from a critical discussion, not a reason for shifting to a negotiation. The principal reason does not recommend any one alternative dispute settlement procedure over others. Why, as they ask (336), negotiate rather than take a vote or throw dice? In answer to their first question, then, van Laar and Krabbe (337f.) identify five motivating reasons that parties might have, or offer, for moving from a persuasion dialogue to a negotiation and for accepting a negotiated compromise in place of the continued dissensus resulting from a failed persuasion dialogue. Briefly, they are: 1. The strategic argument: “By arguing you achieve nothing. By negotiation you achieve at least something” (337). 2. The argument for principled consequentialism: “It would just be wrong to neglect the possibility of realizing at least part of your aims” (338). 3. The epistemic argument: “Neither of us can be sure of having a complete view of all aspects of the situation and a deal might do justice to a wider array of considerations” (338). 4. The democratic argument: “Making a deal with contributions from both sides is more democratic than a one-sided solution” (339). 9 It is worth noting that they do not stipulate that each party is also open-minded—i.e., seriously willing to be convinced by the other’s arguments or critical doubts. 11 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION 5. The community-based argument: “To compromise would help to build a community” (339). Yet, to establish their thesis that it is sometimes permissible to shift from a critical discussion to a negotiation in order to reasonably settle a difference of opinion, van Laar and Krabbe commit themselves not only to the claim that these reasons can motivate such shifts. In stating their permissive policy they also commit to the claim that that these reasons warrant, or justify, such shifts, by providing sufficiently good reasons to shift. In “Criticism and justification of negotiated compromises” (Journal of Argumentation in Context 2019a), van Laar and Krabbe cite these motivating reasons as reasons that may be offered to justify negotiation as a means to dispute settlement, claiming that they may be offered in answer what they call the general compromise criticism, namely the critical questions “Should we settle for compromise [at all]?” and “Should one not rather deal with the problem [of managing the disagreement] in some other way [than negotiation]?” (97f.). In “Splitting a difference of opinion,” in an effort to show that these motivating reasons also can justify the shift to negotiation, van Laar and Krabbe endeavor to demonstrate that the fallacies associated with dialogue shifts can be avoided by their permissive policy for shifting to negotiation. 3.3 Avoiding the fallacies of shifting to negotiation Of the principles and fallacies pertaining to dialogue shifts introduced in §2, van Laar and Krabbe explicitly address the fallacy of middle ground (mistaking a compromise for a resolution) and the fallacy of bargaining (substituting offers for arguments). As they present it, their proposal obviously avoids both. To avoid the fallacy of middle ground, they take care to distinguish compromises from resolutions (343). For example, a concession that not-p may, they claim, be held consistently with the standpoint that p and does not require its retraction. To avoid the fallacy of bargaining, they explicitly claim that negotiation is not a means to resolution (343) and stipulate that the conditions of transparency and intersubjective validity shall be met in order that a shift to negotiation be permissible (335f.). For example, they write: “the Fallacy of Bargaining can be avoided by obtaining the interlocutor’s (explicit or implicit) consent, for example by means of the various motivating reasons for turning to a negotiation dialogue” (2018b: 551). Having sketched the case van Laar and Krabbe offer in support of their permissive proposal for shifting to negotiation, we proceed to identify, locate, and circumscribe the controversial elements of the proposal. 4 Shifting to negotiation: Locating the controversy Importantly, van Laar and Krabbe are concerned not with negotiation per se, but rather with a dialectical shift from a critical discussion to a negotiation—claiming that, in some circumstances, it is permissible (i.e., reasonable) for participants to shift from an unresolved critical discussion to a negotiation in order to settle their difference of opinion. Crucially then, they propose a shift to negotiation only after a critical discussion (at some stage) fails to deliver a resolution. This invites two questions: (i) Why not start off directly 12 DAVID GODDEN AND JOHN CASEY with negotiation? (ii) Why is consensus achieved via a negotiated compromise preferable to the residual dissensus of an unresolved critical discussion? 4.1 Why not negotiate from the get-go? (or, the methodological priority of argumentation) The shift-to-negotiation scenario envisioned by van Laar and Krabbe presupposes the normative legitimacy of the initial critical discussion—it should be attempted first. Call this normative claim the methodological priority of argumentation: insofar as it can be settled at all, the relevant difference of opinion ought to be resolved via critical discussion. The methodological priority of argumentation prizes the goal of resolution above any other state of a discourse because it prioritizes critical discussion over all other means of dispute management and agreement generation. In the absence of such a presupposition, one may readily identify discursive contexts in which a shift to negotiation is unproblematically, indeed trivially, permissible: namely, any context in which one ought to have, or might just as well have, negotiated from the outset. As we noted earlier (§2.4), the normative risks associated with trivially permissible shifts are not to be found in the shift itself; rather they occur in the dialogue preceding the shift, which can disorient or even derail subsequent dialogue. We take van Laar and Krabbe’s permissive policy to endorse at least some non-trivial shifts to negotiation. As a consequence, we hold them to be committed to the methodological priority of argumentation in cases where their permissive policy applies non-trivially. And, we take this commitment to have important consequences for the acceptability of their permissive policy. Specifically, we contend that it commits them to the principle of retrospective evaluation (a point to which we return in §5.2), and that it undermines their principal reason for shifting from a critical discussion (the point to which we now turn). 4.2 The normative force of the principal reason for shifting Recall that the motivating idea of van Laar and Krabbe’s principal reason is that a dialogue shift is warranted since the apparent unrealizability of a resolution impedes or prevents consensus. Therefore, attempts at a resolution should be abandoned in an effort to achieve consensus by some other means. When unpacked in this way, the reasoning behind the principal reason makes it seem as though resolution is valued instrumentally as a means to the end of consensus. Yet, this reasoning contradicts the methodological priority of argumentation. If resolutions are valued only as a means to consensus, then there is no reason other than expediency that one ought to adopt critical discussion as a dispute settlement procedure—more efficient ways of achieving consensus ought to take priority. Yet, this instrumental view confuses the relationship between resolution and consensus. The value of a resolution is not that it produces consensus, but rather that, if and when it produces consensus, it does so in a specific way—namely, on the basis of the public use of reason in critical argumentation that tests the acceptability of standpoints. Put differently, the normative status of a consensus position derives not from the fact of agreement, or what is agreed upon, or even from what is achieved as a result of the agreement, but from how 13 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION that agreement was achieved. If the methodological priority of argumentation holds, then consensus is valuable only insofar as it instantiates a resolution. As such, why should the fact that a critical discussion is in an unresolved state be a reason to abandon the dispute resolution process, given that one’s goal is to resolve the matter of the rational acceptability of the standpoint at issue? Rather, it would seem to be a reason to try harder—e.g., to seek out reasons that would determine the matter. Granted, discussants might reasonably deem the perennial unresolvability of a critical discussion to be a reason to give up on the goal of resolution, judging it to be out of reach. This would be a reason to abandon the disagreement, leaving it unsettled, rather than taking up some other means of agreement generation. Yet, the only way for discussants to determine whether some unresolved state of their discourse is evidence of its unresolvability is to persist in their efforts at critical discussion. (Such evidence might, e.g., consist in their discovering that persistence in their acts of reasons-giving deepened the disagreement thereby worsening their prospects for a resolution.) If correct, this argument demonstrates that van Laar and Krabbe’s principal reason is not a good reason to abandon critical discussion for any other means of dispute settlement where resolution ought to be the goal. 4.3 When is consensus preferable to dissensus? This brings us to our second question: why should consensus achieved via negotiated compromise be preferable to continued dissensus within a critical discussion? Importantly, agreement or consensus is not inherently preferable to disagreement or dissensus. Consider, for example, “deals with the devil”: The devil says that all souls should belong to him and should suffer the torments of hell. You disagree on each point, claiming that no souls should belong to the devil and none should suffer hell at all. What’s the worth of a compromise that the devil should get only half the souls but that each of those will suffer only half of hell’s torments? Moreover, what weight should attach to motivating reasons like the following? By giving up only half the souls and protecting those from half the torments of hell, you at least get part of what you want (strategic and principled consequentialist arguments). Besides, neither of you can be certain that no souls might actually enjoy the torments of hell or would be damned anyway (epistemic argument). And, it’s more democratic to heed at least some of the devil’s views and wishes (democratic argument). Finally, it will “foster a relationship based on mutual concern” (cf. 339) between you and the devil, as well as his minions and the souls of the damned, that the latter should suffer the torments of hell (community-based argument). Deal-with-thedevil cases illustrate that consensus is not intrinsically preferable to dissensus. Often, it’s not only more reasonable, but the only permissible option, to persist in one’s disagreement when, by your best rational lights, the other side is just wrong. The persuasive force of the principal reason depends on the claim that consensus is preferable to dissensus. Yet, this is not necessarily so—particularly in cases where the means by which consensus is achieved matters. As such, it remains to be shown why a normatively inferior end state that is more easily achieved by a normatively inferior process (negotiated compromise) should be preferable to a normatively superior end state that is not easily achieved, as yet unachieved, and perhaps unachievable, via a normatively superior process ((un)resolved critical discussion). Importantly, whatever the sense in which an agreement reached through quid-pro-quo concessions is “preferable” to lasting 14 DAVID GODDEN AND JOHN CASEY disagreement, it is not preferable according to the norms, standards, and goals of the critical discussion. This foreshadows our claim (argued for in §5.2) that non-trivial shifts to negotiation will not satisfy the principle of retrospective evaluation. 4.4 Eligibility, optimality, and reasonableness “Deal with the devil” scenarios appear to prescribe a minimum threshold for the kinds of positions that should even be eligible to serve as starting points in shifted-to negotiations.10 Indefensible or manifestly unreasonable positions ought to be non-starters. If correct, this poses further problems for van Laar and Krabbe’s proposal. First, recall that, according to van Laar and Krabbe, shifts to negotiation most naturally arise in critical discussions where each discussant gives no quarter (i.e., they retract neither their standpoint nor their critical doubts about the other’s standpoint). Yet, in such situations, each discussant finds the other’s standpoint to be manifestly unreasonable, if not indefensible. Seemingly, these are precisely the positions that ought to be rationally ineligible as opening positions in negotiation. Van Laar and Krabbe could reply that each negotiator might well find the other’s opening position rationally ineligible, yet their aim is to negotiate to an eligible, indeed optimal, consensus position.11 Yet, notice how, from the perspective of each discussant, the shifted-to negotiations recommended by van Laar and Krabbe proceed from at least one ineligible starting point. When negotiating, having given no quarter in the critical discussion, each negotiator will see themselves as having to trade away elements of a manifestly reasonable and critically defensible position in order to get their interlocutor to abandon elements of a manifestly unreasonable and critically indefensible position. Prima facie, from the perspective of each discussant, making any such concession is entirely unreasonable. Consequently, negotiations beginning from rationally ineligible opening positions seem capable of producing only rationally ineligible compromises—any consensus achieved via this route should appear fatally compromised. Thus, from the perspective of each discussant having given no quarter, continued dissensus ought to be preferable to any compromised consensus. If this reasoning is correct, then ineligible positions are nonstarters and eligibility marks a minimal acceptability condition for opening positions in shifted-to negotiations. Now consider what might be a maximal condition for opening positions in shifted-to negotiations: rational optimality. Sometimes our collective reasons will not select a single view as maximally plausible; rather sometimes several incompatible (contrary) standpoints can be equally—and sufficiently—supported by the available reasons. Since each view is endorsed by reason and neither is favored by reason this might permit one to adopt a stance of rational indifference between them. (An alternative response might involve seeking out reasons that would tip the balance.) In pure inquiry, where only epistemic goods determine value, such a situation seems to require that one suspend judgement, withholding 10 Eligibility, recall, is the condition stipulating that the contents of a minimally acceptable compromise position should make consensus preferable to continued dissensus (344). 11 Recall that the condition of optimality stipulates that a fully acceptable compromise will not lack any content elements that would be both preferable to some party and mutually agreeable to all parties (345). We understand optimality to entail eligibility. 15 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION commitment from each view by weighting each equally. In practical matters, by contrast, when several courses of action are permissible without any one being recommended over another, seemingly, unless one is obliged to seek out determinative reasons, one may blamelessly choose from a menu of such alternatives. Faced with the choice of one among several optimal but incompatible action or policy options, one might blamelessly negotiate their way to any one of them. Yet, one might just as well have thrown the dice. That one negotiated one’s way to a rationally optimal outcome from a menu of them is not discursively praiseworthy and does not speak to the merits of the selection procedure. Rather, being under the mistaken impression that there is some rational motivation to opt for negotiation in such a circumstance, instead of some other random selection procedure, one may rightly be criticized for failing to apprehend the connection between the starting situation, the selection procedure, and the outcome. Having problematized the motivating reasons critical discussants might have for shifting from their unresolved critical discussion to negotiation as a dispute settlement procedure, we now turn to our critical arguments against the justifiability of such dialogue shifts, beginning with the fallacy of middle ground. 5 Resisting the shift to negotiation 5.1 Revisiting the fallacy of middle ground The fallacy of middle ground mistakes a compromise for a resolution, roughly speaking. This analysis invites the questions: How might this fallacy be committed? In what ways might a compromise be mistaken for a resolution? It can’t be a mistake about content. Since the very same discursive end point can be the result either of a critical discussion or of a negotiation, compromises are not distinguishable from resolutions by their contents. So, the fallacy of middle ground is not committed in virtue of arriving at one discursive end point rather than another. Nor, on the permissive proposal, can it be a mistake about the procedure. The fallacy of middle ground might occur when discussants are mistaken about how a discursive end point was reached or achieved. One might, for example, mistakenly think that they negotiated their way to a resolution, or critically argued to a compromise. Yet, van Laar and Krabbe explicitly insulate their position from this possibility by stipulating that the conditions of transparency and intersubjective validity are met in their permissible shifts to negotiation. Perhaps, then, discussants might mistake a compromise for a resolution in terms of the discursive status that the discursive end state has. Yet, van Laar and Krabbe insulate their permissive proposal from any mistakes about the categorical status of (i.e., the label that properly applies to) discursive end-points by stipulating that compromises are not resolutions and do not settle differences of opinion within critical discussions. Yet, even if one made none of the preceding errors, one might still mistake a compromise for a resolution by treating it as though it were a resolution. That is, the fallacy of middle ground might still be committed if one mistakes the functional discursive status of a compromise for that of a resolution, e.g., by allowing a compromise to play the same discursive role as a resolution in subsequent talk and action. On this point, van Laar and 16 DAVID GODDEN AND JOHN CASEY Krabbe’s position is more murky. While they stipulate that a negotiated compromise does not settle a difference of opinion within a critical discussion, and, by implication, that concessions do not qualify as commitments in critical discussions, we contend that, in limited but salient ways, their policy treats negotiated compromises as discursive proxies (or functionally equivalent discursive substitutes) for resolutions, allowing the compromise to do the same (extra-)discursive work as the (unachieved) resolution would have done. Consider that the goal of the sorts of critical discussions that are the foci of van Laar and Krabbe’s permissive policy is to resolve a matter of policy in order to act on it. This practical urgency and preferability of doing something rather than nothing (i.e., of arriving at some policy decision rather than either engage in further inquiry and critical argumentation or endure lasting disagreement) factors significantly in the motivating force van Laar and Krabbe ascribe to their principal reason for shifting from an unresolved critical discussion. The resolution was sought for the (extra-)discursive end that it might serve as a basis for future action. Now, with the resolved position inaccessible (because un-generated within the unresolved critical discussion), the practical value of adopting a permissive policy of shifting to negotiation is that it will allow a (presumably more easily attainable) compromise position to serve that same role. Whether attained through resolution or compromise, the policy position may, according to van Laar and Krabbe, be implemented. Indeed, on their account, this commitment is a consequence of accepting a compromise: “The primary propositional commitment that an agreement on a compromise incurs is the proposition that, all things considered … [the consensus policy position] P merits acceptance and implementation” (344; emphasis added). Just as the resolved position would have, the compromise position may, according to van Laar and Krabbe, serve as a premise in future practical reasoning and as the basis for future action in matters related to the policy consideration at hand. In this respect, the functional discursive status of the compromise is no different from that of the resolution. Call this the discursive proxy (in contexts of implementation) thesis. Indeed, without the prospect of substituting compromise policy positions in place of resolved policy positions in future practical reasoning and action, there is no practical incentive to shift to negotiation when a critical discussion becomes stuck. As such, van Laar and Krabbe’s principal reason by itself fails to motivate any dialogue shift. Rather, the motivating impetus for reaching agreement, and hence for shifting dialogue, is that consensus is required in order to achieve some (extra-)discursive end. Hence, it is only to the extent that any achieved compromise may properly serve as a discursive proxy for a(n unachieved) resolution in respect of those (extra-)discursive ends that discussants have any motivation to shift from their critical discussion to a negotiation. Thus, we conclude that van Laar and Krabbe’s permissive policy for shifting to negotiation commits the fallacy of middle ground, because it permits—indeed its motivation depends on its permitting—a negotiated compromise to be treated as a discursive proxy for a resolution in the talk and action related to implementing the agreedupon policy position. 5.2 Retrospective evaluation revisited At this point van Laar and Krabbe’s neglect of the principle of retrospective evaluation becomes conspicuous. 17 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION Recall that the principle of retrospective evaluation prescribes that dialogue shifts are permissible to the extent that the results and proceedings of the shifted-to dialogue may be judged favorably against the goals and standards of the shifted-from context of dialogue (Walton and Krabbe 1995: 102-103). Again, this principle seems designed, in part, to ensure that achievements gained in the shifted-to dialogue may, in subsequent talk and action, properly perform the same (extra-)discursive tasks as claims established in the original dialogue. Although van Laar and Krabbe never explicitly apply the principle of retrospective evaluation in justifying or evaluating their prescriptive policy, they seem to commit themselves to something like it as a permissibility condition for dialogue shifts. They write: When two parties choose to engage in a dialogue of a specific type, they commit themselves to contribute to the realization of the main goal of that type of dialogue type. We assume that such a choice brings particular dialogical obligations. Thus, a shift to a different dialogue type can be the subject of normative evaluation. (332, fn.3) Plausibly, the principle of retrospective evaluation would not properly apply to a dialogue change if either the first dialogue was somehow redundant or illegitimate, or claims established in the new dialogue would not perform the same (extra-)discursive tasks as claims established in the original dialogue. Yet, as we argued in §4.1, van Laar and Krabbe’s permissive policy commits them to the methodological priority of argumentation—they cannot deny the propriety of the initial critical discussion. Perhaps, then, it is because they take themselves to have avoided the fallacy of middle ground that van Laar and Krabbe seem to not take themselves to be beholden to the principle of retrospective evaluation. Since compromises are not resolutions, they shouldn’t be evaluated as such; hence the principle of retrospective evaluation simply doesn’t apply. Yet, we have just shown (in §5.1) that van Laar and Krabbe’s permissive policy does not entirely avoid the fallacy of middle ground. Rather, in contexts of implementation, their permissive policy permits consensus policy positions to function as discursive proxies for (unachieved) resolved policy positions in contexts of implementation. Thus, the principle of retrospective evaluation seems as though it should apply to non-trivial dialogue shifts sanctioned by the permissive policy. Yet, the fact that critical discussions aim at generating resolutions while negotiations are only capable of generating compromises entails that negotiations cannot satisfy the goal of a critical discussion. By failing to satisfy the goal of a critical discussion, dialogue shifts sanctioned by the permissive policy fail the goal condition of the principle of retrospective evaluation. And, the force of this criticism is normative, not semantic. As we noted earlier, the value of a resolution is not that it produces consensus, which might be more easily achieved by other means, but rather in how any resulting consensus was produced—namely via the transaction of reasons in argument. Not only does the permissive policy fail to satisfy the goal condition of retrospective evaluation, it also fails the standards condition. Dialectical theories construe normative standards as sets of procedural rules (e.g., lists of obligatory, permissible, and prohibited moves) for the appropriate activity type. The standards condition of retrospective evaluation thus prescribes that the set of permissible move sequences in any shifted-to dialogue should be a subset of the permissible move sequences in the shifted-from 18 DAVID GODDEN AND JOHN CASEY dialogue. To allow any new move sequences in a shifted-to dialogue is to permit moves that are somehow fallacious (i.e., not permitted) by the standards of the shifted-from dialogue. Thus, if the procedural rules for negotiation permit move sequences that are not licensed within a critical discussion, then those move sequences are fallacious according to the evaluative standards of critical discussions and the shift to negotiation fails to meet the standards condition of retrospective evaluation. Yet, given the unresolved state of the shifted-from critical discussion, it is only because such new, unlicensed and fallacious, move sequences are possible in the shifted-to negotiation that it is capable of offering any dispute settlement resources that were not already available within the shifted-from critical discussion. Indeed, this is the promise that informs the principal reason motivating the dialogue shift: our dialogue has become stuck, so perhaps we should try something else. As such, either shifting to negotiation is redundant (and thus imprudent) or it runs afoul of the standards condition of the principle of retrospective evaluation. 5.3 Revisiting the fallacy of bargaining Having shown that the permissive policy fails to satisfy the principle of retrospective evaluation brings us to reconsider the fallacy of bargaining. Recall that, roughly stated, the fallacy of bargaining consists in substituting offers for arguments. As we read them, by distinguishing the activities of critical argumentation, which we have characterized as consisting of reasons-responsive moves, and negotiation, which we have characterized as consisting of trade-off-responsive moves, van Laar and Krabbe take themselves to have avoided the fallacy of bargaining. We disagree. To appreciate our reasons, it is best to start with cases of bargaining for the truth. 5.3.1 Bargaining for the truth: The basic problem 5.3.1.1 The failure of quid-pro-quo moves to offer epistemic reasons Beliefs, and their cognitive and discursive kin (e.g., opinions, assertions), are alethicallyoriented and truth-responsive: they aim at truth; they have a mind-to-world direction of fit. We have the beliefs that we do because we take them to be true—which is to say that, to the extent that they do not accord with the way things are, they are subject to correction and criticism. Relatedly, if reasoning and argumentation function to induce or evaluate truth-apt, or truth-responsive, propositional attitudes (whether these are called “beliefs” or “commitments”), they ought to reliably track the epistemic and alethic properties of the standpoints at issue. Reasons that fail to indicate truth should cease to have epistemic currency. So, upon discovering that our argumentative practices (i.e., the ways by which we value and transact reasons) fail to track epistemically relevant properties of claims, we ought to revise our argumentative practices accordingly. Now, consider the kinds of discursive moves that occur in negotiations. On our count, there are four basic types of concession moves, or quid-pro-quo offers, in any negotiation: Four basic moves in negotiation Commitment for commitment [CC]: “If you commit to p, I will commit to q.” Retraction for retraction [RR]: “If you retract q, I will retract p.” 19 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION Commitment for retraction [CR]: “If you retract q, I will commit to r.” Retraction for commitment [RC]: “If you commit to p, I will retract s.” This list might be supplemented by the following ultimatums: Final offer [FO]: “My final offer is p; take it [i.e., commit to p] or leave it.” Quit move [QM]: “If you (don’t) commit to p / retract q, I will quit the negotiation.” Notice that none of these moves carries any force if ‘believe’ is substituted for ‘commit to’ and ‘disbelieve’ is substituted for ‘retract.’ (Contrast this with the substitutions ‘undertake to f’ for ‘commit to’ and ‘refrain from f-ing’ for ‘retract,’ where ‘f’ names some action.) “If you (will) believe p, I will believe q” is logically incapable of having any doxastic force—it utterly fails to provide a reason to believe. And the reason for this is that your undertaking to believe p is not evidence for the truth of q—it doesn’t count in favor of, or against, the truth of q at all.12 In a phrase: beliefs are not the sorts of things that can be bartered. Because of this, they may not be bartered, and to the extent that one does barter with them, they cease to be belief-like.13 14 12 To appreciate this point in the context of practical reasoning, consider an example Douglas Walton suggested to us in conversation. Suppose a prospective car buyer says to the sales representative at the dealership: “I don’t want to buy car X because I am not convinced that it has adequate safety features, like anti-lock brakes.” Consider now the aptness of a response like the following from the salesperson: “Okay, I’ve just spoken to our sales manager, and we can discount the price of the car by $ 1,000.” Here, the salesperson’s response seems to entirely miss the buyer’s point. The salesperson’s undertaking to do something—specifically something that doesn’t affect the truth-value of the consideration cited by the buyer as a reason—fails entirely to address the buyer’s reason as a reason. (Here is something that the salesperson might do that would address the reason; they could say: “Okay, we’ll install anti-lock brakes on the car for you.”) The very idea of putting a non-epistemic value on the reason, such that it might be given up for a price or other non-epistemic benefit, seems inept. It completely misunderstands move made by the prospective buyer. The salesperson is “playing the wrong game.” Apt responses would include moves like these: “Well, here’s a newer model of car X, where they have upgraded the safety features to include anti-lock brakes,” or “Oh, that’s a misprint in the specifications; car X actually does have anti-lock brakes,” or “You’re right, car X doesn’t have anti-lock brakes; it has autonomous braking are even safer than anti-lock brakes.” 13 Dialectical theories of reasoning and argumentation tend to prefer the language of commitment (or verbal acceptance) over that of belief (or mental acceptance) (Godden 2010: 399ff.). At least part of the reason for this seems to be that commitments, being externalized, are easier to track and hold others accountable to. Yet, the point being made here does not call for a particular interest in the unarticulated, psychological dimensions of argumentation. Rather, it concerns the alethic orientation of those propositional attitudes or commitments (whether externalized through public speech acts or not) that are at stake in argumentation. Asserting that p, for example, expresses one’s commitment to the truth of p. Ordinarily, this is properly understood as a doxastic commitment—as an expression of one’s belief that p. Belief, for purposes of this discussion, may thus be understood as alethically-oriented commitment—e.g., as those (perhaps inner) commitments that are rightly subject to criticism, and ought to be qualified or retracted, to the extent that they are not true. An epistemological concern with the dialectician’s adoption the language of commitment, rather than that of (externalized) belief, is not so much that it neglects the inner, psychological, dimensions of reasoning and argumentation (though this is a concern of normative import; see Godden (2010)), but that it has been applied in ways that neglect or obscure the alethic orientation of much of our discourse, and that it thereby fails to afford truth, rather than acceptance or agreement, its proper place as a discursive norm. Indeed, one might explain inclinations towards discursive shifts of the sort sanctioned by the permissive policy as valuing agreement over the truth, or rational acceptability, of what is agreed upon. 14 This, of course, includes claims of verisimilitude, e.g., claims of the form “p is close enough to the truth for some purpose, F.” Claims of verisimilitude, satisfycing, practical efficacy, within an acceptable margin 20 DAVID GODDEN AND JOHN CASEY 5.3.1.2 The failure of the motivating reasons to shift to negotiation in epistemic contexts Relatedly, as with “deal-with-the-devil” cases, when each party in a critical discussion has given no quarter concerning the epistemic merits of the other’s position, all five of van Laar and Krabbe’s motivating reasons for shifting to negotiation fail not only to justify any such shift, they fail to motivate it at all. The (1) strategic, (2) principled-consequentialist, (4) democratic, and (5) community-based arguments for shifting all fail because they reference the wrong kind of criteria for the epistemic acceptability of some position or procedure. Against the strategic and principle-consequentialist arguments: since truth cannot be bargained, nothing epistemically worthwhile, and none of one’s epistemic goals, can be achieved via negotiation. Against the democratic argument: truth does not side with a majority, or multi-sided opinion. Rather, we explain the rightness of the opinions of the many in terms of their truth, not the truth of opinions in terms of their being held by the many. Incorporating a broader range of viewpoints per se is not epistemically valuable. Additionally, one must have reason to judge that giving audience to those alternative viewpoints will contribute to reducing ignorance and error. Yet, having given no quarter, no such judgement seems plausible in the situation of dialogue shift envisioned by van Laar and Krabbe. Against the community argument: building good relations within a community is not a good reason for adopting manifestly unreliable epistemic practices. Epistemically speaking, building a community around a falsehood, or a truth for which we lack adequate reason to believe, is not to be preferred over having justified truths which might divide a community. As W.K. Clifford pointed out in his (1877) essay “The ethics of belief,” this only fosters credulousness in the community. So far, each consideration cited as a reason for shifting to negotiation, making compromises, and accepting a concession in lieu of a resolution is epistemically inert. Finally, while the (3) epistemic argument cites epistemically relevant considerations—that our own views are incomplete and fallible, epistemically speaking—those considerations do not warrant abandoning critical discussion, let alone abandoning it for negotiation. Just the opposite! Together with inquiry, critical discussion, not negotiation, is the means by which we detect and correct for ignorance and error among our commitments. 5.3.2 The fallacy of bargaining in deliberative contexts Previously, David Godden (2016: 3f.) offered an argument along these lines challenging the legitimacy of shifts to negotiation in “contexts where epistemic considerations are relevant.” In response, van Laar and Krabbe replied: “we agree with Godden that our account cannot, without further ado, be applied to settings that lack a practical aspect” (341). Here, we will show that the mere presence of a practical aspect to the reasoning, either because of the discursive context or the matter at hand, does not answer the criticism or avoid the fallacy of bargaining. While van Laar and Krabbe limit the focus, but not the scope, of their permissive policy to disagreements about action prescriptions or policy proposals, the paradigm situation they have in mind is one where circumstances, i.e., (extra-)discursive goals, motivate of error, etc. are truth-apt and have the nature of beliefs in this respect: that their propositional contents are false always counts as good grounds for their critique. 21 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION discussants to make a practical decision. The type of cases they consider are ones where critical discussants are engaged in a persuasion dialogue about policy options, and this persuasion dialogue becomes deadlocked. We would call the reasoning involved in such cases deliberative or practical reasoning (also called ‘instrumental,’ ‘prudential,’ or ‘means-end’ reasoning). It is reasoning about what to do; its conclusions are action policies. Generically, such reasoning may be schematized as follows: Practical Reasoning Scheme P1. Goal Premise: S wants (desires, values, intends) some end or goal, that g. P2. Means Premise: Effecting (i.e., bringing it about that) m, is a (required, optimal, effective, efficient, available, etc.) means to achieve (i.e., bring it about that) g. \ C Conclusion: Action Policy: S (or some appropriate agent or agency) ought to do (i.e., bring it about that) m. (cf. Walton, Reed, and Macagno 2008: 323f.; Walton 1998: 153f.) Disagreements about action policies, then, can be reduced to disagreements about goals and means. Consider now the extent to which each component of this reasoning is amenable to being bargained. 5.3.2.1 The fallacy of bargaining about the means Let’s start with the means premise. This premise cannot reasonably be bargained for exactly the reasons outlined in §5.3.1. The means premise asserts that some specific relation obtains between some purported end and some putative means thereto. Such claims about the (comparative) efficaciousness of particular means in achieving particular ends are truth-apt and, typically, contingent. Commitment to, and contention of, a means premise is thus a commitment to its truth—the relevant propositional attitude is alethicallyoriented and thus doxastic in nature. As such the rational acceptability of a means premise consists entirely and exclusively in its believability—i.e., our acceptance of such claims should be based on, and proportional to, the evidence we have in their support. Yet, an alethic commitment to a means-end claim cannot be established, and entitlement to it cannot be secured, through compromise. Moreover, any policy decision based on a means premise is criticisable to the extent that disputants are mistaken about, or unjustified in accepting, that means premise. Thus, an action policy cannot be reasonably bargained insofar as its acceptability ought to be partly based on the acceptability of a means premise that cannot properly be bargained. To elaborate, suppose that critical discussants cannot agree on a policy decision because, even though they have achieved a resolution about the goal, their critical discussion has not produced a resolution about the means. Prima facie, this calls not for bargaining about the means, but rather for further inquiry and critical argumentation, seeking out and critically scrutinizing the evidence that would determinatively settle the matter of the means. Suppose though that, unfortunately, practical constraints limit the opportunity for further inquiry, such that any decision must be taken on the basis only of the incomplete, non-determinative, information available. Even in this situation, we argue, 22 DAVID GODDEN AND JOHN CASEY the proper method for selecting the means is to do so on the basis of the best epistemic reasons available, rather than by bargaining. Ceteris paribus, the means that should be employed is the one that the available evidence suggests will be the most likely to be efficacious, over and above some minimal, actionable threshold of likelihood of success. Of several in-principle feasible means each of which satisfy the acceptable likelihood of success threshold, practical considerations might also be weighed in deciding upon optimality, e.g., practical feasibility, cost-effectiveness, etc. Yet, these again are factual matters, the rational acceptability of which is not susceptible to being bargained. Perhaps, after all these considerations have been weighed through a process of critical argumentation, several incompatible means remain such that reason finds each acceptable and none preferable, over all, to any of its competing alternatives. When considering such means we should be rationally indifferent. Even here, while bargaining among these alternatives might be permissible, since it will be no worse than any other selection method, as we noted earlier the view that bargaining is a preferable selection procedure in such circumstances mistakes the connection between the starting situation, the selection procedure, and the acceptability of the outcome. As a final possibility, bargaining in such circumstances might generate a new or hybrid means claim as a compromise outcome. Yet, while negotiation might generate such a claim, its acceptability remains to be established via the rational scrutiny of critical discussion. So, means claims are not susceptible to being bargained. 5.3.2.2 The fallacy of bargaining about the goal Next, let’s proceed to the goal premise.15 Here, there are two kinds goals to be considered: pure preferences and rational preferences. If a goal premise expresses a pure preference then it is not subject to rational justification or criticism; pure preferences are not the sorts of things that call for, or stand in need of, reasons.16 If a goal, g, is a pure preference, then, in response to the challenge, “Why do you want g?”, a sufficient acceptable response (i.e., one that conclusively answers the challenge) is “Because I want it,” or “Because I prefer it to the alternatives.” Relatedly, if a goal, g, is a pure preference, then it is not susceptible to criticism on the grounds that g is a goal that one (or, perhaps, you) ought not to have. If a dispute over action policies reduces to a difference between incompatible purepreferences, where the matters about the means have been resolved, then disputants might as well bargain over the goal from the get-go, as there is nothing to reason about! In cases where a shared goal is a pure preference, agreement, understood as intersecting autonomous preferences, seems sufficient to warrant its acceptability. In such cases, though, the methodological priority of argumentation is not satisfied. The mistake in 15 Analogous considerations to those raised here in the context of goal evaluation apply directly to the evaluation of action policies, or choices, where the performance of the action itself is its own end or goal. 16 If one denies this claim, as did one reviewer who claimed that pure preferences can be unreasonable and criticized as such, then, on our understanding of pure and rational preferences, the preference is rational, not pure. Our argument, nevertheless, stands as written: while pure preferences can, sometimes, be explained, properly understood they are not susceptible to rational criticism or justification. Hence, while they may be bargained, it is a mistake to argue about them. By contrast, rational preferences may be argued about but are not susceptible to bargaining. In either case, shifting from a critical discussion to a negotiation about them is either trivially permissible (in the first case) or impermissible (in the second). 23 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION shifting to negotiation in such cases occurs when disputants mistakenly think that they ought first to argue, indeed that they can meaningfully argue, about pure preferences. Most of our goals, though, are not pure preferences—certainly not the meaningful or momentous ones. Most of our goals are susceptible to criticism on the grounds that they’re ones that we ought not to have, and our choice of goal typically stands in need some reason beyond that we prefer, or want, it. Goals like these are rational preferences. Such rational preferences are not merely motivated by, and intentionally explained in terms of, reasons, but are justified by them. Such reasons establish these goals as permissible or perhaps obligatory (for us). Yet, insofar as we adopt or endorse our goals on the basis of reasons, those reasons cannot be bargained away or gained by bargaining. Generally, goals can have two kinds of value: instrumental and intrinsic, and these two kinds of value provide two types of reasons by which our choice of goals may be justified. Goals have instrumental value insofar as they are valued for the sake of, or as a means to, some other end. So, a goal, g, has instrumental value with respect to some end, h, to the extent that accomplishing g is a means to achieving the further goal, h. As such, our reason for having g as a goal is that it is instrumental to attaining h, and g may be criticized, and must be justified, in respect of its being an effective or optimal means to h. Yet, as we just argued, claims about means are fact-based and consist in expressions of alethicallyoriented commitments which cannot not be bargained. So, just as means are not susceptible to being bargained, neither are instrumental goals. Goals have intrinsic value insofar as they are valued for their own sake or as an end in themselves. So, a goal, g, has intrinsic value insofar as (accomplishing) it is itself valuable. Yet, unless one is a nominalist about intrinsic value, then a thing (e.g., object, state-ofaffairs, event, goal) has the intrinsic value that it does because, independently of anyone’s attitudes or judgements, it has some property or feature in virtue of which it is intrinsically valuable. That is, g’s intrinsic value, v, will be grounded in, or supervene upon, some other (discernable) property, p, of g. Yet, this fact determines the structure of the reasoning capable of establishing, and required to establish, that g has v. Specifically, the acceptability of the claim [C] that g has v, will depend on the truth of [P1] that v is grounded in p and [P2] that g has p. Yet, P1 and P2 are matters of fact and their assertion expresses an alethically-oriented commitment. So, once again, the reasonableness of one’s goal depends on the reasonableness of related alethically-oriented commitments. And, again, these cannot be bargained. Suppose, instead, that one is a nominalist about intrinsic value. In that case, things have the value they do in virtue of someone having assigned them that value, where there is no logical possibility of the assigner being corrected or mistaken in their assignment of value. As such, value assignments are not judgements but expressions of pure preferences on the part of the valuer who is able to nominally confer intrinsic value upon a thing. (For those other than the valuer, value assignments are complex judgements, consisting of two judgements: one about the identity of the valuer and a second about their preferences.) Yet, as we have already argued, pure preferences are not the sorts of things that call for, or stand in need of, reasons. As such, while they may be bargained, it is a mistake to argue about them! Summing up the arguments offered in § 5.3, merely stipulating that some matter under discussion has a practical aspect does not make it susceptible to being bargained. Rather, alethically-oriented standpoints and standpoints whose acceptability properly depends on their logical or justificatory connection to claims having an alethic orientation, are not 24 DAVID GODDEN AND JOHN CASEY susceptible to being bargained. As such, even when applied to the domain of claims van Laar and Krabbe focus on, namely action prescriptions or policy proposals, so long as these claims are properly understood or justified as the conclusions of episodes of practical reasoning or argument, we conclude that their permissive policy is guilty of permitting the fallacy of bargaining. 5.4 Revisiting the fallacy of abandonment of commitment / discussion This leaves us with the fallacy of abandonment of commitment / discussion. Recall that, roughly stated, committing this fallacy involves reneging on one’s discursive commitments, particularly one’s obligation to defend one’s stated views. Being in the discursive vicinity of a dialogue shift amplifies the risk of falling victim to this fallacy. We contend that, despite their stipulation that negotiated compromises do not settle disagreements within a critical discussion, van Laar and Krabbe’s permissive policy nevertheless sanctions such fallacious behavior. Here are some ways that have already been noted in the preceding discussion. Most generally, in citing the principal reason for shifting—that the unresolved state of a critical discussion is a reason to abandon it for negotiation rather than to try harder at finding a resolution—proponents of the permissive policy renege on their commitment to the methodological priority of argumentation and to the value of resolution over consensus. This is exacerbated when compromises are then permitted to serve as discursive proxies for (unachieved) resolutions in contexts of implementation. Championing the epistemic motivating argument for shifting to negotiation involves abandoning one’s commitment to critical discussion as the means by which we reasonably detect and correct for ignorance and error among our commitments. Finally, in falling foul of both the goal and standards conditions of the principle of retrospective evaluation, in shifting to negotiation one reneges on one’s commitments to the norms and goals of shifted-from critical discussions. In permitting the substitution of offers for arguments, the permissive policy allows discussants to abandon claims they are obliged to defend in exchange for adopting claims that they are not in a position to defend. Each of these points depends on a distinction, made at the outset of our discussion, between concessions, understood as a trade-off-responsive moves, and retractions and acceptances understood as reasons-responsive moves. Yet, in their paper “The role of argument in negotiation” (Argumentation 2018b), van Laar and Krabbe argue that “a series of offers and counteroffers in a negotiation dialogue contains a reconstructable series of implicit persuasion dialogues” (549, abstract) consisting of “expediency arguments from consequences” (554). If the parties negotiate in a way that goes beyond mere pressuring and that contains serious attempts to convince one another of the acceptability or unacceptability of particular compromise options, we can understand the quid pro quo moves … implicit in the negotiation dialogue as appealing to, and thereby conveying, expediency arguments by which the speaker attempts to convince the interlocutor of the acceptability of a proposed compromise (offer), given the interlocutor’s own value assignments. (554) 25 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION From this, they conclude that “argument has a role to play in splitting differences of opinion [i.e., settling a difference of opinion via negotiation] in a reasonable way, and the shift from a persuasion dialogue about a difference of opinion to a negotiation dialogue about the same difference … does not amount to the abandonment of argument” (566; emphasis added). We wish to flag this conclusion as misleading. If read as a retort to the fallacy of abandonment of discussion / commitment, this claim distorts the difference between a reason-tracking procedure, consisting of reasons-responsive moves, and a preferencetracking procedure, consisting of trade-off-responsive moves. It seems, though, that van Laar and Krabbe seek to preserve this distinction via their characterization of the kinds of reasons transacted in expediency arguments from consequences. They write: Bargaining can thus be reconstructed as containing a series of argumentative exchanges about what the participants want, about how much they want it, and about the expediency of particular choices in light of information about existing preferences, which gradually becomes available. (557) On this picture, while trade-off responsive moves are not identical to reasons-responsive moves, each trade-off-responsive move (e.g., offer) is typically connected to a reasonresponsive move (having the form of an expediency argument). Yet, insofar as making a trade-off responsive move is explained as motivated by responding to an expediency argument (i.e., a reason-responsive move), the line between the two kinds of moves becomes blurred. Yet, if one accepts our claim that the basic types of concession moves (quid pro quo offers and ultimatums) lack epistemic traction, then it is worthwhile to mark the difference between reasons-responsive moves and trade-off-responsive moves with a bright line, rather than the blurred one found in the above picture. This last point deserves further elaboration, since it marks a fundamental difference in the nature and orienting values of the activities of critical discussion and negotiation. Consider two discussants, Sam and Erin, who disagree about some matter of policy, p, and are discussing which proposal to adopt. Sam advocates for the complex policy proposal {A,B}, while Erin advocates for the complex policy proposal {C,D}. Suppose further that (i) the commitment pairs {A,D} and {B,C} are each consistent, while {A,C} and {B,D} comprise incompatible contraries, and Sam and Erin each recognize this; (ii) Sam and Erin each take themselves to be rationally entitled to their respective commitments; and (iii) in fact, B and C are only supported by misleading evidence or fallacious reasoning, while A and D are supported by good evidence and cogent reasoning. In this circumstance, Sam and Erin should, and ideally would, come to agree upon {A,D} in settling their difference of opinion, since this outcome is more reasonable—and thus objectively preferable in that sense—than {B,C}. Given all the facts, Sam and Erin ought to rationally prefer {A,D} to {B,C}. As such, we suspect that if proponents of the permissive policy thought that negotiation, as an agreement generating procedure, would fail to reliably select for {A,D} over {B,C}, they would not recommend it. That is, we suspect that proponents of the permissive policy tacitly envision that {A,D} will, somehow, tend to be the result of successful negotiation—that negotiation will somehow produce, or at least select for, reasonable, rationally preferable outcomes as compromises. Yet, we 26 DAVID GODDEN AND JOHN CASEY claim that, given the nature of negotiation, there is no reason for this supposition. Indeed, all that need be supposed to show that negotiation will tend towards, if not reliably select for, {B,C} as the compromise view is that (iv) that Sam subjectively prefers B over A while Erin subjectively prefers C over D, where ‘subjectively preferring p over q’ means something like “would rather that p were the case than q” or “would rather give up q than p,” perhaps due to one’s view that p is epistemically superior to q. And, we contend, that there is no reason to suppose that our preferences, so conceived, will naturally align with our rational entitlements. Rather, if they do, this is a result of critical reflection and argumentation. As a compromise position, {B,C} is not criticizable, and nor is a negotiation procedure that generates it as a discursive end state. By contrast, {B,C} is criticizable as a resolution. Our practices and norms of critical argumentation are designed to select for reasonable outcomes, and are faulty, criticizable, and ought to be revised to the extent that they don’t. The same just doesn’t hold for negotiation. The values that orient negotiation as a discursive practice are orthogonal to what is reasonable. Thus, there is no reason to expect that when we elect to split a difference of opinion by negotiating a compromise, this procedure will reliably select for reasonable outcomes. Negotiation remains a preference-tracking procedure and reconstructing the proceedings of a negotiation as a sequence of “reasons-giving” moves does not change this. Concessions remain trade-off-responsive moves, the normative structure of which is categorically different from that of the reasons-responsive moves retraction and acceptance. In whatever sense trade-off-responsive moves are “reasonable” or “unreasonable,” it is not the normative value and attendant standards that critical discussions are designed to track. When viewed from this perspective, shifting to negotiation does amount to the abandonment of argument—and not merely the activity of argumentation itself (together with the commitments one has within argumentation), but, much more importantly, the commitments one has about argumentation as an activity. 6 Rational hazards of a permissive policy for shifting to negotiation 6.1 Incentivizing argumentative overcharge and abandonment of discussion Having presented our critical arguments against the reasonableness of a permissive policy for shifting to negotiation, we proceed argue that such a policy is imprudent. Indeed, by incentivizing two rationally and argumentatively hazardous behaviors a permissive policy acts to undermine the normative structure and rational ends of the activity of critical argumentation itself. First, a permissive policy for shifting to negotiation structurally incentivizes a kind of strategic maneuvering we call argumentative overcharge. Argumentative overcharge involves overstating one’s initial position beyond what one is actually in a position to support or successfully defend with reasons, with the aim of bargaining one’s way to an agreeable compromise, rather than ever having to argue for it as a defensible standpoint. And, this is exactly the opposite of the structural incentive one has if one knows that one must finish out a critical discussion by either defending one’s standpoint from reasonable criticism or retracting it. 27 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION Importantly, committing an act of argumentative overcharge need not entail that the proponent is insincere in their commitments—e.g., in holding or articulating their initial standpoint. (That is, such proponents do not necessarily or obviously violate any sincerity requirement.) A proponent might genuinely accept (i.e., believe) their overstated standpoint. They might simply be overconfident or blithe about the strength of their reasons or about their prospects of dialectical success. Consider that people generally have the views they do, and persist in disagreements about those views, precisely because they take their views to be well-taken—i.e., correct and not unreasonable. Consider also that, at the confrontation stage when disagreement is externalized and recognized, the strength of a proponent’s reasons, and thus the acceptability or defensibility of their standpoint, has yet to be tested in critical discussion. The insincerity that ought to concern theorists here is not the sincerity with which discussants hold the views they have when entering into critical argumentation, but the sincerity with which they commit to the process of critical argumentation as a means of fixing, and determining the rational acceptability of, the views they have at the end of the discursive process. This type of insincerity might be called dialectical insincerity. Van Eemeren and Grootendorst (2004: 77) express the Pragma-Dialectical honesty principle in terms of dialectical insincerity. “The honesty [i.e., sincerity] principle implies that everyone may be held responsible for assuming the obligations linked to the speech act that he or she has performed.” So formulated, the sincerity principle grants critical discussants the right to hold each other dialectically accountable for their respective commitments. Yet, by presenting the option of shifting to negotiation, the permissive policy excuses critical discussants from the responsibilities attendant to the honesty principle and thereby creates a structural incentive for them to not attend to those responsibilities. Proponents in a critical discussion face the risk of retraction. Should they be unable to defend their standpoint from critical doubt, they are obliged to retract it. Thus, if one knows that one will be compelled to defend one’s initial standpoint, and that its rational acceptability depends on one’s dialectical success in defending it from criticism, proponents have a structural incentive to articulate the most defensible—i.e., weakest, most qualified or hedged—version of the standpoint they hold. Adopting such standpoints minimizes the risk of retraction. By contrast, if one knows that deadlocked critical discussions may proceed to negotiations, then retraction is not a risk.17 Rather, a permissive policy on shifts to negotiation provides a structural incentive for proponents to take, as an initial standpoint, the least defensible—i.e., strongest, least qualified or hedged—version of the position they hold. Proponents have a structural incentive to articulate opening positions that are overstated to such an extent that there is enough to “bargain away” such that what remains is the standpoint they initially sought to establish. Argumentative overcharge is not the only rational/argumentatively hazardous behavior encouraged by a permissive policy for shifting to negotiation. Second, the fallacious behavior of abandonment of discussion is not merely permitted, it is structurally incentivized. By presenting, if not recommending, the option that critical discussants turn to negotiation when settling their resolution-resistant differences of opinion, arguers are incentivized to enter into critical discussions with the aim of deadlocking them rather than arguing reasonably—particularly if they feel that their positions are not rationally 17 Notice: this is a consequence of van Laar and Krabbe’s distinction between a resolution and a compromise and their claim that compromises do not end disagreements (see our §2.3). 28 DAVID GODDEN AND JOHN CASEY defensible. Viewed in this way, the permissive policy incentivizes dialectical insincerity. By relieving arguers of the risk of retraction, one excuses them from the obligation to defend, and thereby from the obligation to engage in the activity of arguing altogether. Rather, critical discussion simply becomes an overture to negotiation—it is abandoned before it has even begun. Indeed, one might even understand argumentative overcharge and deadlocking critical discussions as kinds of negotiating pressure (van Laar and Krabbe 2019b)—pressures to negotiate. Like setting a list price, or manufacturer’s suggested retail price, opening positions can have an anchoring effect whereby the “reasonableness” of alternative positions is partly judged according to their relative proximity to the initial offer, with big differences being less reasonable than smaller ones. And, in accordance with van Laar and Krabbe’s principal reason for shifting, when the distance separating those positions is too vast to be reasoned across, some other path to agreement must be found. With each kind of strategic maneuvering “a speaker changes the social circumstances for the continuation of the conversation” (2019b: §3), specifically directing it towards a dialogue shift to negotiation. In sum, we find the structure of incentives for strategic maneuvering presented by a permissive policy for shifting to negotiation to undermine the very goals and standards embodied in the practice of critical argumentation. Arguers are not incentivized to align their externalized standpoints with the strength of the reasons they have for them; nor are they incentivized to be responsive to reasons offered in critical argumentation. Rather, arguers are incentivized to be unreasonable! 6.2 Overcharging and plea bargaining Our concerns about the rational and argumentative hazards of argumentative overcharge and abandonment of discussion find an illuminating example in the practice of overcharging, common in United States civil and criminal law. In its broadest formulation, overcharging consists in alleging more at the charging stage than one intends to prove at trial. (Again, this needn’t be insincere, since the prosecutor may judge the charges to be provable at trial.) The aim of the overcharge is to induce a defendant to waive their right to a trial in favor of a voluntarily accepted negotiated settlement. The incentives to overcharge as a bargaining tool, on the part of the prosecutor, are strong. They can (1) claim success at conviction and (2) avoid expenditure of time and resources in exchange for a less harsh sentence than they would otherwise have been able to secure. The overcharge plays a critical role in this strategy, as the prosecutor is able to overstate the charge or swamp the defendant in an array of similar counts. Viewed from a broader argumentation perspective, the purpose of the overcharge is to induce an opponent in a critical discussion to weigh the costs of arguing and the consequences of their standpoint not prevailing in front of an impartial judge and so opt for a negotiated settlement of the disputed questions. Overcharging can take a few different forms. First, one might allege more than they can hope to prove as a kind of feint. Second, one might charge more than is fair or customarily appropriate to the gravity of the offenses or wrongs alleged. Finally, one might charge more than one intends to prove with the specific intention that the defendant cease resisting and opt for a plea-bargained settlement (Graham 2014). Most relevant, from a 29 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION negotiation perspective, is the third. We would allege that the first two are straightforwardly unreasonable strategies: the first is plainly dishonest while the second is extreme. Besides, though the bar is low, patently dishonest overcharging violates the American Bar Association’s code of legal ethics (American Bar Association 3-4.3). The critical feature of the third variety of overcharging is the question of sincerity: the one making the charge doesn’t actually maintain the claims alleged except for the purposes of strengthening their position for a negotiated settlement. Crucially, the things alleged are not in themselves unreasonable or false. They’re just not claims the person making the charge actually endorses or intends to follow through on; they only endorse them for the purposes of bargaining on them. Alschuler (1968) distinguished two common overcharging strategies. The first, called horizontal overcharging, is where a criminal prosecutor may charge a criminal defendant on every conceivable count under a charge by breaking an offense into distinct parts. Say, for example, a defendant breaks into an abandoned property for the purposes of theft. In a horizontal overcharge of such a case, a prosecutor might break up the single act into several distinct criminal transactions, such as criminal damage to property, unlawful entry, vandalism, and so forth in hopes that a defendant will be encouraged to admit guilt on some of them and so settle the matter before a costly and time-consuming trial (Graham 2014). To put this in a more general form, the possibility of a shift to negotiation encourages an exaggeration of one’s commitments or one’s questions upon the commitments of another, with an eye toward making one’s interlocutor face the prospect of extra burdens should they insist on continuing critical discussion. The second kind of strategy, vertical overcharging, consists in charging a defendant with the more serious version of the type of crime of which they’re accused, including “lesser included charges” (Alschuler 1968; Graham 2014). For instance, a defendant charged with shoplifting (a misdemeanor) may face a much stiffer charge of larceny (a felony), with the understanding that the more serious charge includes the less serious one. While the horizontal strategy threatens work for a defendant should they continue a dispute, this kind of overcharge raises the stakes for the defendant if they lose the factual dispute, and so encourages them to abandon critical discussion for negotiation. Whatever the method, horizontal or vertical, overcharging can be used as a tactic to pressure a shift to negotiation in the broader context of a critical discussion. As with the cases involving negotiation shifts discussed by van Laar and Krabbe, it is a second-best option, born of frustration and economy: frustration because the parties are at an impasse (with a defendant claiming innocence); economy because negotiation is a potentially more efficient way to settle the matter (Schulhofer 1992). The risks brought about overcharging are manifold. The purpose of the overcharge is to provide leverage in bargaining over the provability of a factual claim—a defendant’s guilt or innocence. Given the inherent dangers of a criminal trial, innocent people may be induced to plead guilty in the hopes of exchanging a certain shorter sentence for the risk of a longer one. This is especially true of indigent or otherwise vulnerable criminal defendants, who cannot possibly hope to match the resources of the state at trial. A poignant example of this was the case of Aaron Swartz, who, facing horizontal and vertical overcharging for a simple theft case, took his own life (Dean 2013).18 Cases such as this, 18 Aaron Swartz broke into a computer database at MIT in 2010. He altered an IP address on a server and proceeded to download 4.8 million articles from JSTOR, an academic database. In response, the federal 30 DAVID GODDEN AND JOHN CASEY where the imbalance in resources is extreme, illustrate that the parties to a dialogue at risk of shifting may not be equally poised to negotiate on fair terms. This imbalance is bound to affect not only the outcome of the negotiation but also the tenacity with which one interlocutor may stick to their guns in order to bring negotiation about or, critically, forestall it.19 The risks of illicit shifts to negotiation are not merely borne by the parties to the negotiation. The United States Supreme Court has noted that the bargain obscures the critical factual question of guilt or innocence: It is critical that the moral force of the criminal law not be diluted by a standard of proof [or a procedure for conviction] that leaves people in doubt whether innocent men are being condemned. It is also important in our free society that every individual going about his ordinary affairs have confidence that his government cannot adjudge him guilty of a criminal offense without convincing a proper factfinder of his guilt with utmost certainty. (In re Winship, 397 U.S. 358 (1970)). Negotiations, like critical discussions, can have onlooking audiences as well as stakeholders and participants. The concern, in this case, is that the onlooking audience will view bargains with cynicism—when, for reasons of expedience or economy, the parties walked away from resolving a factual dispute. Here again, the danger of the outcome of such a negotiation taking on the status of a discursive proxy (see above) is particularly acute. Although criminal defendants may negotiate to a lighter sentences, they will nonetheless be treated as though the claims pleaded to are true. The truth of these claims can then weigh significantly on their future lives. That a claim is discovered to be true or bargained for is lost as the claims go forward. For this reason, such bargained claims pollute judicial and social discursive ecosystems. But there are more dangers in the shift to bargaining (Wright and Miller 2002). The complexity of the negotiation even more strongly dilutes the core epistemic questions that initially motivated the exchange. The step to negotiation in criminal cases such as these involves more than two simple parties who differ with regard to a question. Schulhofer (1992: 1987) notes that “the real parties in interest (the public and the defendant) are represented by agents (the prosecutor and the defense attorney) whose goals are far from congruent with their principals.” The agents can have different incentives and different objectives in the bargaining. It might be objected that overcharging takes place within a structured legal system where negotiation is clearly a part of the process so it is not representative of dialogue shifts of the kind van Laar and Krabbe have in mind. It is true that the legal system in which overcharging has a place has a formal structure. We would argue, however, that the underlying incentives and pathologies obtain whenever there is the possibility of a shift from critical discussion to negotiation. As overcharging poignantly illustrates, argumentation over matters of fact is not cost free (Paglieri 2009; Paglieri and Castelfranchi prosecutors charged Swartz with 13 counts of violating the computer fraud and abuse act. He faced 35 years in prison. https://slate.com/technology/2013/01/aaron-swartz-suicide-prosecutors-have-too-much-power-tocharge-and-intimidate-people-for-their-crimes.html 19 We have so far argued that the hazards in shifting to negotiation are borne out on the negotiation end. It seems to be the case, however, that there are clear hazards for the critical discussion stage. 31 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION 2010; Hample, Paglieri, and Na 2012). The budget of any prosecutor’s or public defender’s office will amply illustrate this point. More generally, arguing is costly in terms of time, effort, and lost opportunities. Moreover, it is risky. Losing an argument may damage one’s relationships or one’s status as an arguer. All of these can be leveraged in a kind of strategic maneuvering that we have called an argumentative overcharge: i.e., the temptation to adopt more standpoints or more extreme standpoints in order to raise the costs of arguing to an opponent. And this kind of strategic maneuvering structurally disadvantages the discussant with the worse “exit options” (Stevens 2019) for the dialogue. This will be the party with fewer resources they can devote to resolving the disagreement at trial, or who can worse bear the loss of the argument, or who has the worse consequences for walking away with the dispute unresolved (for a time). As prosecutorial overcharging demonstrates, such overcharges need not be fabricated or dishonest in order to achieve their intended result. Now, it also might be objected that this kind of case does not fit the proposed model precisely because the outcome is coerced, thereby failing to meet van Laar and Krabbe’s “freely chosen” condition (ii) of compromises. Yet, if this kind of strategic maneuvering were really deemed coercive in the contexts in which it actually occurs, such as the example just discussed, then it would be prohibited. Yet, it is not. Indeed such plea bargained settlements can only be considered as just outcomes of the judicial process insofar as they are freely chosen by the accused, whose statement of guilt as part of plea agreement involves an explicit acceptance of the factual claim forming the substance of the charge. And, in argumentative situations that are less strictly regulated, such as those of everyday argumentation, or political policy debate, there is very little in the regulatory environment to prevent its employment. Given what we take to be the manifest normative unacceptability of such maneuvering, we would suggest that the transition from persuasion to negotiation dialogue is a far riskier transition than has been thus far recognized, and one deserving of greater regulation in institutional contexts and greater skepticism or reluctance in the contexts of much ordinary argumentation. 7 Conclusion Proponents of a permissive policy for shifting to negotiation from an unresolved critical discussion in non-trivial cases face the following dilemma. Either the compromise policy position resulting from the shifted-to negotiation is a discursive proxy for a resolved policy position in contexts of implementation, or it isn’t. If it isn’t, then the fallacy of middle ground is thereby avoided; the principal of retrospective evaluation doesn’t properly apply, and the fallacy of bargaining is also thereby avoided. But any such shift is imprudent— indeed inept—since shifted-to negotiations cannot provide the (extra-)discursive goods required (i.e., actionable action policies) to motivate the shift. If it is, then, while the dialogue shift can be motivated by the shifted-to dialogue’s promise to supply the required (extra-)discursive goods, the fallacy of middle ground is not avoided. As such, in all cases where the methodological priority of argumentation holds (i.e., where the dialogue shift is not trivial), the principle of retrospective evaluation will apply but not be satisfied, and shifted-to negotiations will commit the fallacies of bargaining and abandoning of discussion / commitment. In either case, the dialectical shift from critical discussion to negotiation is unwarranted, either because it is imprudent or because it is fallacious. 32 DAVID GODDEN AND JOHN CASEY As a final point, consider two policy proposals about the permissibility of shifting from a critical discussion to a negotiation: Permissive policy: Having given no quarter in a non-redundant critical discussion that has failed (at some stage) to produce a resolution about some matter of policy (or other matter having at least some practical aspect), critical discussants may shift to negotiation in order to reach an implementable compromise policy position. Prohibitive policy: Having given no quarter in a non-redundant critical discussion that has failed (at some stage) to produce a resolution about some matter of policy (or other matter having at least some practical aspect), critical discussants may not shift to negotiation in order to reach an implementable compromise policy position. (Each policy might be qualified to be a default permission or prohibition.) Suppose, that each of these incompatible policy proposals has proponents. Suppose further that these proponents engage in an extended critical discussion about the rational acceptability of their respective standpoints, the result of which is that each gives no quarter. Neither is willing to assent to the other’s view nor to retract theirs, and each maintains their criticisms of the other’s standpoint. Suppose now that it is proposed that they adopt the permissive policy in this instance—that they should enter into a negotiation in order to move past their differences. After all, starting from their initial standpoints, they might bargain their way to some middle ground where each gives a little and gets a little. Should either side be rationally inclined towards adopting the permissive proposal in this case? The permissive proposal is obviously inconsistent with the prohibitive policy standpoint, so any disinclination of its proponent is understandable. Yet, the proposal is consistent with the permissive policy standpoint. Why should this deadlocked critical discussion over incompatible policy proposals not be one of those times when a shift to negotiation is permissible? Suppose there is a practical impetus to settle the matter, because, say, we have to decide on how to revise our procedural rulebooks for argumentation. Should these considerations at all incline the proponent of permissiveness to adopt negotiation as a discursive means to reach some concession position that would settle the difference of opinion as to the rational acceptability of their own standpoint? We suggest that reason should not so incline them. That this difference of opinion remains, seemingly, unresolved is not a reason to give up arguing and begin bartering. If we are correct, then each proponent acts reasonably in their refusal to follow the permissive policy as a means of settling their otherwise deadlocked critical discussion about the acceptability of the permissive policy. While the proponent of prohibitiveness acts consistently in their refusal, it is not obvious that the proponent of permissiveness does also, even though we find that they would be acting rightly. There are times when cooperation should not involve compromise, and argumentation is one of them. 33 NO PLACE FOR COMPROMISE: RESISTING THE SHIFT TO NEGOTIATION References Alschuler, A. 1968. The prosecutor’s role in plea bargaining. University of Chicago Law Review 36: 50-112. American Bar Association. 2015. Criminal justice standards for the prosecution function. https://www.americanbar.org/groups/criminal_justice/standards/ProsecutionFunctionFourthEdition/ Benjamin, M. 1990. Splitting the difference: Compromise and integrity in ethics and politics. Lawrence KS: University of Kansas. Clifford, W.K. 1877. The ethics of belief. Contemporary Review 29: 289-309. Reprinted in: 1886. William K. Clifford, lectures and essays, eds. L. Stephen and F. Pollock. London: Macmillan and Co. Dean, J. https://verdict.justia.com/2013/01/25/dealing-with-aaron-swartz-in-the-nixonian-tradition Jan 25 2013 Godden, D. 2016. Commentary on Jan Albert van Laar and Erik C.W. Krabbe ‘‘Splitting a difference of opinion.’’ In Argumentation, Objectivity, and Bias: Proceedings of the 11th International Conference of the Ontario Society for the Study of Argumentation (OSSA), May 18–21, 2016, eds. P. Bondy and L. Benacquista. Windsor, ON: OSSA. https://scholar.uwindsor.ca/ossaarchive/OSSA11/papersandcommentaries/173/. Godden, D. 2010. The importance of belief in argumentation: Belief, commitment and the effective resolution of a difference of opinion. Synthese 172: 397-414. Graham, K. 2014 Overcharging. Ohio State Journal of Criminal Law 11(2): 201-724. Hample, D., F. Paglieri, and L. Na. 2012. The costs and benefits of arguing: Predicting: the decision whether to engage or not. In Topical themes in argumentation theory: Twenty exploratory studies, Argumentation Library vol. 22, eds. F.H. van Eemeren and B. Garssen, 307-322. Dordrecht, Springer. In re Winship, 397 U.S. 358, 364 (1970). Paglieri, F. 2009. Ruinous arguments: Escalation of disagreement and the dangers of arguing. In: Argument Cultures: Proceedings of OSSA 09, ed. J. Ritola, 1-15. Windsor, ON: OSSA. https://scholar.uwindsor.ca/ossaarchive/OSSA8/papersandcommentaries/121/ . Paglieri, F., and C. Castelfranchi. 2010. Why argue? Towards a cost–benefit analysis of argumentation. Argument and Computation 1: 71-91. Schulhofer, S. 1992. Plea bargaining as disaster. The Yale Law Journal 101(8): 1979–2009. Stevens, K. 2019. The roles we make others take: Thoughts on the ethics of arguing. Topoi 38 (4): 693-709. Tindale, C.W. 1997. Fallacies, blunders, and dialogue shifts: Walton’s contributions to the fallacy debate. Argumentation 11: 341-354. van Eemeren, F.H., and R. Grootendorst. 2004. A systematic theory of argumentation. Cambridge: Cambridge UP. van Eemeren, F.H., and R. Grootendorst. 1984. Speech acts in argumentative discussions: A theoretical Model for the analysis of discussions directed towards solving conflicts of opinion. van Laar, J.A. 2019. Middle ground: Settling a public controversy by means of a reasonable compromise. In: Studies in critical thinking, ed. J.A. Blair, 69-80. Windsor, ON: Windsor Studies in Argumentation. van Laar, J.A., and E.C.W. Krabbe. 2016a. Splitting a difference of opinion. In Argumentation, Objectivity, and Bias: Proceedings of the 11th International Conference of the Ontario Society for the Study of Argumentation (OSSA), May 18–21, 2016, eds. P. Bondy and L. Benacquista. Windsor, ON: OSSA. http://scholar.uwindsor.ca/ossaarchive/OSSA11/papersandcommentaries/128/. van Laar, J.A., and E.C.W. Krabbe. 2016b. Reply to David Godden’s commentary on ‘‘Splitting a difference of opinion.’’ In Argumentation, objectivity, and bias: Proceedings of the 11th international conference of the Ontario Society for the Study of Argumentation (OSSA), May 18–21, 2016, eds. P. Bondy and L. Benacquista. Windsor, ON: OSSA. http://scholar.uwindsor.ca/ossaarchive/OSSA11/papersandcommentaries/64/. van Laar, J.A., and E.C.W. Krabbe. 2018a. Splitting a difference of opinion: The shift to negotiation. Argumentation 32: 329-350. doi: 10.1007/s10503-017-9445-7 van Laar, J.A., and E.C.W. Krabbe. 2018b. The role of argument in negotiation. Argumentation 32: 549-567. doi: 10.1007/s10503-018-9458-x van Laar, J.A., and E.C.W. Krabbe. 2019a. Criticism and justification of negotiated compromises: The 2015 Paris Agreement in Dutch parliament. Journal of Argumentation in Context 8: 91-111. doi: 10.1075/jaic.18009.laa 34 DAVID GODDEN AND JOHN CASEY van Laar, J.A., and E.C.W. Krabbe. 2019b. Pressure and argumentation in public controversies. In Proceedings of the ninth conference of the International Society for the Study of Argumentation, eds. B. Garssen, D. Godden, G. Mitchell, and J. Wagemans, 702-709. Amsterdam: SicSat. Walton, D. 1992. Types of dialogue, dialectical shifts, and fallacies. In Argumentation illuminated, eds. F.H. van Eemeren, R. Grootendorst, J.A. Blair, and C.A. Willard, 133-147. Amsterdam: SICSAT. Walton, D. 1998. The new dialectic: Conversational contexts of argument. Toronto: University of Toronto Press. Walton, D., and E.C.W. Krabbe. 1995. Commitment in dialogue: Basic concepts of interpersonal reasoning. Albany, NY: SUNY Press. Walton, D., C. Reed, and F. Macagno, F. 2008. Argumentation schemes. Cambridge: Cambridge UP. Weinstock, D. 2013. On the possibility of principled moral compromise. Critical Review of International Social and Political Philosophy 16(4): 537-556. Wright, R., and M. Miller. 2002. The screening/bargaining tradeoff. Stanford Law Review 55(1): 29–118. 35