Download Dialectical Foundationalism vs. Egalitarianism: Burden of Proof in Discourse and more Exercises Skepticism in PDF only on Docsity!
Shifting the Burden of Proof?
Michael Rescorla
Abstract: Dialectical foundationalists , including Jonathan Adler, Robert Brandom, Adam Leite, and Michael Williams, claim that some asserted propositions do not require defense just because an interlocutor challenges them. By asserting such a proposition, the speaker shifts the burden of proof to her interlocutor. Dialectical egalitarians claim that all asserted propositions require defense when challenged. I elucidate the dispute between dialectical foundationalists and egalitarians, and I defend a broadly egalitarian stance against several prominent objections.^1
§1. Dialectical foundationalism versus dialectical egalitarianism If I assert a proposition and an interlocutor challenges me to defend it, who has the burden of proof? Am I always responsible for defending my assertion, or does the onus sometimes lie with my challenger? Dialectical foundationalists claim that certain propositions have a privileged role in our reasoning with one another, serving as “default” premises that do not require defense just because someone challenges them. Dialectical egalitarians deny the existence of these “default” propositions. The dispute between dialectical foundationalism and egalitarianism recurs throughout the history of philosophy. Dialectical foundationalism underlies Aristotle’s doctrine of “first principles” and Wittgenstein’s discussion of “hinge propositions.” It figures more explicitly in writings of J. L. Austin, Wilfrid Sellars, and Stephen Toulmin. Recent advocates include
Jonathan Adler, Robert Brandom, Adam Leite, Andrew Norman, and Michael Williams. Dialectical egalitarianism was implicitly embraced by the Pyrrhonian skeptics and, during the protocol sentence debate, by Otto Neurath. More recently, Peter Klein has endorsed it, as have and Frans van Eemeren and Rob Grootendorst.^2 I will try to clarify the debate between dialectical foundationalists and egalitarians, to disentangle it from some epistemological debates with which it is often conflated, and to defend a broadly egalitarian stance against some popular objections.
§2. Dialectically basic propositions The debate between dialectical foundationalists and egalitarians centers upon what Sellars calls “the game of giving and asking for reasons”: the activity through which we rationally assess propositions by providing one another with arguments and counter-arguments. I will refer to this activity as reasoned discourse. Systematic study of reasoned discourse stretches back at least to Aristotle, continuing through medieval discussions of the Obligation Game, with recent contributions by Robert Brandom, Alvin Goldman, C. L. Hamblin, Wilfrid Sellars,
(^1) I am grateful to helpful feedback from Jonathan Adler, C. Anthony Anderson, Anthony Brueckner, Kevin Falvey, Warren Goldfarb, Matthew Hanser, Elizabeth Harman, Richard Heck, Jason Newman, Charles Parsons, and NathanSalmon. I am also indebted to the anonymous referees for this journal for comments that helped improve the paper. (^2) L. Wittgenstein, On Certainty , D. Paul and G. E. M. Anscombe (trans), G. E. M. Anscombe and G. H. von Wright (eds.),O. Urmson and G. J. Warnock (eds), (Oxford: Clarendon Press, 1979), pp. 76- (New York: Basil Blackwell, 1969); J. L. Austin, “Other Minds,” repr. in his116; W. Sellars, “Some Reflec Philosophical Papers , 3rdtions^ ed., J. on Language Games,” repr. in hispp. 321-358; S. Toulmin, The Uses of Argument Science, Perception, and Reality , (Cambridge UP, 1958), at pp. 222-228; J. Adler, (New York: Routledge and Kegan Paul, 1963), Belief’s Own Ethics Leite, “A Localist Solution (MIT Press, 2002), at pp. 159-185; R. Brandom, to the Regress of Justification,” Making it Explicit Australasian Journal of Philosophy (Harvard UP, 1994), at pp. 176-178; A. 83 (2005), pp. 395- 421; A. Norman, “Regress and the Doctrine of Original Epistemic Sin,”477-494; M. Williams, “Skepticism,” in J. Greco and E. Sosa (eds), The Blackwell Guide to Epistemology The Philosophical Quarterly 47 (1997), pp., (Malden: Blackwell, 1999), pp. 35- Walter Sinnott-Armstrong (ed), Pyrrhonian Skepticism 69; M. Williams, “The Agrippan Argument and Two Forms of Skepticism, ” in (Oxford UP, 2004); O. Neurath, “Protocol Sentences,” repr. in R. S. Cohen and M. Neurath (eds and trans) 1983); P. Klein, “How a Pyrrhonian Skeptic Might Respond to Academic Skepticism,” in S. Luper (ed), Philosophical Papers: 1913-1946 (Boston: D. Reidel Publishing, The Skeptics: Contemporary EssaysSystematic Theory of Argumentation (Aldershot: Ashgate Publishing, 2003); F. van Eemeren R. Grootendorst, (Cambridge UP, 2004), at pp. 135-140. For discussion of Aristotle, see J. A Barnes, The Toils of Skepticism (Cambridge UP, 1990).
challenges. Call the view that such propositions exist dialectical foundationalism. Dialectically basic propositions provide a foundation of stable, if not immutable, resting points. By asserting such a proposition, I shift the burden of proof to my interlocutor. She can oblige me to defend the proposition only by providing additional considerations, such as reasons to doubt the proposition or reasons to doubt that I am justified in believing it. In this manner, the norms of reasoned discourse insulate certain assertions from gratuitous disputation. Dialectical foundationalists disagree regarding which propositions are basic. Aristotle favored a privileged class of “first principles,” providing an axiomatic foundation for scientific inquiry. Adler appears to claim that all propositions are basic (p. 159).^4 Brandom, Leite, and Williams allow the division between basic and non-basic to vary with conversational context. Leite, who uses the label terminating claim instead of dialectically basic proposition , argues that “a claim is a terminating claim when the defendant correctly and responsibly takes there to be no reason to doubt it” (p. 405). On this view, the line between basic and non-basic propositions varies with the speaker’s epistemic circumstances. Although dialectical foundationalists reject the Defense Norm, they can endorse: The Default-Challenge Norm : When faced with a legitimate challenge to defend an asserted proposition, one must either defend it or else retract it. What counts as a “legitimate challenge”? According to Aristotle, dialectically basic propositions never require defense, so that no challenges to them are legitimate.^5 Two other possible views: a legitimate challenge to a basic proposition requires compelling reason to doubt that proposition, or it requires asserting the proposition’s negation. Brandom entertains each of these two weaker views (p. 178). Many variants are possible.
(^4) He eventually qualifies that strong thesis (p. 160), but only for cases where the “backing” for the challenge is already “evident.”
Dialectical foundationalism is opposed by dialectical egalitarianism , according to which all assertions require defense in light of brute challenges. The Defense Norm enshrines this egalitarian perspective. While the Default-Challenge Norm does not explicitly endorse an inegalitarian conception, it makes room for such a conception by incorporating an extra parameter, legitimacy , that can insulate certain propositions from certain challenges. Dialectical egalitarians hold that all challenges are legitimate, and hence that the Default-Challenge Norm collapses into the Defense Norm. Dialectical foundationalists offer two main arguments for their position. First, they raise the specter of infinite regress.^6 When I defend some assertion with an argument, I produce additional assertions. My interlocutor can challenge those additional assertions, which I may defend with further assertions, and so on. To avoid the ensuing regress, dialectical foundationalists classify certain propositions as dialectically basic. Second, dialectical foundationalists adduce specific propositions that allegedly require no defense in light of brute challenges, at least in certain contexts. Favorite examples include: (a) Propositions about one’s own mental states: I am in pain ; I think that p. (b) Perceptual reports: I see a red cube in front of me ; I seem to see a red cube in front of me ; I saw a red cube in front of me yesterday. (c) Basic autobiographical facts: My brother’s name is Eric. (d) Propositions about one’s observable surroundings: A red rose is on the table (uttered while pointing at a red rose on a table). (e) Propositions that are common knowledge in our linguistic community: George W. Bush is President of the United States. (f) “Hinge” propositions: The world has existed for more than five minutes ; I have a physical body. (g) Elementary logical and mathematical truths. (h) Elementary analytic truths. I intend (a)-(h) to be neither exhaustive nor mutually exclusive.
(^5) See Barnes, The Toils of Skepticism , pp. 120-123.
§3. Indefensible propositions An ancient objection to dialectical egalitarianism is that some propositions cannot be defended through rational argument. “First principles” were supposed by Aristotle to fit this description. Similarly, Toulmin urges that it sometimes “does not even make sense to talk about producing grounds for one’s assertion” (p. 222). The point is not that certain propositions cannot be derived from others. As Wittgenstein observes, “any proposition can be derived from other ones” ( On Certainty , §1). The point is that certain propositions cannot receive rational support from others. The objection seems most plausible for elementary logical and mathematical truths. As Tyler Burge puts it, some of these truths are “so basic that if one understands them, one realizes that they are true. No argument for them could provide them with a justification that adds force that is not already present in understanding them.”^8 Even here, we should not underestimate the potential for rational argument. For instance, contra Toulmin (pp. 225-227), I think that 2 x 2= admits a non-circular argument, as first shown by Dedekind. Eventually, though, we will reach a level so basic that further argument is impossible: If p, then p is a likely example. Elementary analytic truths, if such there are, may illustrate the same phenomenon. Taking Burge’s formulation as our guide, say that a proposition is defensible iff some argument could provide it with justification that adds force to whatever justification is already present from understanding it. A proposition is indefensible iff no such argument is possible. An indefensible proposition admits no justification beyond whatever justification arises from grasping it in the first place. No further propositions one might adduce can increase whatever rational support it already possesses in itself.
(^7) D. Davidson, Inquiries into Truth and Interpretation (Oxford UP, 1984).
If we concede that indefensible propositions exist, then it becomes natural to emend the Defense Norm as follows: The Defense Norm (reformulated) : When challenged to defend an asserted defensible proposition, one must either defend it or else retract it. The reformulated Defense Norm cedes some ground to dialectical foundationalism. It allows that indefensible propositions are dialectically basic. If there is literally nothing one could say to buttress one’s assertion, then one need not defend it. How much ground does the reformulated Defense Norm cede to dialectical foundationalism? That depends on which propositions are indefensible. I submit that virtually all empirical propositions are defensible. This is not to say that any such proposition p is true, or that a given speaker is justified in believing p , or that a given speaker could provide a non- circular argument for p. It is just to say that there exist further propositions that rationally support p. An interlocutor might not believe those propositions. But, if she does believe them, then they provide added justification for believing p beyond whatever arises from mere grasp of p. Recent work on “transmission failure” challenges the thesis that virtually all empirical propositions are defensible. Crispin Wright contends that various “hinge propositions,” such as There is an external world , cannot be defended through non-circular arguments.^9 Inspired by Wittgenstein, he then argues that we are entitled to believe hinge propositions even though we have no evidence for them. I focus on Wright’s claim that hinge propositions are indefensible. Wright critiques the following “Moorean” argument: I have hands; therefore, there is an external world. He urges that one is entitled to believe the premise only if one is independently
(^8) T. Burge, “Concepts, Conceptions, Reflective Understanding: Reply to Peacocke,” in M. Hahn and B. Ramberg (eds), (^9) C. Wright, “Wittgensteinian Certainties,” in D. McManus Reflections and Replies: Essays on the Philosophy of Tyler Burge (ed), Wittgenstein and Skepticism (MIT Press, 2003), pp. 383-396, at p. 389. (New York: Routledge, 2004).
they provide added justification for their conclusions. If so, then they further confirm the thesis that hinge propositions are defensible.^13 I henceforth assume that virtually all empirical propositions are defensible. (I say “virtually” to circumvent notorious cases like “I exist.”) Given this assumption, the reformulated Defense Norm seems more congenial to dialectical egalitarianism than to dialectical foundationalism. Indefensible propositions play a significant role mainly in mathematical and philosophical conversation. They cannot provide a foundation for ordinary empirical discourse. Thus, by treating indefensible propositions as exceptions to the Defense Norm, we induce a fairly minor perturbation in the normative architecture of quotidian rational conversation. Technically speaking, of course, the reformulated Defense Norm is “foundationalist” rather than “egalitarian,” as I defined those terms in §2. Perhaps we should redefine our terms so as to avoid that consequence. For instance, we might redefine “egalitarianism” as claiming that no defensible propositions are dialectically basic. This terminological issue is less important than the substantive question I will pursue: does the reformulated Defense Norm yield a plausible theory of dialectical interaction?
§4. Epistemology and reasoned discourse Dialectical foundationalists often attack egalitarianism by citing alleged undesirable epistemological consequences. Supposedly, egalitarianism entails the implausibly stringent principle that I am justified in believing p only if I can defend p against brute challenges. Thus,
(^13) For defense of this abductivist strategy, see L. BonJour, In Defense of Pure Reason (Cambridge UP, 1998); A. Goldman,Inference to the Best Explanation,” Empirical Knowledge (University of California Press, 1988); J. Vogel, “Cartesian Skepticism and Journal of Philosophy 87 (1990), pp. 658-666; J. Vogel, “The Refutation of Skepticism,” in M. Steup and E. Sosa (eds),criticism of the strategy, see Alston, The Reliability of Sense Perception Contemporary Debates in Epistemology (Cornell UP, 1993) and J. Beebe, “The (Malden: Blackwell, 2005). For Abductivist Reply to Skepticism,” Philosophy and Phenomenological Research (forthcoming).
Toulmin motivates dialectical foundationalism by observing that “I shall not expect to be asked how I know that [my brother’s name is Roger]… Having once learnt my brother’s name, I need no grounds or premises in order to continue knowing it” (p. 224). Such arguments elide epistemic status and normative standing within reasoned discourse. The reformulated Defense Norm says that, if I assert a defensible proposition p that another speaker challenges, then I must defend p or retract it. If I do neither, I engage incorrectly in reasoned discourse. It does not follow that I am unjustified in believing p or that I should suspend judgment in it. Neither consequence follows even from the stronger assumption that I cannot defend p. Claims about my ability to participate correctly in reasoned discourse do not directly entail claims about whether my beliefs are justified. Brandom, Leite, Williams, and many other dialectical foundationalists maintain that epistemic justification intimately involves an ability, at least in principle, to defend one’s beliefs against legitimate challenges. Call this the Justification Thesis. The basic idea, pioneered by Austin, Sellars, Toulmin, and Wittgenstein, is to illuminate the nature and structure of epistemic justification by “reading them off” justificatory transactions between speakers. If we accept the Justification Thesis, then dialectical egalitarianism entails that far fewer of our beliefs are justified than one might have expected.^14 We should reject the Justification Thesis. Following Alston and Audi, we must distinguish the state of being justified from the ability to justify a proposition to other speakers. Justification is a positive epistemic status possession of which does not require the ability to participate a certain way in reasoned discourse. Young children apparently have justified beliefs, but they cannot mount arguments for those beliefs. As Adler notes (p. 182), adults justifiably
hearer has asked the speaker] without explanation” (p. 160). Our question here is not whether the hearer’s challenge is polite. Our question is whether the hearer’s challenge, polite or not, requires a response. The norms of reasoned discourse may well conflict with norms of etiquette, morality, legality, pragmatic utility, etc. Dialectical egalitarians can also admit that there are many contexts in which it seems legitimate not to answer a brute challenge. If I introduce myself to a stranger at a cocktail party, it would be ridiculous for the stranger to retort, “How do you know that your name is Michael? Justify your assertion.” I would surely decline to answer. Even when a challenge is not so absurd, I might cite various reasons for not answering: I am too tired, too busy, etc. These examples do not undermine the reformulated Defense Norm. In each example, I decline to engage my interlocutor’s challenge. I refuse to enter into reasoned discourse. My refusal does not violate the norms of reasoned discourse, any more than a refusal to play chess violates the rules of chess. What if John, while engaged in reasoned discourse, answers a challenge to some assertion by saying: “Let me think about it and get back to you later”? Again, we should not say that John violates the reformulated Defense Norm. As long as John does not invoke the disputed proposition as a premise when defending other propositions, then his response seems admissible. In effect, John postpones a particular episode of reasoned discourse, just as a chess-player might call “time-out” when facing a particularly challenging board-configuration. Hence, the reformulated Defense Norm is not as demanding as it might initially appear. It is a norm of reasoned discourse , not a norm of conversation in general. One can engage in conversation without engaging in reasoned discourse, and one can suspend an episode of reasoned discourse while continuing a conversation.
Do these maneuvers drain the reformulated Defense Norm of all force? They do not. We must distinguish two scenarios in which a speaker leaves a disputed proposition undefended. If the speaker does not rely upon the disputed proposition when defending other propositions, then the reformulated Defense Norm does not apply. We may describe the speaker as either averting, postponing, or ending an episode of reasoned discourse. But if the speaker invokes the disputed proposition while defending additional propositions, the reformulated Defense Norm applies, classifying his linguistic conduct as defective. Thus, the crucial context for deciding between dialectical foundationalism and egalitarianism is one where a speaker bases an argument upon a disputed premise. Are there any disputed, defensible premises that a speaker may legitimately leave undefended? Do certain premises shift the burden of proof from speaker to challenger? Dialectical foundationalists sometimes cite an alleged intuition that certain premises require no defense in light of brute challenges. The trouble is that dialectical egalitarians deny experiencing this intuition, so that the debate descends into stalemate and intuition-mongering. For instance, Brandom (p. 222) writes that [t]here are sentence types that would require a great deal of work for one to get into a position to challenge, such as “Red is a color,” “There have been black dogs,” “Lighting frequently precedes thunder,” and similar commonplaces. These are treated as “free moves” by members of our speech community --- they are available to just about anyone any time to use as premises, to assert unchallenged. Brandom does not explain how he arrives at these examples. He apparently assumes his readers will experience an intuition that they are dialectically basic. That assumption may be fair for Red is a color , which is arguably indefensible. It seems illicit for There have been black dogs and Lightning frequently precedes thunder. We can agree with Brandom that, in practice, speakers do
§6. The pursuit of rapprochement A basic purpose of reasoned discourse is to isolate mutually acceptable premises relevant to the truth of disputed propositions. Mutually acceptable premises provide a neutral evidentiary base for adjudicating disputes. A neutral evidentiary base may not decisively resolve a dispute, but it serves as common ground. By isolating it, speakers achieve what I will call rapprochement. Only by achieving rapprochement do participants in reasoned discourse engage one another rationally. One might say that, if two speakers cannot agree upon any relevant premises, then they succeed only in talking at one another, rather than reasoning with one another. In this sense, the pursuit of rapprochement is constitutive of reasoned discourse. Philosophical theories of dialectical interaction frequently, albeit often implicitly, invoke rapprochement. (a) van Eemeren and Grootendorst urge that a basic purpose of dialectical interaction is consensus , i.e. resolution of the dispute. Rapprochement is a weak form of consensus. It involves agreement regarding relevant premises but not necessarily regarding the main dispute. Moreover, speakers must achieve rapprochement to resolve the main dispute through rational argument. As van Eemeren and Grootendorst put it, “[t]here is no point in venturing to resolve a difference of opinion through an argumentative exchange of views if there is no mutual commitment to a common starting point” (p. 60). (b) Walton and Krabbe develop a theory of dialectical interaction, harkening back to the Greek Sophists, on which each participant aims to persuade the others to adopt his viewpoint. Ralph Johnson develops an alternative theory in the same vein.^16 Rational persuasion requires convergence upon mutually acceptable premises. (c) Many philosophers, including Goldman, argue that a basic purpose of dialectical interaction is convergence to the truth. Achieving this goal entails achieving consensus: if two speakers
(^16) R. Johnson, Manifest Rationality (Mahwah: Lawrence Erlbaum Associates, 2000).
disagree about some proposition p , then one of them must either endorse a falsehood or else withhold assent from a truth. So convergence to the truth entrains rapprochement. As these examples suggest, the pursuit of rapprochement is implicit in many, if not most, philosophical theories of reasoned discourse. It is a fundamental constraint upon fruitful disputation. Any decent theory, whether foundationalist or egalitarian, should treat it as in some sense constitutive of reasoned discourse. In what sense, exactly? That is a difficult question. Luckily, we need not explore it in much detail. I will make a few comments. Consider a tennis player who deliberately misses certain crucial shots so as to let his opponent win. He plays the game deviantly, because he does not make a genuine effort to win. Yet his deviant conduct differs markedly from that of a player who cheats , perhaps by calling a ball out when it is in. The injunction that one try to win is constitutive of tennis, but it is not a rule of tennis. To capture this difference, Joseph Raz distinguishes between a practice’s norms (such as the rules of a game) and its values (such as winning the game).^17 I find it more suggestive to instead speak of constitutive norms and constitutive goals. Winning is a constitutive goal of tennis, while the injunction “Do not call a ball out when it is in” is a constitutive norm. Constitutive norms describe how to engage in a practice correctly. Constitutive goals describe what ends one pursues, qua participant in the practice. One can engage in a practice correctly without pursuing its constitutive goals, as illustrated by our tennis example. But non-deviant participation in a practice requires promoting its constitutive goals. I submit that rapprochement is a constitutive goal rather than a constitutive norm. For instance, suppose a speaker defends a controversial assertion with a non-circular, cogent argument whose premises she knows her interlocutor is unlikely to accept. Her linguistic conduct
Consider a variant upon The Matrix. In the variant scenario, most people justifiably believe they are disembodied brains communally linked in a computer simulation. I do not suppose that they are disembodied brains. I suppose only that they possess compelling evidence, for instance, that they encounter physical anomalies best explained by the computer-simulation hypothesis. I also assume that the relevant evidence is common knowledge. These people could assert, challenge, and retract claims. They could engage one another in reasoned discourse. Clearly, the assertion I have a physical body would not enjoy default status within their linguistic interaction. If anything, it would seem particularly dubious. A brute challenge to it would not elicit interlocutor-deviancy or speaker-immunity intuitions. In response to the proposed scenario, dialectical foundationalists can say that the line between basic and non-basic propositions varies with context. In our own epistemic context, the proposition I have a physical body is extremely well-confirmed. (Needless to say, philosophers disagree about the precise nature of the confirmation.) In the Matrix-style scenario, I have a physical body is disconfirmed. This contrast explains why the proposition is dialectically basic in one scenario but not the other. I claim that, once we acknowledge a role for epistemic context, we do not need dialectical basicness. Egalitarians can explain relevant interlocutor-deviancy and speaker- immunity intuitions just as well as foundationalists. I begin with interlocutor-deviancy. Speakers ordinarily find little difficulty in achieving rapprochement. No matter how deep our disagreements, we usually agree on a large class of “background” propositions that are well- confirmed for normal participants in the conversation. (Cf. Adler, pp. 164-167.) These propositions serve as a dialectical tribunal of last resort. They might not resolve the dispute, but
they serve as common ground. I have a physical body is one such background proposition. Since it is so well-confirmed, I usually feel confident my interlocutor will not challenge it. What if she does challenge it? My dialectical tribunal of last resort under attack, I no longer know how to achieve rapprochement. My interlocutor disputes a premise we ordinarily regard as beyond dispute. What can I say that would satisfy her? Even if I want to answer her challenge, I feel unsure how to do so. If she disputes I have a physical body , which premises will she concede? The main problem here is not my suspicion that my interlocutor believes the disputed proposition. Speakers can play “devil’s advocate”, disputing propositions that they believe. I can conduct a rational argument with such an interlocutor. The main problem is that, even if I suspect my interlocutor of playing devil’s advocate, I do not know what perspective this particular devil advocates. Her challenge is so fundamental that I have no idea which premises she will leave unchallenged. Assume that my interlocutor grasps the broad character of my epistemic circumstances. (Without this assumption, no interlocutor-deviancy intuition arises.) Then she will have anticipated the shock and amazement elicited by her challenge. To assist me in answering that challenge, she must explain her perspective more fully. She must indicate what kind of relevant premise she might accept. For instance, she might present a “skeptical scenario” that I allegedly cannot rule out, such as a brain-in-a-vat scenario. By presenting this scenario, she sets the terms of the debate. She indicates that she will accept certain premises, such as propositions about my apparent perceptual experiences, while simultaneously illustrating why these premises fail to establish the disputed proposition. In this manner, she partially demarcates the boundaries of mutual agreement. Only given some such demarcation can we achieve our common goal, as
