SS10005: Norms of Belief

Lecture 03: The Truth Norm for Belief Dr Florian Steinberger∗

Dr Ole Thomassen Hjortland†

May 9, 2012

‘[Belief is] a map of neighbouring space by which we steer.’ (Ramsey 1931)

1

Introduction

1.1

Normative concepts for belief • Propositions are true or false; beliefs are not. When we say that an agent has a true belief, we say that the content of the belief is true. • A belief can have a variety of ways of being good, different doxastic virtues: it can be correct, justified, appropriate, or rational (reversely, incorrect, unjustified, inappropriate, or irrational). • An epistemic norm (or doxastic norm) is a norm governing when a belief is correct (incorrect), rational (irrational), and so on.

1.2

The truth norm • “Belief aims at truth”. (Williams 1976, 136) • The truth norm For any agent S, proposition p: S’s belief that p is correct if and only if p is true. Velleman, Railton, Wedgewood

∗ Munich Center for Mathematical Philosophy (MCMP), Ludwig-Maximilians-Universität (LMU),

80539 München, Germany, [email protected]. † Munich Center for Mathematical Philosophy (MCMP), Ludwig-Maximilians-Universität (LMU), 80539 München, Germany, [email protected].

1

1.3

The explanatory power of the truth norm • The truth norm is typically applied to explain other features of belief, for example: (1) evidential norms about justified belief and rational belief as means to an end; (2) doxastic in voluntarism, that we cannot (fully) control which beliefs we form; (3) the difference between belief as an attitude, and other attitudes such as imagining that or supposing that. (Such attitudes involve considering p as true, but do not aim at truth.) (4) More generally, propositional attitudes are individuated by the conditions under which they satisfy normative concepts (Wedgewood 2002, 270)

2

Wedgewood on the aim of belief Main reading: Wedgewood, R. (2002).

2.1

A framework for normative concepts • Wedgewood (2002, 268) characterizes sufficient conditions for what it is for a concept to be normative for a practice.

(NORM) If it is constitutive for a concept C that it plays a regulative role in a practice, then C is normative for the practice. – Example: it is constitutive for the concept legal chess move that it plays a regulative role in chess games. • Commitment: Thus, if C plays such a regulative role, then whenever an agent engaged in the practice judges whether or not C applies to a circumstance, she is committed to regulate her actions according to the judgements. • Rationality: Reversely, it is irrational for an agent engaging in the practice to judge that an action is C, while failing to act accordingly.

2.2

Norms for theoretical reasoning • Strong supervenience: It is impossible for two belief states b and b0 to be identical in all non-normative respect, but differ with respect to a normative concept C. • Thus any two beliefs which differ with respect to C must differ with respect to some non-normative property A. 2

• According to Wedgewood, epistemic norms are necessary, general principles which say, for some non-normative property A and normative concept C, that: whenever belief b has property A, then it satisfies the concept C. • Epistemic norms are part of the essence of belief—they are constitutive of what it is for an attitude to be a belief.

2.3

Fundamental epistemic norms • Primitive vs non-primitive epistemic norms. • Fundamentality: An epistemic norm F is fundamental just in case it is primitive and, for any other epistemic norm E , F explains E . • For Wedgewood, a belief is correct just in case it satisfies the fundamental epistemic norm.

2.4

Correctness for credal states • Wedgewood’s main thesis: S’s belief that p is correct if and only if p is true. • A qualification: The norm should apply to all types of credal states, e.g. (i) outright beliefs and partial beliefs; (ii) disbelieving that p; non-believing that p; suspending judgement about p. • For Wedgewood, for any p, it is neither correct nor incorrect to suspend judgement about p. • Note: this brackets a number of trouble cases where suspending belief might have independent epistemic value, e.g. vagueness, reference failure, paradoxical sentences.

2.5

The Fundamentality Thesis • Wedgewood’s fundamentality thesis: The truth norm is not only a correctness norm for belief—it is fundamental. • The truth norm explains epistemic norms for rational belief. • Note that it is normally assumed a true belief can be irrational (recall Sven, the amateur chess player); and, correspondingly, that a rational belief can be false (recall Magnus, the chess grand master).

3

2.6

Objections

Two objections to the claim that the norm of rational belief can be (fully) explained by the truth norm. • Objection 1: Rationality must be at least in part a matter of practical considerations. Desires, needs, interests cannot be part of the explanation if the truth norm entirely explains the rationality norm. • Objection 2: The correctness norm ranks attitudes to a true proposition p as follows: believing p > suspending judgement about p > disbelieving p. But the ranking does not determine a unique scale for measuring how much worse it is to suspend belief about p than to believe p, when p is true. Or how much better it is to disbelieve that p than to suspend judgement about p, when p is false. Thus, the truth norm does not have enough structure to fully determine what it is rational to believe (disbelieve).

2.7

Reply Reply: The truth norm for belief underdetermines the norm for rational belief. There is a number of faithful precisifications of the norm for rational belief. The range of admissible precisifications is narrowed down by the context in which the proposition is considered. The context consists, among other things, of practical considerations. (High stakes vs low stakes.)

2.8

Evidential norms as the means to an end • “Even though irrational beliefs can be correct, the only way in which it makes sense to aim at having a correct belief is by means of having a rational belief.” (Wedgewood, 2002, 276) • Hitting the target is achieved by being guided by rules which are: (1) error-avoiding in the circumstances; and (2) reliably belief-yielding in the circumstances. • A rule is construed broadly: it can be a rule which is followed spontaneously and unreflective. • The rules are rules for revising beliefs, in the sense of either forming new beliefs or abandoning old beliefs. 4

2.9

Rational rules • Recall BonJour’s example about the clairvoyant. • The clairvoyant has a reliable method for forming beliefs, yet it may not be rational to believe that the method is reliable (clairvoyance is a mystical faculty, etc). • “I propose that [...] it is rational for one to follow a rule just in case it is rational for one to believe that the rule is reliable.” (Wedgewood 2002, 280)

2.10

Two objections

(1) The requirement on rational belief is too strict: an agent S can form a rational belief by a reliable method (in a set of circumstances) without any beliefs about the method itself (e.g, infant’s perception). Reply: The requirement is only that it is rationally permissible for S to believe that the rule is reliable, not that S actually believes it to be reliable. (2) There is a vicious regress: for it to be rationally permissible for S to believe that a rule is reliable, there must be a further rule with which we could form the rational belief about the first rule.

3 3.1

Against the truth norm The Exclusivity Objection • Shah (2003), Owens (2003), Steglich-Petersen (2009), McHugh (2011). • Truth is not the aim of belief since it is never the case that this aim is weighed against other aims (e.g. practical goals). • Deliberation on what to believe does not involve non-epistemic considerations about our practical interests. • Aims are typically non-exclusive, e.g. I might weigh up my goal of becoming a great general against my goal of being a good husband. • “Now such purposes interact with each other in certain familiar ways, so if a subject really does form a belief with the purpose of forming it only if it is true, his pursuit of that goal should be constrained by his other goals and objectives in (something like) the usual fashion.” (Owen 2003, 295)

5

3.2

The Triviality Objection • “Trouble arises, however, in the case of the truth norms since condition C [the normative condition], in this case, refers to the truth of a proposition. For instance, in order to follow [the truth norm], the subject has to have a belief about whether or not p is true: if S considers whether p and comes to form the belief that p is true then [the norm] gives her reason to conclude that she ought to believe that p. In other words, [the norm] tells her that if she believes that p, she ought to believe that p. It is rather obvious that no guidance can be had from this. If, instead, S considers whether p and she forms the belief that not p, [the truth norm] gives her reason to conclude that it is not the case that she ought to believe that p. This, clearly, does not give her any reason to revise her belief that not p. Hence, there cannot be truth-norms (of whatever form) that serve to guide our belief formation.” (Glüer & Wikforss 2009, 44) • Bykvist & Hattiangadi (2007), Whiting (2010), Hattiangadi (2009), Wedgewood (ms.).

6

Bibliography Bykvist, K. and Hattiangadi, A.: 2007, Does thought imply ought?, Analysis 67(296), 277–285. Glüer, K. and Wikforss, Å.: 2009, Against content normativity, Mind 118(469), 31– 70. Hattiangadi, A.: 2009, Some more thoughts on semantic oughts: a reply to daniel whiting, Analysis 69(1), 54–63. McHugh, C.: 2011, Belief and aims, Philosophical Studies pp. 1–15. Owens, D.: 2003, Does belief have an aim?, Philosophical Studies 115(3), 283–305. Shah, N.: 2003, How truth governs belief, The Philosophical Review 112(4), 447– 482. Steglich-Petersen, A.: 2009, Weighing the aim of belief, Philosophical studies 145(3), 395–405. Wedgwood, R.: 2002, The aim of belief, Noûs 36, 267–297. Whiting, D.: 2010, Should i believe the truth?, dialectica 64(2), 213–224. Williams, B.: 1976, Problems of the self: philosophical papers 1956-1972, Cambridge Univ Pr.

7