Belief and acceptance

What is it to believe something? What sort of mental attitude is belief? On the face of it, folk psychology treats belief as a single kind of mental state, but this surface unity hides important distinctions. Building on ideas from Daniel Dennett, Ronald de Sousa, and Jonathan Cohen, I distinguish partial beliefs (degrees of confidence) from flat-out beliefs (categorical commitments), and argue that flat-out beliefs are premising policies—personal decisions to use certain propositions as premises in truth-directed reasoning. I also distinguish beliefs from acceptances—premising policies adopted for pragmatic rather than epistemic reasons.

This account connects to the two-level mind, where basic beliefs (partial, passive) support superbeliefs (flat-out, active premising policies). It also illuminates dual-process theories, language and thought, and self-knowledge.

Partial belief and flat-out belief

There is a duality in the folk-psychological concept of belief. Sometimes, we talk about belief as coming in degrees—I’m 50% confident it will rain, fairly sure the train will be late, and almost certain that I locked the front door before leaving. These are partial beliefs, which indicate how disposed I am to trust their contents. But we also talk about categorical, yes-or-no beliefs—I believe it will rain, I believe the train will be late, I believe I locked the door. I call these flat-out beliefs.

Philosophers and psychologists have long tried to reduce one type of belief to the other. Perhaps flat-out belief is confidence that exceeds some threshold, or perhaps partial beliefs are flat-out beliefs in explicit probability claims. I argue that neither reduction works. Instead, flat-out beliefs are premising policies. To flat-out believe that p is to be disposed to use p as a premise in conscious reasoning, to assert p when asked, to defend p when challenged, and so on.

This view treats the two types of belief as distinct psychological attitudes, while explaining their relationship. Flat-out beliefs guide conscious, deliberate reasoning and action, while partial beliefs guide unconscious, intuitive responses. The relation between them is one of dependence or realization. A flat-out belief is a disposition to engage in a certain pattern of premising behaviour, and this disposition is the result of a set of partial beliefs and desires. One takes p as a premise (flat-out believes p) because one has high confidence (partial belief) that one is committed to taking p as a premise, has a strong desire (partial desire) to adhere to that commitment, and so on.

Both types of belief play essential roles in the human mental economy.

Not all premising policies are beliefs. Sometimes we adopt propositions as premises for pragmatic or emotional reasons rather than epistemic ones. L. Jonathan Cohen called such states ‘acceptances’. I might accept a scientific theory for research purposes without being convinced that it is true, or accept my friend’s word out of loyalty to them, or accept a religious doctrine as a matter of faith.

I use the term ‘acceptance’ for all premising policies, including ones that constitute flat-out beliefs, and I classify them by their purpose and the reasoning contexts in which they are active. Paradigm flat-out beliefs are formed and sustained for epistemic reasons and are active in reasoning whose overriding aim is to get at the truth. Paradigm pragmatic acceptances are formed and sustained for non-epistemic reasons and are active in reasoning where truth is not the overriding aim, such as defending a friend or working out the consequences of a hypothesis. There are also mixed cases. For example, where the evidence is evenly balanced, pragmatic considerations can tip the balance in favour of a belief commitment. Moreover, if epistemic support for a proposition changes, acceptances can change their status, perhaps without our noticing it. Pragmatic acceptance can transmute into genuine belief, which is relied on in truth-critical contexts, and belief can degrade into pragmatic acceptance, which would not be trusted in high-stakes situations.

This framework helps explain puzzling phenomena. Delusions, I propose, are acceptances maintained to help make sense of anomalous or distressing experiences. Religious faith is acceptance motivated by a spiritual goal that is unsupported by, or even belied by, mundane evidence. Self-deception involves accepting a proposition for the comfort it affords while non-consciously manipulating the evidence to shield it from critical scrutiny.

Active belief formation

Can we form beliefs at will? The traditional answer is that we cannot: belief aims at truth, and we can’t make things true simply by wanting them to be true. So the view that we can believe at will—voluntarism—is necessarily false. I argue that this conflates two claims, which I call strong voluntarism and weak voluntarism. Strong voluntarism is the claim that we can decide to believe things for which we have no epistemic support. Weak voluntarism is the claim that we decide to believe things for pragmatic reasons provided we have a certain level of epistemic support for them. I argue that strong voluntarism is indeed false, but that weak voluntarism is true.

The core idea is that, when the evidence leaves a question open—when it enables belief without compelling it—we can settle the matter of what to believe by accepting the relevant proposition for truth-critical deliberations, which amounts to forming a flat-out belief in it. This isn’t believing at will in the objectionable sense, since the belief is epistemically supported. But it is active belief formation. We are actively making up our mind rather than passively having it made up for us by the evidence.

This supports a broader picture of mental agency within the two-level framework. Forming and executing premising policies are intentional actions—things we do rather than merely things our brains do. In forming flat-out propositional attitudes and reasoning with them, we become active thinkers, whose reasoning is open to reflection, evaluation, and improvement. We construct a supermind for ourselves, whose workings we can articulate and take responsibility for.

Key publications

Mind and Supermind (2004)

Develops the two-level framework comprehensively, showing how flat-out beliefs (superbeliefs) are premising policies realized in partial beliefs and desires. Applies this to akrasia, self-deception, and first-person authority. Argues that the distinction between basic beliefs and superbeliefs resolves longstanding puzzles in philosophy of mind while vindicating folk psychology. Cambridge University Press.

Partial belief and flat-out belief (2009)

Addresses the duality in folk psychology between partial beliefs and flat-out beliefs. Reviews attempts to reduce one to the other and argues none succeed. Defends the premising view: flat-out beliefs are complex behavioral dispositions realized in partial beliefs and desires. Shows how this resolves the Bayesian Challenge—both partial and flat-out beliefs can guide action without conflict, since flat-out states are effective in virtue of underlying partial ones. In F. Huber and C. Schmidt-Petri (eds.), Degrees of Belief. Springer, pp. 75–93.

Deciding to believe again (2007)

Distinguishes weak and strong voluntarism about belief. Strong voluntarism—forming beliefs at will without epistemic support—is necessarily false. But weak voluntarism—forming beliefs directly and for pragmatic reasons, provided one has epistemic support—remains defensible. Argues that actively formed beliefs are premising policies: resolutions to use propositions as premises in conscious reasoning. Sketches a new argument against strong voluntarism while defending weak voluntarism and direct activism. Mind 116: 523–47.

A matter of opinion (1998)

Sets out the case for distinguishing belief from opinion, building on Daniel Dennett’s work. Argues that this distinction can account for tensions within our commonsense concept of belief. Criticizes Dennett’s treatment of the cognitive role of opinion and sketches an alternative that provides a richer account. Philosophical Psychology 11(4): 423–442.

Natural language and virtual belief (1998)

Argues that actively formed beliefs (virtual beliefs, later called superbeliefs) are premising policies, and that forming them typically requires language. Virtual beliefs are realized in basic beliefs about one’s cognitive commitments. Language provides scaffolding for articulating, manipulating, and evaluating these commitments. In P. Carruthers and J. Boucher (eds.), Language and Thought. Cambridge University Press.

Delusions, levels of belief, and non-doxastic acceptances (2011)

Applies the two-level framework to delusions. Argues that delusions are level 2 beliefs—conscious premising commitments—and suggests some may be non-doxastic acceptances adopted for emotional reasons rather than epistemic ones, analogous to religious faith. Explores how rationality constraints operate differently at each level. Neuroethics 4: 153–57.

Adaptive misbelief or judicious pragmatic acceptance? (2009)

Commentary on positive illusions, highlighting the distinction between belief and pragmatic acceptance. Asks whether positive illusions are adaptive misbeliefs or judicious pragmatic acceptances, and discusses how to determine which. Behavioral and Brain Sciences 32(6): 520–21.

Partial belief and flat-out belief