Permission to believe: descriptive and prescriptive beliefs in the Clifford/James debate

Lawrence, Christopher

Permission to believe: descriptive and prescriptive beliefs in the Clifford/James debate

Doctoral Thesis

2020

Abstract

William Clifford's ‘The Ethics of Belief' proposes an ‘evidence principle': …it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence (1877, 1879:186). Its universal, absolutist language seems to hide something fundamentally correct. We first argue for excluding prescriptive beliefs, and then consider further apparent counter-examples, culminating in more restricted, qualified wording: If anything is morally wrong, then it is morally wrong within the category of descriptive belief to believe anything knowingly or irresponsibly on insufficient evidence in the absence of any conflicting and overriding moral imperative except when the unjustified believing is outside the believer's voluntary control. We test this against William James's counter-claim for qualified legitimate overbelief (‘The Will To Believe', 1896, 2000), and suggest additional benefits of adopting an evidence principle in relation to the structured combinations of descriptive and prescriptive components common to religious belief. In search of criteria for ‘sufficient' and ‘insufficient' evidence we then consider an ‘enriched' Bayesianism within normative decision theory, which helps explain good doxastic practice under risk. ‘Lottery paradox' cases however undermine the idea of an evidence threshold: we would say we justifiably believe one hypothesis while saying another, at the same credence level, is only very probably true. We consider approaches to ‘pragmatic encroachment', suggesting a parallel between ‘practical interest' and the ‘personal utility' denominating the stakes of the imaginary gambles which Bayesian credences can be illustrated as. But personal utility seems inappropriately agent-relative for a moral principle. We return to Clifford's conception of our shared responsibilities to our shared epistemic asset. This ‘practical interest we ought to have' offers an explanation for our duty, as members of an epistemic community, to get and evaluate evidence; and for the ‘utility' stakes of Bayesian imaginary gambles. Helped by Edward Craig's (1990, 1999) ‘state-of-nature' theory of knowledge it provides a minimum threshold to avoid insufficient evidence and suggests an aspirational criterion of sufficient evidence: Wherever possible, a level of evidence sufficient to support the level of justification required to be a good informant, whatever the particular circumstances of the inquirer.

Keywords

Philosophy

Reference:

Collections

PhD / Doctoral

Full item page