In this chapter, I advance a non-reductive, proposition-directed, sui generis account of agnosticism called the questioning-attitude account. The questioning-attitude account is non-reductive because it denies that agnosticism is reducible to other mental states like belief, desire, or intention. It is a proposition-directed account because it holds that the object of agnosticism is a proposition, as opposed to a question or another mental state. It is a sui generis account because it holds that unlike belief, which involves an affirming stance towards a proposition, or disbelief, which involves a denying stance towards a proposition, agnosticism involves a distinct questioning stance towards a proposition.

In this chapter, I defend two theses: (1) a narrow version of permissivism, according to which it may be rationally permissible to either believe or be agnostic about P, given some fixed body of evidence, e, or disbelieve or be agnostic about P given e, but it is never rationally permissible to believe and disbelieve P relative to e, and (2) the thesis that agnosticism is the rationally appropriate response to cases of revealed peer disagreement. I conclude the chapter by arguing, contra Michele Palmira, that in cases of agnostic disagreement, the agnostic party is rationally justified in retaining her attitude of agnosticism.

In this chapter, I summarise the central theses defended in my monograph and explain how they fit together to provide us with a more complete picture of the nature and normative significance of agnosticism. I conclude that agnosticism is best conceived of as the mental state of questioning both the truth and falsity of some proposition, P – which can, but need not, be preceded by the mental act of refraining from judging P to be true or false – and is rationally appropriate whenever we take our evidence for and against P to be inconclusive.

In this chapter, I mount a sustained argument against Friedman’s claim that one is agnostic about whether P if and only if one is in an inquiring state of mind about whether P. I reject the claim that an inquiring state of mind entails agnosticism on the grounds that it fails to accommodate cases in which an agent inquires with the aim of ratcheting up an instance of (justified) believing to the status of knowledge or an instance of knowledge to the status of complete certainty. I reject the claim that agnosticism entails being in an inquiring state of mind on the grounds that it fails to accommodate cases in which a subject is agnostic towards P but are unmotivated to inquire about whether P because they believe or know that the question of whether P is unanswerable.

In this chapter, I contend that there is no practical attitude that stands to intending to do X and intending not to do X as agnosticism towards P stands to believing P and disbelieving P. In short, there is no practical analogue to agnosticism. Call this the non-existence thesis. I defend the non-existence thesis against potential objections and highlight some of its implications for the norms governing belief and intention.

In this chapter, I detail some of the main considerations that have led contemporary epistemologist to largely reject non-attitudinal accounts, which conceive of agnosticism chiefly in terms of a combined absence of belief and disbelief, in favour of views that see agnosticism as involving one or more attitudes. I conclude with a brief survey of the major contemporary attitudinal accounts of agnosticism.

In this chapter, I advocate for a bipartite act-attitude account of doxastic neutrality, according to which the mental act of withholding judgement stands to the attitude of agnosticism as the mental act of judging stands to the attitude of belief. My proposed account stands in contrast with that of Matthew McGrath, who argues that there are at least three distinct ways of being neutral – namely agnosticism, refraining from judgement, and suspension of judgement. I argue that suspension of judgement, as conceived of by McGrath, is not a distinct way of being neutral. This leaves only the mental act of refraining from judgement (or what I call ‘withholding judgement’) and the mental state of agnosticism as the two genuine ways of being doxastically neutral.

In this chapter, I defend the thesis that there may be pragmatic reasons to be agnostic. Given that agnosticism is one of the possible outcomes of doxastic deliberation – that is deliberation about whether to believe P – it follows that pragmatic considerations may determine the outcome of doxastic deliberation. However, while I hold that pragmatic considerations may be reasons to refrain from belief, I deny that they may be reason to believe.

In this chapter, I vet various criteria for a satisfactory account of agnosticism that have been proposed in the literature. I conclude that a satisfactory account of agnosticism should satisfy the following seven criteria: (1) Cognitive Contact, (2) Inconsistency, (3) Spontaneity, (4) Termination, (5) Neutrality, (6) Commitment, and (7) Appropriateness.

In this chapter, I apply the criteria vetted in Chapter 2 to the non-attitudinal accounts of Russell and Crawford, the metacognitive accounts of Masny and Raleigh, Wagner’s endorsed-indecision account, and Friedman’s question-directed attitude account. I demonstrate that each account fails to satisfy one or more of the criteria for a satisfactory descriptive account of agnosticism.