One of the criticisms directed at the decision-theoretic model discussed in Chapter 2 (§ 2.2) is that the probability (or odds) that the claim is true cannot capture all of a decision maker's epistemic concerns. The question persists as to whether some additional epistemic concept or factor needs to be put into the mix of decision making. A similar concern presents itself for criteria of discriminatory power that are focused not on posterior odds assessments but rather solely on assessments of the strength of the publicly produced evidence (§ 2.3.1) or that are theory focused in that they require the fact-finder to compare the probability (or “plausibility”) attaching to the claimant's theory of the case with that attaching to the defendant's theory of the case or to select the most plausible explanatory theory (§ 2.3.2). It is this quite general concern that gives rise to the question of Keynesian weight.

I begin by motivating this concern through a look at the problem of “naked statistical evidence” (§ 3.1). I then examine the nature and importance of Keynesian weight (§ 3.2), and identify a crucial but underappreciated dimension of Keynesian weight, its excusable preference structure (§ 3.3). In the rest of the chapter, I relate the understanding thereby developed to modern theorizing about informal reasoning (§ 3.4), comment on the possibility of quantifying Keynesian weight (§ 3.5), illustrate how the misunderstanding of the role of Keynesian weight is sometimes attributable to the erroneous assumption that it must figure as merely one ingredient in a holistic judgment of epistemic warrant (§ 3.6), relate it to decision making with belief functions that do not conform with the axioms of probability (§ 3.7), and conclude by relating the solution of the problem of Keynesian weight to the goals of adjudication (§ 3.8).

While I draw on Keynes's insights, this is not a work in intellectual history, and I do not purport to follow Keynes in any detail.