According to non-cognitivists about judgments of positive, negative, or neutral value, such judgments are non-cognitive attitudes of favoring, disfavoring, or indifference. The most famous problem for this sort of view—the Frege-Geach problem—is that these attitudes lack the kind of structure needed to make systematic sense of compositionality and inferential relations involving these judgments. My primary focus in this talk will be on a related but lesser-known problem of insufficient structure, raised by Michael Smith. Judgments of value can vary both with respect to the magnitude of positive or negative value attributed to an object and with respect to the degree of certitude of that attribution. By contrast, it seems that attitudes of favoring, disfavoring, or indifference only vary in one corresponding dimension—the strength of favoring and disfavoring involved.

In this talk, I suggest that this problem of insufficient structure might be solved by extending a partial solution to the Frege-Geach problem to the problem of normative certitude. Elsewhere I have argued that a proper general account of acceptance and rejection in acts of judgment makes sense of negation in evaluative and non-evaluative domains alike, without ad hoc assumptions. Here I propose that a corresponding account of partial acceptance and rejection might similarly make sense of degrees of certainty applicable across domains.