Normative ought-statements are often said to come in two sorts. Consider (1) (1) The president ought to re-join the Paris climate accord.

On a standard way of construing the meaning of ought, (1) means something like the state of affairs wherein the president rejoins the Paris climate accord is better than one where he does not. Evidently, a different kind of meaning is exhibited by (2).

(2) John ought to return the lost money.

It’s not simply that the world where John returns the lost money is better than one where he doesn’t; (2) enjoins John to perform a particular action– return the money. Philosophers moved by this distinction point out that ought-sentences like (2) play a special kind of role in our practical deliberations. (1) and (2) exemplify the distinction between evaluative and deliberative ought, respectively. (The distinctions between ought-to-be/ ought-to-do (Castaneda [1975], Feldman [1986]) and relativized/ unrelativized oughts (Grice [2001]) are in the conceptual neighborhood.)

There is no consensus on how to account for this distinction in meaning. Schroeder [2011] offers an ambiguity thesis, contending that the difference in meaning is due to the syntactic difference between a raising ought and a control ought. On his proposal, deliberative ought is a control verb that takes two arguments, an agent and a property, as opposed to a single, propositional argument. This provides a principled basis for the distinction, but the claim that ought is a control verb has been disputed on empirical grounds. Another proposal, due to Chrisman [2012], suggests that deliberative ought can be given a non-ambiguity-theoretic analysis with Kratzer’s modal semantics if the conversational background against which ought is interpreted makes explicit the relevant agent carrying out the action. However, this proposal fails to give us a precise recipe for distinguishing the two senses.

I will give my own account of the distinction between deliberative and evaluative ought without positing a lexical ambiguity. My contention is that it owes to a difference in the logical form of ought sentences and to pragmatic mechanisms of exophoric reference. The analysis will make use of two independently needed grammatical resources for seemingly unrelated problems in the semantics of modals; the source of future-orientation in modal sentences with eventive prejacents, and a composition principle for so-called “external arguments” in verbal semantics, due to Kratzer [1996]. With this we have a robust, uniform, and linguistically motivated distinction between deliberative and evaluative ought.

​

References

H. Castaneda. Thinking and Doing: The Philosophical Foundations of Institutions. D. Reidel Pub. Co., 1975.

M. Chrisman. ’ought’ and control. Australasian Journal of Philosophy, 90(3):433–451, 2012.

F. Feldman. Doing the Best We Can: An Essay in Informal Deontic Logic. D. Reidel Publishing Company, 1986.

P. Grice. Aspects of Reason. Clarendon Press, 2001.

A. Kratzer. Severing the external argument from its verb. In L. Z. Johan Rooryck, editor, Phrase structure and the lexicon, pages 109–137. Springer, 1996.

M. Schroeder. Ought, agents, and actions. Philosophical Review, 120(1):1–41, 2011.