A brief note about some things I found puzzling on reading Friederike Moltmann’s “Events, Tropes, and Truthmaking,” Philosophical Studies, 2007, 134: 363–403.
1. Moltmann characterizes the semantic value of a nominal as the truthmaker for the associated sentence. E.g., the value for “John’s run” = the truthmaker for “John ran”. Her aim seems to be to avoid having the value for “John’s run” determined by the semantic value, or values, of “John ran,” as it would be on e.g. a neo-Davidsonian account.
2. What is the truthmaker for “John ran”? Moltmann considers appeal to the following condition:
(TP) An entity e makes true a sentence S iff necessarily e exists entails S is true.
However, she settles for appeal to only the “left-to-right” version, which I presume means:
(TP*) An entity e is truth-maker for a sentence S only if, necessarily, if e exists, then S is true.
(For present purposes, I think we can ignore the use of sentences rather than propositions and the consequent issues about the modality in the r.h.s.)
3. But as far as I can see, (TP*) allows no means of making a transition from “John ran” to a particular truth-maker. We can use (TP*) to filter out candidates. But, first, in any case with respect to which there’s a candidate truthmaker f (presumably, it may be a merely possible f) that also meets the r.h.s. condition (so, f exists entails S is true), (TP*) won’t determine which of e or f is the truthmaker for S. Second, with respect to any e that meets the necessary condition in (TP*), it will be left open that it fails some additional necessary condition. (Moltmann in fact adds in some additional conditions, but as far as I can tell, they appeal to further characterisations to do with temporal extent, and standard issues raised by first order operators. She doesn’t claim that they combine to furnish a sufficient condition on being the truthmaker for a sentence.)
4. So, what we get is something like this. The semantic value for “John’s run” = the truthmaker for “John ran” ≠ any e such that possibly, e exists and S is not true. That type of negative characterisation is not really satisfactory.
5. I think what Moltmann wants is an addition to the effect that e exists: The semantic value for “John’s run” = an e such that (e exists and e exists entails “John ran” is true). So, the semantic value for “John’s run” = an e such that e exists entails “John ran” is true. The problem with this is that it runs into the types of difficulties that figure in people wishing to avoid commitment to the right-to-left direction of (TP). In particular, unless suitable restrictions are imposed, too many distinct things will count as values for “John’s run”: any finite stretch of John’s running, the singleton set of any such finite stretch, the singleton set of the singleton set, and so forth. (It may be that those difficulties are less severe in this context, because the subject matter is contingent nominalisations about events or tropes.)