Abstract
Theories of propositions as complex acts, of the sort recently defended by Peter Hanks and Scott Soames, make room for the existence of distinct propositions which nonetheless represent the same objects as having the same properties and standing in the same relations. This theoretical virtue is due to the claim that the complex acts with which propositions are identified can include particular ways of cognizing, or referring to, objects and properties. I raise two questions about this sort of view—one about what it means to stand in a propositional attitude relation to a complex act of this sort, and one about which ways of cognizing can be parts of propositions. Both questions turn out to be difficult for the complex act theorist to answer in a satisfactory way.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
There are of course other routes to the conclusions that these pairs of sentences express different propositions. Many have argued that (1)/(2) express different propositions, on the grounds that believing what is expressed by (1) might have a more immediate connection to action than believing what is expressed by (2). And many have argued that (3)/(4) express different propositions, on the grounds that (3) is trivial, uninformative, and a priori, whereas (4) is non-trivial, potentially informative, and a posteriori.
Remember the above remarks about the neutral sense of ‘express.’ Soames, now as in his previous work, takes pairs like (1)/(2) and (3)/(4) to semantically express the same proposition. He’s thus out to provide a pragmatic solution to Frege’s puzzle. But his theoretical apparatus could be used just as well by someone who took (1)/(2) and (3)/(4) to semantically express different propositions.
Soames (2015, p. 46). Emphasis in the original.
There are problems with understanding the nature of these complex acts which parallel in certain ways the traditional problem of the unity of the proposition. The problems to be discussed in what follows would arise even if these problems could be solved. For discussion of the problem of the unity of the proposition when propositions are thought of as complex acts, see Speaks (ms).
This is a slight oversimplification. Given that (2) involves a name, the proposition it expresses, in the relevant sense of ‘express,’ will contain a Millian mode of presentation of the sort discussed below. Since that is irrelevant to issues involving first person propositions, I stick with the simpler rendering of (2P) here.
See, for discussion, Soames (2015, Chaps. 3–5).
Soames (2015, p. 18).
The ‘in general’ qualification is needed, since some propositions are explicitly about cognitive acts—and to judge those, subjects must of course have thoughts about the cognitive acts which those propositions are about.
This is a point of difference between Hanks and Soames. While both identify (at least some) propositions with acts of predication, Hanks takes predicating a property of an object to involve commitment to the object’s having that property. Hence performing the act to which a proposition is identical is, for him, sufficient to judge that proposition. (Unless the act takes place in a cancellation context; see Hanks 2015, Chap. 4.) So Hanks is better placed to handle the special case of judgement than is Soames. But problems parallel to the ones discussed below will, I think, arise for a view like Hanks’ for propositional attitudes other than judgement. I don’t argue the point here, however.
This argument relies on the plausible assumption that causation is transitive. This is not, however, an assumption which everyone accepts. But even if causation is not in general transitive, it will presumably still be possible to set up indirect cases of this sort.
One might reply that we have here a case of causal overdetermination, and that, had the relevant dispositions not already been formed and activated, the act of entertaining (10P) would have formed or activated them. But that won’t be in general true; it could be the case that, had the subject not judged (9P) to be true, she would have entertained but not judged (10P).
Might one solve these problems by adopting a ‘primitive connection’ theory, which takes the ‘in doing’ relation as a primitive of the theory? An obvious worry is that such a theory would not provide a very information account of judgement. But a more serious worry is that such a theory would seem to be open to the same sort of objection as the causal theories just discussed. Cases in which subjects judge p after the p-relevant dispositions are already formed and activated would require the primitive connection theory to be weakened in a way analogous to (3\(c*\)); but this would lead to a closure principle like [RC2].
There’s an additional, but related, problem as well. In some cases, distinct ways of cognizing may not be extensionally equivalent, but it might still be the case that the subject in question has no attitudes toward the objects which he cognizes in one way which he does not have toward the objects which he cognizes in the other way. Let’s say that distinct ways of cognizing are practically equivalent for a subject iff they are so related. If two propositions are representationally identical and differ only in ways of cognizing which are practically equivalent for the subject, then if a subject entertains both and judges on, it will follow from [J3] that she judges both. But this sort of closure principle will be false, for reasons familiar from the closure principles discussed in the main text.
Soames (2015, p. 73).
One way to avoid this problem would be to take acts of thinking about objects in particular ways, not as parts of complex acts, but as modifications of targets of predication. On that sort of view, a proposition like (1P) would not be a complex act with the act of thinking of JS in a first person way as one of its constituents; rather, it would be an act of predicating the property of being on fire of the following entity: JS-as-thought-of-in-a-first-person-way. If the act theorist took this route, it is hard to see how we could get the problem with the Twinkie act started. We could begin by pointing out that the subject does cognize o while thinking of a Twinkie. We could then accept while-eating-a-Twinkie as a mode of presentation of o. But we would have no guarantee that the subject predicated redness of the entity ‘while-eating-a-Twinkie(o).’ If we accept that modes of predication can be modify targets of predications, then the fact that the subject predicates something of one mode of presentation of o will not entail that the subject predicates that property of some other mode of presentation of o. This is the analogue of the fact that, on standard Fregean views, believing a proposition in which an object is cognized via one sense is not sufficient for believing a corresponding proposition in which that same object is cognized via a different sense. The present problem is that once we treat the ‘way of cognizing’ as one sub-act of a complex proposition, then we cannot simply deny that the subject performs the relevant act. This would just be false; the subject described above does, after all, really cognize o while eating a Twinkie. Ultimately, I think that taking ways of thinking as modifications of targets of predication leads to many of the problems that plague more traditional Fregean views, as well as potentially violating the principle [Innocence] discussed above. I discuss the issue briefly in Speaks (ms).
Thanks to an anonymous referee for suggesting this.
Thanks to Kris McDaniel for pressing this objection.
One could also spin this as a worry about self-knowledge. If which propositions I judge depends on the number of trees in Canada, then one might think that I should be in a position to know, by introspection alone, the number of trees in Canada, which is of course absurd. To this the liberal might reasonably reply by pointing out that the issues here are parallel to well-known apparent incompatibilities between various externalist theses about mental content and plausible-seeming assumptions about the scope of introspective knowledge, and that whatever ends up being the correct solution to the latter problem will generalize to the present case.
As above, we are staying neutral on the question of whether the contrast between (11) and (12) is to be explained in terms of semantics or pragmatics, and hence whether the apt/inapt distinction is a distinction between truth and falsity, or something else.
Soames (2015, p. 71).
Soames (2015, p. 71).
Another principle is suggested by Soames’ claim that propositions are ‘purely representational’ cognitive acts. (See, among other places, Soames 2015, p. 16.) One might then point out that producing certain sound waves is not a purely representational act, since it includes details which are not essential to the representational properties of the act. But of course that is also true of the act of referring to Venus by using the name ‘Hesperus,’ since one can use that name to refer to other things and can refer to Venus in other ways. Parallel points apply to the first person, and any ways of cognizing objects or properties.
Soames (2015, p. 17).
References
Caplan, B., Tillman, C., McLean, B., & Murray, A. (2014). Not the optimistic type. Canadian Journal of Philosophy, 43(5–6), 575–589.
Hanks, P. (2011). Structured propositions as types. Mind, 120(477), 11–52.
Hanks, P. (2015). Propositional content. Oxford: OUP.
Soames, S. (2012). What is meaning?. Princeton, NJ: Princeton University Press.
Soames, S. (2013). Cognitive propositions. Philosophical Perspectives, 27(1), 479–501.
Soames, S. (2014). Cognitive propositions. In New thinking about propositions (pp. 91–125). Oxford: Oxford University Press.
Soames, S. (2015). Rethinking language, mind, and meaning. Princeton, NJ: Princeton University Press.
Speaks, J. (ms.). Complex acts and the unity of the proposition.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Speaks, J. Act theories and the attitudes. Synthese 196, 1453–1473 (2019). https://doi.org/10.1007/s11229-016-1219-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-016-1219-5