What are the primary bearers of truth?

Abstract

According to the traditional account of propositional content, propositions are the primary bearers of truth. Here I argue that acts of predication are the primary bearers of truth. Propositions are types of these actions, and they inherit their truth-conditions from their tokens. Against this, many philosophers think that it is a category mistake to say that actions are true or false. Furthermore, even if we grant that token acts of predication have truth-conditions, there are reasons for doubting that types of these actions also have truth-conditions. I respond to these objections in this paper. I also clarify what it means for propositions to inherit truth-conditions from token acts of predication.

Keywords::

• truth
• propositions
• predication
• types
• inheritance

Notes

^1. See (Hanks 2007a, 2007b).

^2. See (Hanks ms).

^3. In performing acts of predication we exercise capacities for noticing similarities between objects and grouping them according to these similarities. Many of these capacities are innate, and without them we would not be capable of thought or speech. As Quine put it, ‘surely there is nothing more basic to thought and language than our sense of similarity; our sorting of things into kinds,’ (Quine 1969, 272).

^4. Echoing Strawson, Searle remarks that ‘the view that it is the act of stating which is true or false is one of the most serious weaknesses of Austin's theory of truth,’ (Searle 1968, 423).

^5. See (Parsons 1990, ch.4). ‘Truly’ the verb modifier should not be confused with ‘truly’ the sentence modifier, as in ‘Truly thou art damned, like an ill-roasted egg, all on one side,’ (Shakespeare, As You Like It).

^6. In assessing the naturalness of these examples it helps to imagine conversations in which they might occur. Here is a possible conversation for (3b):

  Reporter 1: Clinton and Obama both made assertions yesterday about the health   care bill.

Reporter 2: Where did Clinton's assertion take place, and was it true?

Reporter 1: Clinton's assertion occurred at the press conference and was false.

It is easy to devise similar conversations for any of the examples in (3)-(6).

^7. Here is a context in which one might hear (9a):

There was a debate last night in which the candidates spent a lot of time talking about how the government should respond to the current recession. Many years ago, Obama stated that deficit spending can pull the country out of a recession. Obama's statement, which is true, occurred seven times during the debate. Obama himself made it twice, as did five others.

For this to make sense, ‘Obama's statement’ must be read as denoting a type of statement.

^8. See (Hanks 2009) for a survey of the philosophical issues raised by these sorts of examples.

^9. This doesn't work for (10c). ‘*The assertion of Obama asserting that Clinton is eloquent is true’ is terrible. But again, this seems to be due to the gerundive nominal ‘Obama asserting that Clinton is eloquent’. ‘*The movement of him moving attracted attention’ and ‘*The reaction of him reacting to the noise was unexpected’ are equally bad.

^10. Thanks to Michelle Mason for insightful comments that led to this idea. Note that (10b-d) are about token acts of assertion. My suggestion is that these examples sound bad because they foreground the practical aspects of these token actions. Nothing I have said implies that propositions, considered as types of acts of assertion, are subject to practical norms.

^11. This is true for many material properties, but not all. If no tokens of the Union Jack exist then the Union Jack, the type, is not flying from any flagpoles.

^12. In their contribution to this special issue, Tillman, Caplan, Mclean and Murray criticize the inheritance model for not being able to explain why propositions inherit their semantic properties from their tokens. As they understand it, an answer to this question would provide a general principle that tells us which properties of token acts of predication are inherited by which types. Now, obviously it would be nice to have such a general principle, and later on in this paper I will offer a necessary condition on the inheritance of truth-conditions. But the success of the inheritance model does not depend on providing a general principle of this kind. The inheritance model is meant to answer the question about how propositions have truth-conditions, where this is understood as a metaphysical or constitutive question. The question is: what is it for a proposition to have truth-conditions? The answer given by the inheritance model is that a proposition is a type of act of predication, and the possession of truth-conditions by this type is constituted by the possession of truth-conditions by its actual and possible tokens. There is no further need to provide a general principle indicating which properties of token acts of predication are inherited by which types. Consider, by way of comparison, the view that it is primitive that propositions have truth-conditions. This is a different answer to the constitutive question about how propositions have truth-conditions. This view holds that it is a brute, primitive fact that a proposition has truth-conditions, not constituted by any more basic facts. Clearly, the success of this answer does not depend on providing a general principle that tells us which properties of which propositions are primitive and which are not.

^13. See (Hanks 2011, 2013).

^14. When the relation involved is not symmetric then the type must also determine roles for the various objects to play in the relation. Compare asserting that Desdemona loves Cassio with asserting that Cassio loves Desdemona. Both of these assertions fall under the type of act of predicating the relation of loving of the two people Desdemona and Cassio, but this type does not have truth-conditions since it fails to specify who is the lover and who is the one being loved.

Additional information

Notes on contributors

Peter Hanks

Peter Hanks is an associate professor in the Department of Philosophy at the University of Minnesota. He received his Ph.D. in philosophy from U.C. Berkeley in 2002 under the supervision of John Searle. His research is in philosophy of language and the history of analytic philosophy, with a focus on propositions, names, attitude reports, Russell, and early Wittgenstein. His published work has appeared in Nous, Philosophical Studies, Synthese, Erkenntnis, Mind, Philosophy and Phenomenological Research, and the Journal for the History of Analytical Philosophy. He is currently trying to finish a book, Propositional Content, in which he defends the view that propositions are types of actions.