Stalnaker (1978): a Guided Tour (For Beginners)

Stalnaker (1978): a guided tour (for beginners)

Stalnaker, R. 1978. Assertion. Published as chapter 4 in Context and Content (Oxford University Press). Up to the first paragraph of p. 92. Available at:

http://www.princeton.edu/~harman/Courses/PHI534-2012-13/Lepore/stalnaker78.pdf

This fundamental paper defines the dynamic notion of the meaning of an assertion as a context update effect, thus representing the incremental grow of information in the course of a conversation.

a) An act of assertion expresses a proposition – representing the world as being in a certain way.

b) Assertions are made in a context, and sometimes the (propositional) content of an assertion depends on the context in which the assertion is made. (This dependence requires a two-dimensional system, cf. also Kaplan 1978. We will NOT discuss Stalnaker’s two-dimensional system.)

c) An act of assertion is intended to modify the context, in particular the attitudes of the participants.

We need to define content and context.

d) As for content, Stalnakers defines the notion of a proposition as a set of possible worlds (or the characteristic function thereof), taking possible worlds to be primitives;

– the content of an assertion or belief is a set of possible wolds which accord to a certain representation; to locate the actual world in that set is to represent the world as being in a certain way.

– Any two assertions or beliefs will represent the world as being in the SAME way iff they are true in all the same possible worlds.

e) As for the speech context, this must be characterized in terms of the notion of speaker presupposition (p. 321):

“a proposition is presupposed if the speaker is disposed to act as if he assumes or believes that the proposition is true, and as if he assumes or believes that his audience assumes or believes that it is true as well. Presuppositions are what is taken by the speaker to be the common ground of the participants in the conversation”.

Gunlogson 2003, 25: “The mutual beliefs constituting the CG are, crucially, mutual and not just shared. That is, for each p in the CG, each participant is not only taken to believe p but to be aware that other participants believe p as well; and they, likewise, not only share the belief but recognize that the belief is shared. In other words, mutuality involves beliefs about one’s own and others’ beliefs. Beliefs that remain private, and beliefs that the participants happen to have in common without mutually realizing it, are not part of the CG.”

– The context can be represented more fundamentally not as a set of presupposed propositions, but rather as the set of possible worlds that are compatible with all that is presupposed in the common ground. This is called the context set.

f) Why represent the context as a set of possible worlds? Because this allows us to capture the essential purpose of the conversation: to distinguish among alternative possible ways that things may be:

– the presuppositions define the limits of the set of alternative possibilities among which speakers intend to distinguish; these are the worlds that are recognized by the speaker[s] to be “live options” relevant to the conversation [i.e. candidates for being the actual world].

– The purpose of expressing propositions is to make such distinctions.

g) A context is defective when there are discrepancies between what the speaker assumes that his addressees assume, and what the latter actually assume. If the divergences to not affect the issues that arise in the course of the conversation, they go unnoticed and the context is close enough to being nondefective.

Idealizing somewhat, we can take the context sets of the various participants to normally coincide, hence we can talk of the context set of a conversation.

How does an assertion change the context?

h) Nonessential effect: information is added to the common ground as to aspects of the speech act itself: who the speaker is, what language she is speaking etc. (The first information allows the anchoring of the indexical ‘I’). This is because, in general, any obviously observable change in the phiìysical surroundings of the conversation will change the presumed common knowledge (GOAT EXAMPLE).

i) But how does the content of the assertion – the expressed proposition – change the context? The essential effect of an assertion is to SHRINK the context set: “all of the possible situations incompatible with what is said are eliminated”. The asserted content is added to the presuppositions in the CG – provided that nobody rejects the assertion.

j) Wouldn’t it be sufficient to assume that by an act of assertion the speaker is simply trying to get the audience to accept that the speaker himself accepts the expressed proposition as true [namely, that the speaker commits herself to the truth of that proposition]? No, this won’t allow us to represent the conversation as an attempt at establishing a body of common information: the common context set is the playing field, and the moves are either attempts to reduce the size of the set or rejections of such moves by others. The participants have a common interest in reducing the size of the set, but their interests may diverge when it comes to the question of how it should be reduced.

– Three principles governing conversation: essential principles of rational communication (Gricean perspective: conversation as a cooperative activity involving rational agents). We will limit ourselves to the first two.

I: A proposition asserted is always true in some but not all of the possible worlds in the context set.

This is just a principle of informativity. If the proposition is already true in all the worlds of the context set, then adding it to the CG will not reduce the context set in any way. So the assertion is uninformative because its content is, essentially, already presupposed by the participants.

On the other hand, if the proposition is false in all the world of the context set, then it is effectively presupposed to be false. Adding it to the CG will lead to an empty intersection with the context set, so the updated context set will be empty. This corresponds to an information state that contains a contradiction.

There are various ways of reacting to an assertion that violates this principle. A violation may lead the addressees to realize that the speaker’s context set was actually larger than was realized – with respect to that larger context set, the assertion would count as informative. So, the conversational community realizes that the context was defective. Alternatively, the addressees may look for an alternative interpretation of what the speaker said which would not violate the principle.

II: Any assertive utterance should express a proposition relative to each world of the context set, and that proposition should have a truth value in each possible world in the context set.

This principle concerns truth-value gaps, and connects semantic presupposition with pragmatic (speaker) presupposition. Suppose a sentence S semantically presupposes a proposition P – in the sense tha tS expresses a truth or falsehood only if P is true –, then P is presupposed by the speaker, namely it is taken to be part of the CG.

If the semantic presupposition of S is not satisfied, the proposition expressed by S will be a partial function, undefined by some or perhaps all of the possible worlds of the context set.

[My example: “The gorilla living in Hyde Park is sick” is undefined for any possible worlds in the context set in which there is no (unique) gorilla in Hyde Park: in these words, the proposition yields neither True nor False. If it is even presupposed by the addressees that there is no gorilla in Hyde Park, then the proposition expressed will be undefined for all the worlds of the context set. The proposition will succeed in shrinking the context set only if its semantic presupposition is satisfied in all the worlds of the context set, namely, everyone presupposes that there is a (unique) gorilla living in Hyde Park.]

The rationale for principle II is one of completeness of the update effect. If the proposition expressed is not true or false at some possible world in the context set, it will be unclear whether that possible world must be included in the reduced set of not. That is, the addressees will not know how to reduce the context set.

It has been noticed, since Stalnaker (1974) and Lewis (1979), that the presuppositions of an assertion are sometimes not supported by the common ground, but they are ‘tacitly accommodated’ by the addressees in order to satisfy Principle II. Suppose for instance that at a meeting Susan tells her colleagues “I have to leave now, I have to take my dog to a vet”. Her collegues might have been ignorant, up to that moment, of the existence of a (unique) dog owned by Susan, but they accommodate this information, since this is unproblematic (it does not clash with any other information contained in the CG, and it does not raise any potentially relevant issue).