Subscribe to Elucidations:

Robert Stalnaker is Professor of Philosophy at MIT.  Click here to listen to our conversation with him.

Philosophers of language have always been interested in how the context of a conversation can affect what the participants in the conversation are saying. (For some examples of how the context of a conversation can affect the meaning of what the conversational participants are saying, see our episode on contextualism.) But in this episode, Robert Stalnaker draws a distinction between two senses in which philosophers of language talk about ‘context.’ In the first sense of the term, the context of a statement is the situation in which it takes place. In the second sense of the term, the context of a statement is the whole of what has been said or established so far: something like ‘the conversation up to that point.’

As an example, consider the conversation I had with my roommate last week. At one point in the conversation, I said, ‘Her tea is delicious.’ The context of that remark, in the first sense of the term, was something like this:

Matt’s apartment, July 21, 2011, 10:53am, with Matt speaking, Matt’s roommate being addressed, and Matt pointing to a cup of tea.

But in the second sense of the term, the context of that remark was something like this:

Matt: What is that?
Roommate: It’s tea, made from my mom’s old recipe. Would you like some?
Matt: I’d be delighted to try it.
Roommate: Here you go.
Matt: Thanks! Mmm…her tea is delicious.

As Robert Stalnaker points out, this second sense of the term ‘context’ is what we’re talking about when we talk about being quoted out of context. To quote someone out of context is to ignore how the meaning of what they said may have been affected by what they said previously.

Stalnaker claims that mostly everything we say is, as it were, incomplete. The sentences we utter leave certain details unspecified, and it’s up to the listeners to fill them in. What’s interesting is that the two notions of context alluded to above correspond to two different ways of filling in that extra information. For instance, when I asked, ‘What is that,’ in order to figure out what ‘that’ meant, my roommate had to be standing in front of me, so as to be able to see that I was pointing at the tea. The question itself contained no information about what I was pointing at. So in that case, my roommate drew on context (in the first sense of the term) to figure out what I was asking.

Later on, you may recall, I responded by saying, ‘Her tea is delicious.’ Here my roommate again had to do a little work to figure out who I was talking about when I said ‘her.’ But he didn’t figure that out by looking at what I was pointing at–he figured it out by recalling that earlier in the conversation, he had mentioned his mother. So in that case, my roommate drew on context (in the second sense of the term) to figure out whose tea was under discussion.

Stalnaker’s most striking claim is that the things we say are unavoidably incomplete. You might think that by speaking very carefully and pedantically, I could always phrase things in such a way that my roommate wouldn’t have to do _ any _ extra work to determine who or what I was talking about. But Stalnaker would like to argue that this is impossible.

Matt Teichman