MIT OpenCourseWarehttp://ocw.mit.edu 24.910 Topics in Linguistic Theory: Propositional Attitudes Spring 2009 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.Apr. 28 24.910, Spring 2009 (Stephenson) Part I – Conversation and Common Ground (cont’d) 5. Further Applications 5.1. Semantics / pragmatics of questions 1. [examples of normal information-seeking questions] 2. Is it raining? 3. How are you? 4. Who’s the governor of Iowa? One standard view of the semantics of questions: ¾ The denotation of a question = the set of possible answers to the question (i.e., a set of propositions)  (2 {it’s raining, it isn’t raining}  (3 {The addressee is doing well, the addressee is doing okay, the addressee is doing terribly, …}  (4 {Mitt Romney is the governor of Iowa, Barack Obama is the governor of Iowa, Chet Culver is the governor of Iowa, …} ¾ Observations about questions:  The speaker believes that some member of the set is true [this seems to act like a presupposition]  The speaker believes that the addressee (might) know which one is true, and be willing to tell the speaker ¾ Formally, given a question Q which denotes set A Asking Q does the following:  Adds to the common ground that (at least) one member of A is true [presupposition]  Invites the hearer to assert one of the members of Q ¾ Norm: For a speaker to appropriately ask question Q, where Q has the set of possible answers A, the following must hold:  the speaker can be expected to accept whichever assertion the hearer makes in response to the question  the hearers can be expected not to object to adding to the common ground the proposition that A has some member that’s true ¾ Note: We don’t have to add that the hearer should answer correctly, since this falls under the norm of assertion 5.2. (Indicative) Conditionals 5. If it’s raining, I’ll take an umbrella. 1Apr. 28 24.910, Spring 2009 (Stephenson) One way to think of this: the speaker temporarily adds the proposition that it’s raining to the common ground, and proposes (in that new context) to add the proposition that the speaker will take an umbrella. [After the assertion is over, the proposition that it’s raining is taken back out of the common ground.] Toy example: ¾ 6 worlds in context set:  w1: Rain; Speaker takes umbrella; Umbrella is blown inside-out  w2: Rain; Speaker takes umbrella; Umbrella is not blown inside-out  w3: Rain; Speaker does not take umbrella; Umbrella is not blown inside-out  w4: No rain; Speaker takes umbrella; Umbrella is blown inside-out  w5: No rain; Speaker takes umbrella; Umbrella is not blown inside-out  w6: No rain; Speaker does not take umbrella; Umbrella is not blown inside-out ¾ Step 1: Remove non-rain worlds (w4, w5, w6) temporarily from the context set, giving the following temporary context set:  w1: Rain; Speaker takes umbrella; Umbrella is blown inside-out  w2: Rain; Speaker takes umbrella; Umbrella is not blown inside-out  w3: Rain; Speaker does not take umbrella; Umbrella is not blown inside-out ¾ Step 2: Proposes to remove from this derived context set all of the worlds where the speaker does not take an umbrella (namely, w3). ¾ If the assertion is accepted, we get a new derived context set:  w1: Rain; Speaker takes umbrella; Umbrella is blown inside-out  w2: Rain; Speaker takes umbrella; Umbrella is not blown inside-out ¾ Add the non-rain worlds back in, giving the new context set:  w1: Rain; Speaker takes umbrella; Umbrella is blown inside-out  w2: Rain; Speaker takes umbrella; Umbrella is not blown inside-out  w4: No rain; Speaker takes umbrella; Umbrella is blown inside-out  w5: No rain; Speaker takes umbrella; Umbrella is not blown inside-out  w6: No rain; Speaker does not take umbrella; Umbrella is not blown inside-out ¾ Result [if conditional is accepted]: we’ve removed from the context set worlds where it’s raining and the speaker doesn’t take an umbrella Note: This view of conditionals is highly influential but controversial, and many (perhaps most) linguistic semanticists take a different view. 5.3. Exclamations: Ouch, oops, etc. One simple way to think of these: they are purely presuppositional – that is, add something to the common ground without giving the addressees a fair chance to reject it. (But in order to say one of these things, it must be the case that the hearers can be expected not to object to adding it to the common ground.) On this view: 2Apr. 28 24.910, Spring 2009 (Stephenson) ¾ Ouch: presupposes that something has caused the speaker pain [etc.] ¾ Oops / Oh Sh*t! / etc.: presupposes that the speaker has made a mistake 5.4. Expressive Content Examples: that bastard Sam, nonrestrictive relative clauses (maybe), the f**k-ing TV (and variations), … 6. That bastard Sam got a new job.  asserts that Sam got a new job  Adds to the common ground that the speaker thinks Sam is a bastard 7. Turn off the damn T.V.!  = imperative expressing ‘turn off the T.V.’  Adds to the common ground that the speaker has a negative attitude towards the T.V. Two ways to think of these: ¾ They contain presuppositions about the speaker’s attitude (for example ‘the speaker thinks that Sam is a bastard’) ¾ They [??] put the speaker on record as wanting to add a “judgment” to the common ground (for example, ‘Sam is a bastard’) but presumes that the hearers will not accept it. (Perhaps it gives hearers an opportunity to accept it.) 5.5. Other conversational games: Guessing games 8. The prize might be in my right hand or it might be in my left hand. You have to guess! ¾ Assume:  ‘Speaker believes the prize might be in their right hand’ ≈ ‘it’s compatible with the speaker’s beliefs that the prize is in their right hand’  [and similarly for the left hand] ¾ Also assume: If ‘the prize might be in the speaker’s right hand’ is added to the common ground, what happens is that the interlocutors simply check to make sure there are worlds in the context set where the prize is in the speaker’s right hand [and similarly for left] Reason this kind of example is puzzling: Presumably the speaker knows what hand the prize is in, and so they can’t possibly believe both of these might statements. One way to think of this situation: The norm of