I have been thinking a lot about Kamp/Vlach sentences recently, for instance
(1) A child was born who will become the ruler of the world
Sentences like (1) are puzzling for a number of reasons. For instance, they cause trouble for Priorean tense logic. However, I have a question about (1) that is at least to some extent unrelated to the issue of whether Priorean tense logic could be revived. It seems that (1) requires for its truth that there is someone who was a child but isn't anymore and who will in future become the ruler of the world. In other words, it does not require for its truth that there is someone who is currently a child. But relative clauses are movable. So let's move the relative clause:
(2) A child who will become the ruler of the world was born
(2) could be true, it seems, if there is someone who is currently a child, who was born, and who will become the ruler of the world. As far as I know, moving the relative clause should not be able to affect truth-conditions. What is going on?
Here are some other cases that are structurally similar to (1)
(3) Tomorrow we will talk to a war veteran who survived for more than a week without food or water. (x is a war veteran now)
(4) In 5 months we will go to a demonstration, and we will talk to a protester who went to this school. (x is not a protester now)
(5) In 2030 Alice will be sitting on a chair that Clinton was sitting on in 1990. (x is a chair now)
(6) In 2030 Alice will marry a single father who was abused as a child. (x is not a single father now)
(7) On Friday I will be meeting a former student of mine who is going to China. (x is a former student now)
(8) (oracle) In two years you will go on a long trip, and you will meet another traveler who was a dear friend of yours many years ago (x is not a traveler now)
(9) The baby who was born four months early will finally be able to go home tomorrow (x is a baby now)
(10) The baby who was born four month early will begin college next week (x is not a baby now)
You get the idea. But what causes the different readings? Prima facie, at least, it seems to have something to do with our knowledge of the general background assumptions about properties and persistence. If that is true, then perhaps (2) will not cause trouble for Priorean tense logic after all (the relative order of the tenses is not a problem for Priorean tense logic, as relative clauses + noun phrases can be assumed to move out of the scope of the matrix clause, as argued by Ogihara).
Tuesday, July 18, 2006
Kamp/Vlach Sentences
Subscribe to:
Post Comments (Atom)
7 comments:
(1) A child was born who will become the ruler of the world.
About (1) you say,
"(1) requires for its truth that there is someone who was a child but isn't anymore and who will in future become the ruler of the world"
I can't see it. Why couldn't (1) be an answer to the question (2),
(2) What happened yesterday?
In that case, (1) might be true though the child who was born yesterday is still a child. Maybe I'm missing something.
Hi Mike,
Thanks for your comment. Maybe you are right, in which case the example doesn't pose a problem for theories that treat the tenses as sentential operators. But there are other cases that will make the same point, for instance:
(oracle) You will go on a long trip, and you will meet a traveler who was an old friend of yours many years ago.
Here x is definitely not a traveler now.
Brit, yes I was wondering about the other examples. You write,
"(oracle) You will go on a long trip, and you will meet a traveler who was an old friend of yours many years ago."
And you observe about this sentence,
"Here x is definitely not a traveler now"
But ask the oracle this question: "where is this old friend now, as we speak?"
Oracle: "right now, he's traveling round London"
That's not a coherent answer? He is a traveler now, though he is likely not on the journey during which he meets you. But even that might not be true. He might be on the very journey during which he meets you--though perhaps very early on in that journey. No?
That seems right. Here is what I was worried about, at least initially. Many linguists and some philosophers seem to think that Kamp/Vlach sentences give us a reason to abandon theories of tense that treat the tenses as sentential operators. The reason is not that the sentences cannot get translated. The reason is that the translations are messy and ad hoc (according to them).
For example, "A child was born who will become the ruler of the world" can be translated with Kamp's N operator. But that's somewhat messy. Without the N operator, it seems that we can only translate it as follows (where P is past and F is future):
Ex(P(child x & born x) & F(ruler x))
or on an alternative reading:
Ex(child x & P(born x) & F(ruler x))
Good translations (I think). But many think they are somewhat ad hoc. Some people have even told me that they do not think the truth of the sentence requires the child to exist now (imagine a scattered individual who persists but does not exist now PS! What do you think about that?).
Taking the tenses to be quantifiers supposedly solves the problems. I should probably add another post on this later; I haven't seen enough recent discussions of these examples.
Brit, your weaker formulation struck me as correct:
it does not require for its truth that there is someone who is currently a child.
which seems to be not-box, as opposed to the box-not formulation that Mike is objecting to. (In other words, there could be someone who is currently a child, but there doesn't have to be.) But I thought that your claim is that (2) does require for its truth that there is someone who is currently a child (as in (9)), in which case it still makes your point: (1) and (2) have different readings.
Thanks for the clarifications, Matt (and Mike). Yes, I should have said that it does not require for its truth that there is someone who is a child now. (The reading I had in mind does not rule out that there is a child). But I still think these kinds of sentences are potentially ambiguous.
So, I think that (using old-fashioned Priorean tense logic) we do get the following two possible readings (they may not be salient in all cases):
Ex(P(Cx & Bx) & F(Rx))
Ex(Cx & (P(Bx) & F(Rx))
Or do you think it only has the first reading?
Thanks, Alan!
Right, so
(A) A child was born who will become ruler of the world
(B) A child was born who would become ruler of the world
seem to have different truth-conditions. But you're suggesting that (A) can be read in the same way as (B), right?
If so, then there is a further reading, namely:
P(Ex(Cx & Bx & F(Rx)))
The problem for the Priorean is to explain why these readings are not ad hoc or unwieldy.
Post a Comment