FOM: Pratt on imperatives

[Ben Abraham]

On Sun, 8 Feb 1998, John Mayberry wrote:

> 	Vaughn Pratt has called attention to the important fact that 
> not all sentences are used to convey facts. His example is instructive, 
> and illustrates his point nicely.
> 
> If I say "Now construct the perpendicular to AC at B to intersect EF at 
> G", I have surely spoken meaningfully, yet where is the truth in what I 
> have said?

hate to say the obvious ... but doesn't it say something about the
relationship between AC and EF, in euclidean space for example
there exist B and EF (EF perpendicular to AC) such that the construction
cannot be made, of course there are other statements which are always
statements which are always true, but doesn't that say something also?

> 
> This is a point well made. But it never occurred to me to deny that 
> there are sentences that are used for other purposes than conveying 
> facts. There are interrogatives as well as imperatives, and they do not 
> admit of being pronounced true either. But if *no* declarative sentence 
> can express a truth, I cannot see what the meaning of such  sentences 
> might be. And if declarative sentences don't have meanings, where does 
> that leave imperatives and interrogatives?
> 	Suppose I try to carry out the command in Pratt's example by 
> drawing a line L. Then I might say
> 
> The line L is perpendicular to the line AC at B and intersects EF at G.
> 
> Now if *no* sentence of that form *can* be said to be true, whatever 
> could the original command have *meant*?
> 	When I said that if there is no such thing as speaking truly 
> then there is no such thing as speaking meaningfully, I thought I was 
> pointing out the obvious, not putting forward some subtle point in 
> theoretical semantics.
> --------------------------
> John Mayberry
> J.P.Mayberry at bristol.ac.uk
> --------------------------
> 
>