Re: Question 15 Challenged

From: Harnad, Stevan (
Date: Wed May 21 1997 - 11:25:25 BST

> From: Cherry, Sandra <>
>> (15) According to Sperber & Wilson's relevance theory, which would be
>> most relevant to you now:
>> A. a statement that gives you the answer to question 15?
>> B. an explanation of how to compute the answers to all 36 questions in
>> 18 minutes of calculation?
>> C. an explanation of how to compute the answers to 18 of the questions
>> in 9 minutes of calculation?
>> D. a list of the answers to 6 of the 36 questions?
>> E. ***a list of the answers to all 36 of last year's questions?
> I though that Sperber & Wilson's relevance theory was that you get the
> most RELEVANT information with the least amount of cognitive effort.

No, it was that relevance = most implications/effort.

[Information would be another approach, but that would be defined by
the alternatives that matter to you (as in the 6-button lunch machine
example), and your uncertainty about alternatives; then information
would reduce that uncertainty. That is the classical definition of
information, but not S & W's definition of relevance.]

> So I thought it was A: a statement that gives you the answer to
> question 15, but it could also be both B and C depending on how much
> time you had for the test

How much time you have for the test is not relevant: It is how many
implication, weighted by the amount of effort (= time):

A: 1 (0 minutes effort)
B: 36 (18 minutes effort)
C: 18 (9 minutes effort)
D: 6 (0 minutes effort)
E: 36 (0 minutes effort)

Of these, using S & W's criteria alone, E is the only correct answer.

I will, however, discard this question as no one seems to have gotten

> but on the other hand it just says to all 36/18 questions should we
> take it that they are the questions that we are meant to be answering
> or could they be other ones? Sandra

Again, that's not relevant to S & W's criteria for relevance:
Only implication count and effort cost are relevant. I had said
in lecture that theit criteria do not seem to capture relevance, because
it is surely the relevance of implications, not their quantity, that
matters, whereas S & W are saying relevance = number of implication
(weighted by cost in effort).

But the question will not be counted.

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:53 GMT