Lecture 26: Introduction to Pragmatics
1
Goals Today
✤
Motivate the need for a uniform solution to various challenges to the
adequacy of our translations of the basic logical particles like “and”,
“or”, “not”, “if . . . then. . . “ and “some” and “all.”
✤
Outline the elements of Grice’s theory of conversational implicature-a key idea of pragmatics-- and see how it provides for such a uniform
solution, as well as indicate some problems.
2
Meaning of the Logical Particles
and Problem Cases for Translation
✤
By “logical particles” we just
mean words like “and”, “or”,
“not”, “if . . . then. . . “ and
“some” and “all.”
✤
In some sense, the entire course
thus far has been concerned
with articulating one specific
answer to the question:
✤
what is the meaning of the
logical particles?
3
✤
Our answer: the meaning of the
logical particles is given by
their translations into
propositional and predicate
logic via ∧, ∨, ¬, →, ∃, ∀, and
this answer is adequate because
the translation preserves truthconditions.
✤
Let’s motivate pragmatics by
looking at cases where our
translations ostensibly don’t
preserve truth-conditions.
Problems with ‘And’
✤
Consider:
(1) Claire returned from work
and she fell asleep.
(2) Claire fell asleep and she
returned from work.
✤
If we translate into
propositional logic, we get:
(1’) w ∧ s
(2’) s ∧ w
✤
The problem is that know that
(3) w ∧ s ⊨ s ∧ w
✤
But intuitively: the situations in
which sentence (1) seems true-those situations in which she
first returns from work and
second falls asleep-- are those in
which (2) seems false.
✤
So given (3), (1’) seems like a
bad translation of (1), and (2’)
seems a bad translation of (2).
✤
4
How can we respond to this
problem?
Problems with ‘And’, Con’t
✤
Consider:
(1) Claire returned from work
and she fell asleep.
(2) Claire fell asleep and she
returned from work.
✤
✤
If we translate into
propositional logic, we get:
(1’) w ∧ s
(2’) s ∧ w
✤
✤
The problem is that know that
w∧s ⊨ s∧w
5
You might respond: shouldn’t
we read (1) with a suppressed
and isolated “then”, as in:
(1’’): Claire returned from
work and then she fell asleep.
This response seems a little adhoc: if you’re allowed to fix a
bad translation by adding
words to the original, are there
any limits to what you can add?
Are you allowed to delete
words too when it suits you?
Problems with ‘Or’
✤
Example 1:
Lily: Where is the coffee shop?
Susan: on 6th St or on 7th St.
✤
Example 2:
Lily: When is the exam?
Susan: On Monday or Tuesday.
✤
Example 3:
Lily: Where is Claire?
Susan: at work or at the coffee
shop.
6
✤
So Lily asks Susan a question
and Susan responds with p∨q.
✤
Intuitive Thought. Lily should be
able to infer that (i) Susan does not know p and
that (ii) Susan does not know q.
✤
What could possibly account
for Intuitive Thought besides the
meaning of “or”? After all,
that’s the only thing that’s
common to all Susan’s answers.
Problems with ‘Or’, Cont’
✤
So Lily asks Susan a question
and Susan responds with p∨q.
✤
Intuitive Thought. Lily should be
able to infer that (i) Susan does not know p and
that (ii) Susan does not know q.
✤
What could possibly account
for Intuitive Thought besides the
meaning of “or”? After all,
that’s the only thing that’s
common to all Susan’s answers.
7
✤
You might think: we should
amend the truth-conditions for
“or” so that “p or q” means that
one of them happens but we
don’t know which.
✤
But this won’t work in general
because in other settings we do
know which. For instance,
perhaps I am working on an
algebra problem, and I infer
from my knowledge that x<5 to
my knowledge that x≤5.
Problems with ‘All/Some’
✤
Suppose I say:
✤
✤
✤
1. All of the flowers at the
park are blooming.
✤
1.’ ∀x (Fx →Bx)
✤
2.’ ∃x Fx
Should’t you infer that:
✤
✤
✤
We would translate:
2. There are some flowers in
the park.
But trouble arises when we
translate into predicate logic in
the usual way:
8
But we know in general that 1’
can be true but 2’ be false. For
instance, consider Fx= “x is a
unicorn.” Since there are no
unicorns, for each a in the
model Fa is false, and hence we
have that Fa →Ba is true.
Problems with ‘All/Some’, Con’t
✤
Suppose I say:
✤
✤
1. All of the flowers at the
park are blooming.
You might respond: let’s just
change the rules so that if 1’ is
true then 2’ is true, where:
✤
1.’ ∀x (Fx →Bx)
✤
2.’ ∃x Fx
Should’t you infer that:
✤
✤
✤
2. There are some flowers in
the park.
✤
But trouble arises when we
translate into predicate logic in
the usual way.
But if 1’ is true, then so is:
✤
✤
9
3. ∀x ((Fx ∧ Gx) →Bx)
Trouble with G = ￢B.
Problems. Many Problems.
. . . . . . Grice to the Rescue
✤
✤
It looks like we’re just
scratching the surface here. This
is bad because all these
problems piled one atop the
other suggest that our
translations are in general bad.
Hence, this suggests that the
approach we’ve adopted is
simply misguided: we’ve failed to
identify the meanings of the
logical particles like “and” and
“or” and “some/all.”
10
✤
Grice invented the field of
pragmatics in part to respond to
worries like these.
✤
Grice’s idea was that we can
uniformly dissipate all of these
problems once we realize that
there are certain inferences
which are generated not only
by the meanings of the logical
particles, but by general rules
or maxims of conversation.
A Motivating Example for Grice’s
Theory
✤
✤
As a warm-up, consider:
✤
(1). [Lily:] I need to print my
homework.
✤
(2). [Susan:] There’s a
computer lab in the Social
Science Tower (SST).
✤
But note that it’s not logic itself
which guarantees that (2)
implies (3). After all, (3) is true
even if the computer lab didn’t
have a printer.
✤
What makes us think that (3) is
the case when we hear Susan
say (2)? It’s that we know that
Susan was trying to answer (1).
✤
In the Gricean terminology,
Susan was being cooperative.
Obviously Lily should infer
✤
(3). Susan believes there’s a
printer in SST
11
Grice’s Cooperation Principle
✤
The Cooperation Principle: “[...] Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged”
(Studies in the Ways of Words p. 26, italics added, cf. discussion in last
full paragraph of Gamut vol. 1 p. 202).
✤
In short, the cooperative principle just says that you should try to
make sure that your contribution to the conversation actually furthers
the common aims and goals of the conversation you are engaged in.
✤
Grice amplified on the content of this with some further maxims.
12
Grice’s Maxims
✤
✤
Maxims of Quantity:
1. Make your contribution as informative as is required (for the
current purposes of the exchange)
2. Do not make your
contribution more informative
than is required
✤
Maxims of Relation:
1. Be relevant.
✤
Maxims of Manner:
1. Avoid obscurity of expression.
2. Avoid ambiguity
3. Be brief
4. Be orderly
Maxims of Quality:
1. Do not say what you believe to
be false.
2. Do not say that for which you
lack adequate evidence.
(Studies in the Ways of Words pp.
26-27, cf. Gamut vol. 1 pp.
204-205).
13
Defn: Conversational Implicature
✤
✤
✤
Suppose two people, S and L,
are talking with one another,
where S stands for the speaker
and L stands for the listener.
✤
Definition. q is a conversational
implicature of S’s saying p if
✤
(i) prior to saying this, S has not
violated Grice’s maxims
✤
(ii) if ¬q, then S violates Grice’s
maxims by saying p.
Suppose that S says p.
Then, relative to this situation,
we define the following notion
of conversational implicature
(cf. Levinson Pragmatics p. 113,
or Grice, Studies in the Ways of
Words pp. 30-31):
✤
14
(iii) S and L know (i)-(ii) A Motivating Example for Grice’s
Theory, Con’t
✤
Recall:
✤
✤
✤
(1). [Lily:] I need to print my
homework.
(2). [Susan:] There’s a
computer lab in the Social
Science Tower.
Obviously Lily should infer
✤
(3). Susan believes there’s a
printer in SST.
15
✤
Now, (3) is a conversational
implicature of S’s saying (2).
✤
For, suppose (3) was not true.
✤
So S doesn’t believe there’s a
printer in SST.
✤
Seems like S’s saying (2) is
not relevant, since (after all)
Lily was asking after a
printer.
Conversational Implicature and
Problems with ‘And’
✤
(1) Claire returned from work
and she fell asleep.
✤
Then (3) is a conversational
implicature of S’s saying (1).
✤
Now suppose S said (1) and
that S has been cooperative.
✤
For, suppose (3) was not true.
✤
Since (3) isn’t true, the order of
events was different.
✤
But then S, in saying (1), wasn’t
being orderly, and hence
violated maxim of manner.
✤
Then consider:
✤
(3) Susan believes Claire
returned from work and then
she fell asleep.
16
Conversational Implicature and
Problems with ‘Or’
✤
So Lily asks Susan a question
and Susan responds with p∨q.
✤
Intuitive Thought. Lily should be
able to infer that (i) Susan does not know p and
that (ii) Susan does not know q.
✤
What could possibly account
for Intuitive Thought besides the
meaning of “or”? After all,
that’s the only thing that’s
common to all Susan’s answers.
17
✤
Indeed, both (i) and (ii) are
conversational implicatures of
Susan’s saying p∨q.
✤
For, suppose (i) was not true.
✤
Then S does know that p.
✤
Then S probably believes p∨q
because she believed p. And if
that’s right, the maxim of
quantity should have directed
her to reply with p itself.
Conversational Implicature and
Problems with ‘Some/All’
✤
Suppose S says to you:
✤
✤
(1). All of the flowers at the
park are blooming.
Should’t you infer that:
✤
✤
✤
✤
Then S thinks that there are no
flowers in the park, and this is
no doubt her admittedly trivial
reason for believing (1).
✤
So the maxim of quantity
directs S to have communicated
rather that there are no flowers.
(2). There are some flowers in
the park.
For, “S believes (2)” is a
conversational implicature
of (1).
18
For, suppose S did not believe (2). Since S was
presumably at the park, she
should believe not-(2).
Next time . . . .
✤
Next time we’ll look at whether
the Gricean theory can give us a
response to the paradoxes of
material implication.
19
Goals Today
✤
Motivate the need for a uniform solution to various challenges to the
adequacy of our translations of the basic logical particles like “and”,
“or”, “not”, “if . . . then. . . “ and “some” and “all.”
✤
Outline the elements of Grice’s theory of conversational implicature-a key idea of pragmatics-- and see how it provides for such a uniform
solution, as well as indicate some problems.
20
Ω
21