Grice and conditionals (long, no Greek)
Jonathan Robie
jonathan at texcel.no
Sat Sep 19 18:49:09 EDT 1998
At 01:51 PM 9/19/98 EDT, Paul S. Dixon wrote:
>Would you care to summarize Paul Grice's thesis, or at least what were
>some of his main points which stood out to you and why? I suppose I
>could get the book and read it, and if you can wet my appetite perhaps
>I will, but a brief thought regarding it would be delicious.
<WARNING> The rest of this message is relevant to natural language use of
conditionals, but makes no direct reference to Greek. I'm not liable to
follow up on this thread after this message. </WARNING>
Here's a quick take on Paul Grice's "Studies in the Way of Words", tailored
to those who are studying Greek conditionals ;=>
First off, chapter 4 of this book is all about indicative conditionals in
natural language, so it is probably well worth consulting. I should warn
folks that chapter 4 is best read after reading chapters 2 and 3, since
without having read these chapters, you really don't have the background to
read chapter 4. I also suggest people *not* try to read chapter 1 first
unless they have some background in epistemology or philosophy of language.
After reading some of the other chapters, chapter 1 makes more sense.
To understand anything Grice says, you have to know that he believes there
*is* a logic to natural language reasoning, and he attempts to explain the
valid logic that underlies normal conversation. In chapter 2, Grice says:
"It is a commonplace of philosophical logic that there are, or appear to
be, divergences in meaning between, on the one hand, at least some of what
I shall call the formal devices ... and ... what are taken to be their
analogues or counterparts in natural language -- such expressions as "not",
"and", "or", "if", "all", "some (or at least one)", "the". Some logicians
may at some time have wanted to claim that there are in fact no such
divergences; but such claims, if made at all, have been somewhat rashly
made, and those suspected of making them have been subjected to some pretty
rough handling."
He then goes on to discuss the difference between two traditional ways of
viewing this discrepancy. (1) Formalists basically think that formal
systems are good, and natural language reasoning is inferior. After all,
formal systems are easier to handle clearly. They view those elements of
meaning found in natural language but not in formal logic as imperfections
of natural language. The best approach, therefore, is to construct an ideal
logical language, the formalisms for reasoning in it, and stop trying to
use natural language as the basis for reasoning. (2) Informalists say that
language serves many purposes besides formal logical reasoning, we can know
perfectly well what an expression means without knowing its analysis,
natural language reasoning includes many valid inferences and arguments
that are not expressible in terms of the formal devices, and there is a
need to understand the logic of natural language apart from the formal
devices if you want to know what people are saying (since few people speak
in syllogisms).
Grice then says he is neither a formalist nor an informalist. He maintains
that natural language reasoning is, in fact, valid logical reasoning, but
only if you take into account the conditions that surround conversation.
You have to know more than what is actually stated; you have to know what
is *implied* (he says "implicated", a word I hate), and how things come to
be implied in conversation. For instance, if someone asks how someone else
is doing on their new job, the reply "oh, quite well, I think; he likes his
colleagues, and he has not been to prison yet" does not state that this
person has been to prison, but most of us would assume an implication that
this person may be likely to yield to some temptation associated with the
job, that the colleagues are the kind of people who might lead him to do
illegal things, etc. But none of these things are stated. Why, then, do we
assume these things are implied? According to Grice, because we assume
people don't say something that is irrelevant, and that we can assume, with
validity, that there was a reason for mentioning prison.
Grice goes on to say that we normally assume people are being cooperative
in conversations, and list a number of assumptions that we associate with
normal, cooperative discussion. These are good principles to follow on
B-Greek as well!
Quantity
1. make your contribution as informative as is required (for the current
purpose of the exchange).
2. do not make your contribution more informative is required.
Quality ("try to make your contribution one that is true")
1. do not say what you believe to be false.
2. do not say that for which you lack adequate evidence.
Relation
1. be relevant
Manner
1. avoid obscurity of expression.
2. avoid ambiguitiy.
3. be brief.
4. be orderly.
Grice argues that these are assumptions that underlie most of our
conversation. These principles are violated all the time, for specific
reasons, e.g. when I say "I knew a man with a wooden leg named Sam", I'm
violating the principles "avoid obscurity of expression" and "avoid
ambiguity". But the violation of any of these principles has it's own
meaning - in this case, it means that I must have a reason for being
ambiguous and obscure, and the reason for that is that I'm making a joke.
So to understand the logic of natural language, you have to study the
shared assumptions under which communication occurs, what these assumptions
imply in a given context, and what the violation of any of these
assumptions might imply in a given context. Natural language reasoning is
valid logic, you just have to look at a much bigger picture than just that
which is literally said.
Now I'm pretty rusty on what Grice has to say about conditionals, and the
chapter is not easy reading (Grice never is), so I think I'll leave you
with a teaser, an example that Grice uses in his chapter on conditionals.
If "if p,q", "either not-p or q", and "not both p and not-q" are
equivalent, why are there many "if" statements that sound wrong in natural
English when transformed to one of these other forms? For instance, using
the rules of formal logic, we can transform the sentence "if he rings, the
butler will let him in" into "either he will not ring or the butler will
let him in", but in most conversational contexts, native speakers of
English do not consider these two statements to have the same propositional
value.
Grice goes on to claim that the natural language meaning of the word "if"
does in fact correspond to the meaning defined by truth tables, but a
stronger condition than that found in the truth table is often implied by
the rules of conversation, particularly the Cooperative Principle and the
maxim of Quantity (say what is relevant, and no more). In other words, our
intuition of what is implied by a natural language construct is generally
correct, even when it disagrees with the truth tables. When there is such a
disagreement, it is often helpful to ask ourselves *why* something is being
said in the first place, and we find that the reason for stating something
implies a condition not found in the truth table. In other words, Grice
would say that when you find yourself falling into the Negative Inference
Fallacy, you should ask yourself *why* you are making assumptions not found
in the truth tables, and examine the broader conversational context to
discover these reasons.
But Grice also discusses other, more intricate reasons that our perfectly
valid natural language reasoning for indicative conditionals may seem to
violate formal logic, while remaining perfectly logical if we look at the
larger context in which conversation occurs. Get the book!
Jonathan
___________________________________________________________________________
Jonathan Robie jwrobie at mindspring.com
Little Greek Home Page: http://sunsite.unc.edu/koine
Little Greek 101: http://sunsite.unc.edu/koine/greek/lessons
B-Greek Home Page: http://sunsite.unc.edu/bgreek
B-Greek Archives: http://sunsite.unc.edu/bgreek/archives
More information about the B-Greek
mailing list