Monday, March 07, 2005

The Meaning of "Meaning"

In order to understand how we use and understand language, whether this is in ordinary conversation or on encountering it in advertising or a political speech, or in any other domain, it is critical to recognize that what we or others say will often have a "meaning" which is at variance with what the sentence uttered "means." Some examples:
1. The sentence "You're a well-behaved group of students!," if uttered to a group of misbehaving students by their teacher, will be understood to have a "meaning" which is the opposite of that of the sentence uttered.
2. A sentence like "Can you reach the flour?," which is interrogative, and therefore would normally call for a "Yes" or "No" response, can be used either to request information (the "default" use we might say) or to request the flour. Compare, the following two situations. In the first, one friend is helping out another who is confined to a wheel chair by arranging his/her kitchen and has no apparent need for flour. In such a case, were he/she to say "Can you reach the flour? "he/she would be understood as attempting to acquire information. In the other situation, a chef who is about to construct a roux and therefore can be expected to need flour turns to his/her assistant and says, "Can you reach the flour?" In this case, the chef would be understood not to be requesting information, but to be requesting the assistant to provide him/her with flour.
3. As a third example, suppose you want to go to a movie, and say to your room mate, "Wanna go to a movie tonight?," and he/she replies, "I've got a physics exam tomorrow." You will undoubtedly take him/her as rejecting your invitation even though the sentence uttered makes no explicit reference either to a movie or to what he/she might be doing tonight. Suppose on the other hand, your roommate were to call his/her tutor and say "I've got an exam tomorrow." In that context, what is being said might be understood as a request for help with the exam.
In order to understand what is going on in these cases, we must make a distinction between sentences and utterances, for what we find in each case is that utterances can have a meaning in context not fully determined by the meaning of the sentence uttered (whether the sentence uttered is spoken, written, signed, as in American Sign Language, or given expression in some other way). A second thing that must be recognized is that when I said, "utterances can have a meaning in context not fully determined by the meaning of the sentence uttered" I used the word meaning in two quite distinct senses.

Talk about meaning is notoriously confounded by the fact that the
words "mean" and "meaning" are ambiguous. Consider, for instance, the
following four sentences:

4 a. What is the meaning of "sesquipedalian"?
b. I did not mean to hurt you.
c. Life without faith has no meaning.
d. Dark clouds mean rain.
In the case of (4a), the speaker seems to be asking for something akin to the accepted meaning, or what we might call the "conventionalmeaning" (sometimes also called the "literal meaning") of the word "sesquipedalian." In the case of (4b), the word "mean" means (i.e., has the conventional meaning of) `intend.' In (4c), "meaning" means (conventionally) `significance.' And, in (4d), "mean" means (conventionally) `signifies.' The question is: what senses of the term "meaning" do we (should we) have in mind when we say that an utterance can have a meaning in context not fully determined by the meaning of the sentence uttered?

My answer is that whereas sentences can be said to have conventional meanings, utterances should be said to have significances instead and therefore my claim that "utterances can have a meaning in context not fully determined by the meaning of the sentence uttered" could be made
more explicit by saying "utterances can have a significance in context
not fully determined by the conventional meaning of the sentence uttered."

The ambiguity of the terms mean and meaning is a quite interesting
one, for three of its senses play a role in how we go about understanding
what others say or write or sign. In a nutshell, in any conversation
or other language use, the significance (meaning as significance) we assign
to any utterance will be a function of the conventional meaning of the
sentence uttered and facts about the context that support assumptions about
the goals (intentions) of the speaker.

Some years ago, Arnold Zwicky and I wrote a paper called Invited Inferences noting that when people say things of the form "P if Q" they often mean to imply "not P if not Q." So, if someone says (5a) they will often mean for you to assume that (5b) is also true.

5 a. I'll give you $5 if you mow my lawn.
b. I won't give you $5 if you don't mow my lawn.

However, this is not always true. Consider the following two conversations:

6. An adult telephones a neighborhood child, with
whom she does not have a particularly close relationship, and the following
conversation ensues:

Adult Hello, Terry. Are you doing yard work this summer?
Child Yes.
Adult I'll give you $5 if you mow my lawn.
Child Okay.

7. A child approaches her mother and the following
conversation ensues:

Child: I need $5, Mom.
Adult: I'll give you $5 if you mow the lawn.
Child: The lawn mower doesn't work.
Adult: Well, I'll give you $5 if you clean out the garage.
Child: Okay.

There is in these two cases a close connection between conventionalsentence meaning, utterance significance, and speaker intention (goal). In the case of conversation (6), the adult has the goal of getting herlawn mowed and she is negotiating with the child in an attempt to inducethe child to mow her lawn. In the case of conversation (7), the child hasthe goal of obtaining $5 and he is negotiating with his parent in an attemptto induce the parent to provide him with the $5. We might say then thatutterance of a sentence which has the conventional meaning of "If P, then Q," which is true of (5a), has the contextual significance of communicating that "If not P, then not Q"" is true if and only if the goal of the initiator of the interaction to make "Q" true. Thus, in the case of the adult who called the neighbor boy to try to induce him to mow the lawn (i. e., make "Q" true), the inference went through, but in the case of the child who wanted to obtain $5 from his mother (i. e., to induce his mother to actso as to make "P" true), the inference did not go through.

The critical thing to note here, and it is as true of how we interpret "You're a well-behaved group of students!" in (1), "Can you reach the flour?" in (2), and "I've got a physics exam tomorrow" in (3), as well as "I'll give you $5 if you mow my lawn" in (6) and (7), is that we engage in a process of utterance interpretation that takes into consideration the conventional meaning of what is said and facts about the context that support inferences as to the
goal of the speaker.

Before concluding this brief discussion, let's consider example (4d). Paul Grice, in his very influential paper, Meaning, referred to the kind of meaning we find in (4d) as "natural meaning." In the clearest cases of natural meaning, there is some causal nexus linking the two things. In (4d), the causal nexus is presumably that the set of circumstances that cause dark
clouds also cause rain (rather than that dark clouds cause rain or that rain causes dark clouds). Hence, dark clouds are a sign or symptom of impending rain. In the case of a sentence like Those spots mean measles the causal nexus is that measles is are directly responsible for the red spots.

There is, of course, no causal nexus linking the word "sesquipedalian" to persons who are given to using long words (though it can be admitted that this word is long). Nor is there a causal nexus involved in the determination of utterance significance. There is, however, a close connection between cases like (4c) and (4d). The difference is that while there is a causal nexus of some sort in the case of natural meaning, there is a rational nexus in the case of utterance significance. The nature of this rational nexus is not fully understood. However, in one way or another the answer will be found through pursuing the ideas advanced by Paul Grice in his seminal paper The Logic of Conversation and John Searle's seminal work on Speech Acts. I, of course, prefer the account I gave in Speech Acts and Conversational Interaction, in which I provide a synthesis of the ideas of Grice and Searle and work in artificial intelligence and conversation analysis.

Tweet This!


Anonymous Anonymous said...

Surprisingly entertaining! I set up an RSS feed so I'll know when you post.

9:24 PM

Blogger The Language Guy said...

Thanks for the sort of kind words. :-)

6:47 AM

Blogger The Language Guy said...

Alex, you are quite right about the effects of stress on meaning in the sense of "significance." And you are right about the problems this poses for machine understanding of ordinary discouse. Typically, such work is done on easy stuff -- written English. That reduces the problems you mention.

Getting a machine to engage in normal conversation is not going to happen in my lifetime.

3:41 PM

Anonymous Anonymous said...

Great Blog! I wanted to share this with you. Are you tired of the Internet Marketing Rat Race? based business home opportunity successful

Check this out when you have time :-)

10:57 PM

Blogger Derk said...

Hello Mr Language Guy,

I'm 19 and am a freshman taking Linguistics as an undergrad course. Just started out this summer, as one might put it.

I find your blog not only enjoyable but more importantly, edifying! I've recommended this blog to my peers, and I hope you don't mind.

Sad to say at this moment I cant think of much to add to the intellectual discourse, but I hope its alright if I say thanks and good going, all the way from Malaysia. =)

6:58 AM

Blogger The Language Guy said...

Thanks for your comments. Of course, you can suggest that your friends give it a look. That's what its for.

8:35 AM

Blogger Luigi Speranza said...

Invited Inference Guy!

You are MORE THAN invited to drop a post at GriceClub.Blogspot!

J. L. Speranza

-- Love all you write

8:10 PM


Post a Comment

<< Home