What is the difference between syntax, semantics and pragmatics?

Question

I would be interested to find more about the way semantics intertwines with pragmatics.

Naturally, I associate semantics with more conventional theories like Russell`s as opposed to Gricean pragmatics.

Could you give a practical example of sentence analyses from both points of views?

Answer 1

In general, semantics relates to what sentences mean, and pragmatics to how they are used. There is no clear boundary line as to where one starts and the other ends, because typically an utterance must be understood by reference to who is uttering it, to whom, on what occasion, in front of what audience, and with what common knowledge. Those who emphasise semantics tend to think of sentences themselves as having a meaning independently of their use, while at the other extreme there are some who hold that one can only speak of the use of a sentence, not its meaning. When Russell was writing, in his earlier years, theories of pragmatics did not yet exist, so he is taken as defending a strict semantic understanding of sentences. Pragmatics came along in the 1950s and 60s with John Austin's book "How to do Things with Words" and John Searle's "Speech Acts", and by Paul Grice's theory of implicatures and the cooperative principle. It is now a standard part of linguistics.

One of the powerful features of Grice's theory is that it serves to explain how utterances can be misleading even when they are true. For example, suppose Alice says, "Bob drove home and had a beer". We understand this to mean something different from, "Bob had a beer and drove home". In propositional logic, these two sentences are just "P & Q" and "Q & P" and so by the commutativity of & they should mean the same. We could try to adjust the semantics and say there is something different about 'and' in English that conveys the sense 'and then', but according to Grice there is no need. The fact that we understand the events as happening in the order stated is because the cooperative principle includes the maxim "be orderly". We are entitled to expect Alice to state the events in an orderly manner and it would be misleading, though not actually false, if she gave the wrong order. We can recognise the difference between the meaning of an utterance and an implicature, because an implicature is cancellable. Alice could say, "Bob drove home and had a beer, but not necessarily in that order". This would be expressly cancelling the suggestion of orderliness. If the semantic view were correct, i.e. that 'and' in English means 'and then', then such a cancellation would be a self contradiction.

Another example. Suppose Alice asks Bob, "Where is Charlie?" Bob knows that Charlie is in the library but he answers, "Charlie is either in the library or the kitchen". P entails P or Q (in classical logic at least) so Bob has not said anything false. But we understand his answer to be defective in some way. He should not say 'either/or' unless he does not know which. Again, we could try to bend the semantics and say that 'or' means 'or ... and I don't know which', but Grice comes to the rescue again and says that Bob's response is true but misleading. Another of the maxims of the cooperative principle is "be as informative as required". Bob is being less informative than he could reasonably be in the circumstances. Alice is entitled to suppose that Bob is following the cooperative principle and that he doesn't know which room Charlie is in. As before, meanings can be distinguished from implicatures by the fact that the latter are cancellable. Bob could say, "Charlie is either in the library or the kitchen; I know which, but I'm not telling". If not knowing were part of the meaning of 'or' then this would be a self contradiction.

Answer 2

Tripartition ( See Carnap in Encyclopedia Of Unified Science, page 145 § Analysis Of Language https://archive.org/details/B-001-015-449/page/n159/mode/1up) :

• syntax deals with the relations holding between linguistic items( words , sentences) ; it tries mainly to answer the question : what are the formation rules of a syntactically correct sentence? what is the structure of a sentence? Syntax explains why the sentence : " Colorless green ideas sleep furiously" is correct, though it has no meaning.

• semantics deals with the relations holding between linguistic items and (1) their sense ( the objective thought they express) (2) their reference ( the objects they denote) ; it mainly tries to answer the question : what are the truth conditions of sentences in terms of sets, and of inclusion or membership relations

Note : on the distinction sense / denotation ( reference), see Frege

• pragmatics deals with the relations holding between linguistic items and their users; it tries to answer the question : how can one express more than one actually says , by using a given sentence in such and such a way in a given context? paradigmatically, pragmatics studies " implicatures".

---

-Suppose a person asks you : " Did you like my cookies?" .

Suppose also that you answer : " yes , some of them".

• From a purely semantical point of view, your answer means :

for some x, ( x is a cookie you made and I've liked x).

In other words, your answer means that the intersection between the things you like and the cookies made by the person is not empty.

Note that " some " does not rule out " all" . Semantically, you did not say that you did not like all the cookies the person has made.

• But from a pragmatic point of view the sentence means more than you actually said.

It means, pragmatically, that there are some cookies you did not like.

Why? Because in case you had really liked all the cookies, by saying only " some" you would not have respected the gricean conversational rule regarding quantity.