Review Semantic

 

What is Semantic?

            Semantic theory is the study of the linguistic aspects of the meaning of expressions. It is not always easy to distinguish semantic properties – the linguistics aspects of meaning – from general background knowledge. What is required to understand a sentence like (1), for example, will be a tiny fraction of what most English speakers actually know about witches and princes and charms. The sentence can be understood by people who believe there are witches and by people who do not. That is, understanding this sentence does not require a knowledge of what things, if any, the noun witch refers to, or denotes, nor does it require that we know whether it is true or not. Understanding a sentence does, however, require that we recognize some more basic things, such as certain relations between this sentence and others.

 

            Semantics in linguistic theory

The need for semantics in a theory of human language is controversial. Noam Chomsky (1995) says, for example, “It is possible that natural language has only syntax and pragmatics, it has a ‘semantics’ only in the sense of ‘the study of how this instrument, whose formal structure and potentialities of expression are the subject of syntactic investigation, is actually put to use in a speech community’.” One of Chomsky’s ideas here is that natural language is internal to the speaker, and it is not fundamentally for representing things about the world, things that are true or false depending on matters possibly external to the speaker. Chomsky’s says, “it is not all clear that the theory of natural language involves relations of ‘denotation’, ‘true of’, etc., in anything like the sense of the technical theory of meaning.”

 

Compositionality

Any account of our ability to recognize semantic relations among sentences has to face the fact that this ability is not limited to just a few sentences, but extends through the unlimited range of sentences that any human language allows. The proposal here is the commonsense one that we understand a sentence by understanding its parts and by assembling those parts according to the structure of the sentence. This allows us to recognize the meaning of familiar parts and the meaning of familiar ways of combining parts even in sentences we have never heard before. Intuitively, the meaning of a sentence has, as parts, the meanings of the parts of the sentence, much as a picture has parts corresponding to parts of the whole scene portrayed.

In other words, the rules for determining semantic properties are recursive, just as syntactic and morphological rules are. People understand sentences by understanding the parts of sentences and by recognizing the semantic implications of how those parts are assembled.

 

Entailment

Entailment and possibilities

Understanding our definition of entailment requires getting the right of possibility, and this takes some practice because we commonly refer to many different kinds of possibility. It is an interesting fact that human languages have, in addition to sentences that are true or not depending on the way things are, some sentences that are true simply because of what they mean, so they carry no information about the world. It is impossible, in the sense that it is incoherent, to imagine that one of these sentences is false, given what the sentences mean. These sentences are sometimes said to be true in every logically possible situation, since logics also can have sentences that are always true simply because of what they mean. This notion of possibility differs from physical possibility.

 

Extensional Semantics

1. Predicates and arguments

Whether a sentence is true or not often seems to depend just on what the subject refers to and on what things have the property represented by the verb phrase. The verb phrase is sometimes called a predicate, and the subject denotes the argument of the predicate.

-          Cassandra laughed

It is natural to assume that sentences like these assert that the thing referred to by the subject Cassandra has the property named by the verb phrase laughed. Letting [Cassandra] be the person named by the name Cassandra, and letting [laughed] be the set of things that laughed, we can say:

-          The sentence Cassandra laughed is true if, and only if, [Cassandra] is in [laughed].

 

2. Extensions and intensions

The sets of things with the property named by a noun like witch or a verb phrase like sings or loves Juliet is sometimes called the extension of the expression. This approach to semantics, an approach based on reference, extensions and truth, is sometimes called extensional semantics. It is rather surprising that extensional semantics has been so illuminating, since it is easy to see that this sort of approach to semantics will not capture everything we might want to know about meaning. For example, there are noun phrases that refer to the same thing but which have different meanings. The phrases the author of Cymbeline and the author of Romeo and Juliet refer to same person, and yet they mean different things.

 

3. Determiners and nouns

The previous section suggests a first, very simple idea about the relation between meaning and syntactic structure, an idea about how references and extensions are related in simple subject-predicate sentences. The idea is that a subject refers to something, the VP denotes a property, and the sentence asserts that the reference of the subject DP has the property named by the VP. This idea is wrong. We can see that it does not work when we move to even just slightly more complicated sentences. Consider the sentences:

-          Every man is odd.

-          No man is odd.

-          Most men are odd.

-          Less than 5 men are odd.

In these sentences, the subjects do not name single objects. The different determiners in these sentences relate the set of men ti the set of odd things in different ways.

 

4. Modifiers: recursion and compositionality

In the syntax, the principles of phrase structure allow the construction of infinitely many sentences because they are recursive. In human languages, a phrase of a given category can have other phrases of the same category inside of it; the category can recur any number of times. This idea was introduces in chapter 3, where it was observed that a prepositional phrase inside a DP may contain another DP.

 

Assertion and Presuppositionin

In the preceding sections we observed that knowing what the entailment relations between sentences are constitutes a core part of linguistic semantic competence. In this section we distinguish between two species of entailment: assertion and presupposition.

 

Cross-Linguistic Comparisons Involving Word Meanings

In the preceding sections we argued that linguistic semantics is not the study of what sentences are true and what sentences are not: it studies entailment relations between sentences, and truth is but an auxiliary notion in characterizing entailment.

            We might ask a somewhat similiar question about the meanings of words. Are word meanings (lexical meanings) simple reflections of conceptual distinctions; distinctions that are directly determined by how the world is and how our perceptual and cognitive apparatus works? For instance, it is natural enough to wonder whether we can find that some relations are easily recognized and named by humans, while others are not so easily recognized and will tend not to be named by expressions of human languages. Returning to the consideration of prepositional phrases, an interesting case of this issue presents itself.

 

Summary

Competent speakers recognize certain relations among sentences, relations that hold by virtue of what the sentences mean. A true sentence can be said to assert the things that it entails, but the sentence can also presuppose some things, namely the things that are entailed by both the sentence and its negation. Entailment is an especially simple and important relation of this kind. We have assumed that meanings of sentences are determined compositionally, that is, on the basis of reasoning about the meanings of the parts of the sentences involved.

From the book "Linguistics An Introduction to Linguistic Theory" Written by Victoria A. Fromkin.