Syntax

I

Introduction

Syntax, branch of linguistics concerned with the structure of sentences in human languages. Syntax deals with the arrangement of words into phrases and phrases into sentences. Syntax contrasts with the study of morphology in linguistics, the branch of grammar that deals with the internal structure of words.

II

The Study of Syntax

Every native speaker of English recognizes sentence (1) as a well-formed sentence of English, in other words, (1) is something that is part of their language whereas the sentence in (2) will be characterized as no more than a word salad. (Note: in syntax, an asterisk is used to denote an ungrammatical utterance.)

Most people will attribute the strangeness of (2) to the fact that words are not in the right place. Syntax studies exactly that, i.e., what is the correct place for words to be in an utterance in order to form grammatical sentences of any given language. The data that syntacticians use to determine the right set of rules describing the general syntactic organization of a language consist of grammaticality judgements provided by native speakers just like the judgements referred to with respect to the contrast between (1) and (2). Of course, this is sometimes impossible as happens in the case of historical syntax, which is the study of the syntactic organization of earlier stages of a language. In such cases, where native speakers are not available, syntacticians resort to surviving texts as a more or less accurate reflection of the speakers’ knowledge at that time. Syntacticians attempt to construct a model of that knowledge. The goal of this model is to tell us what are the grammatical sentences of a language.

As such, syntax is thus independent from both phonology and semantics. Saying that syntax is independent from phonology and semantics does not mean that it is not closely related to those two linguistic components. Indeed, all theories of syntax provide a transparent and systematic way of mapping syntactic structures to phonological and semantic representations. The independence of syntax is, rather, the fact that we can judge a sentence whose meaning is gibberish as a syntactically perfectly well formed sentence. This is shown by the following famous example (after Noam Chomsky):

Although the meaning is rather impenetrable structurally (and that is what is of interest to syntacticians) it has exactly the same structure as (4):

III

Syntactic Organization

A sentence is much more than just a string of words. It has syntactic structure. If we consider the sentence in (1) we can first analyse its components in terms of lexical categories (what traditional grammars call parts of speech) and assign to each word a category:

Lexical categories, in turn, combine to form phrasal categories or just phrases. This combination can be represented in the form of a tree diagram as follows:

In this diagram the symbol S stands for Sentence. NP and VP stand for Noun Phrase and Verb Phrase respectively. Notice that phrases are projections of lexical categories. The lexical element that projects a phrase is called the head of the phrase. If we take a careful look at (5) we see that, for example, Det and N are constituents of NP. If we look at the VP more carefully we will also notice that clearly the relation between V and VP is rather different from that between N and VP. We say that V is an immediate constituent of VP whereas N is not (N is an immediate constituent of NP).

A

Phrase Structure and Phrase Structure Rules

One of the earliest ways to express the syntactic structure of phrases, clauses, and sentences is phrase structure rules, which take the form of so-called “rewriting rules”. A rewriting rule has the following form:

Essentially, this is an instruction to replace (rewrite) the symbol S by the symbols NP and VP. This operation gives rise to the following binary branching structure:

NP and VP are the immediate constituents of S. Now the structure of VP itself will be given by another rewriting rule such as:

Graphically:

where V and NP are the immediate constituents of VP. We can proceed further in the same way and apply the following rules:

(11)
a. V ? walked
b. NP ? Det N
c. Det ? the
d. N ? dog
e. N ? boy

Now by application of these rules we generate the structure in (5). In this system, we can also express how language makes an “infinite use of finite means” (Wilhelm von Humboldt). Consider a phrase like (12):

The rules required to generate its structure are:

In the first of the above rules the symbol NP appears on both sides of the arrow. This is called a recursive rule. By applying such a rule repeatedly we can generate an infinite number of different phrases without having a need for separate rules each time.

Now, if we consider more carefully the rules that we mentioned above we will notice that there are two kinds, those whose right-hand side is a sequence of categories (each of which can be the input to another rule) and those whose right-hand side is a word. The latter give us the lowest level of syntactic organization, in other words, the terminal symbols or leaves of the tree. In the examples that we used so far the terminal symbols were words. This is not necessary; we can go even further to the level of morphemes. Morphemes are the smallest meaningful units of a language. An example of a morpheme in English is the plural -s or the past tense -ed morpheme, and thus we can think of a rule set like this:

Applying these rules will give us the structure in (15):