Phrase and Sentence Structure

What Is Syntax?

No matter how difficult it may be to define a term as basic as ‘word’ in a precise way, it is taken as assumed that a sentence in any language is a collection of units, most of which are words. Some of these units are verbs, some are nouns, some are adjectives and so on, and some represent larger entities. Languages also differ in certain respects. For example, some languages permit prepositions to stand as free words while others use bound affixes to represent English prepositional meanings; some languages make regular use of pronouns while others avoid them; some languages turn concepts into verbs that English would denote through the use of adjectives. However, no language creates sentences simply by stringing words together without some higher level of organization. This higher level of organization is called a syntax. While each language has a syntax that is unique to that language, there are common features that recur across languages, and the observation of these commonalities has led many linguists over the years to seek a ‘skeleton key’ that will reveal a universal structure common to natural (i.e., human) languages.

In the 1950s and 1960s, the linguist Joseph Greenberg surveyed large numbers of languages with the goal of finding what he called language universals, i.e., statistical generalizations that generate predictions about features of the next language we might study, once we learn a few facts about the language (Greenberg, 1966). From the 1960s onward, linguist Noam Chomsky took the idea of universality a step further in attempting to ground human language in cognition with the goal of specifying basic features of syntactic structure that are part of our biological inheritance. The goal has been to characterize a universal grammar (UG) that transcends, but is reflected in, the surface-level differences among world languages (Chomsky, 1957, 1968, 1986). In general, the ‘core’ of UG represents the common platform or matrix of principles upon which language learning takes place; this platform includes language features which are unlikely to be taught explicitly as part of a second-language classroom syllabus because they are already shared by the first language. One key principle is that sentences have organizational structures of a type that will be sketched in this chapter, using English as the raw data for analysis; the basic framework is held to be biologically given. However, some structural facts are also highly idiosyncratic to individual languages. To take one example, there seems no reason why the fourth sentence below, which seems organized in a way parallel to the other three, should be unacceptable to those who learned English from an early age:

• I don’t knowthe place whereshe went.

• I don’t knowthe time whenshe left.

• I don’t knowthe reason whyshe left.

• *I don’t knowthe way howto find her.

Yet speakers of most dialects of English do not produce the way how; if they are presented with the last sentence above, they will judge it to be ungrammatical (or at least odd) in a way that the other three sentences are not. In the absence of a better explanation, this ‘hole’ in the system can only be described as a quasi-random language-particular fact that warrants special attention by teachers of English as a second language. It does not seem predictable by any general principles.