TOPIC XI Sentence. Syntactic Structure. Models of Analysis

1. The model of parts of the sentence.

2. The distributional model. The model of immediate constituents.

(IC-model).

3. The transformational model (TM).

1. In order to state general rules of sentence construction it is necessary to refer to smaller units. The process of analysing sentences into their parts, or constituents, js known as parsing. '

The syntactic structure of the sentence can be analysed at two levels: pre-functional (constituents are words and word-groups) and functional (constituents are parts of the sentence).

Parts of the sentence are notional sentence constituents which are in certain syntactic relations to other constituents or to the sentence as a whole. Accordingly we distinguish between principal parts of the sentence, constituting the predication, or the basic structure of the sentence, and secondary parts of the sentence, extending, or expanding the basic structure.

Parts of the sentence are notional constituents as they name elements of events or situations denoted by the sentence: actions or states, different participants and circumstances. The formal properties of parts of the sentence are the type of syntactic relations and the morphological expression.

Principal parts of the sentence are interdependent. The subject is the structural centre of the sentence — the predicate agrees with the subject in person and number. The predicate is the semantic and communicative centre of the sentence.

Secondary parts of the sentence are modifiers of principal and other secondary parts: attributes are noun-adjuncts, objects and adverbial modifiers are primarily verb-adjuncts.

Structurally parts of the sentence may be of three types: simple, expressed by words and phrases; compound, consisting of the structural and notional part (compound verbal and nominal predicate, subject with the introductory it and there); complex, expressed by secondary predications (typical of secondary parts of the sentence).

As there is no direct correspondence between units of different levels of sentence structure and as grammatical phenomena have fuzzy boundaries which often overlap, there are difficulties in distinguishing between certain parts of the sentence:

(1) / want to leave (object or part of the predicate?)

(2) Features of her mother and fattier were blended in her face

(adverbial modifier or prepositional object?)

Besides the three "traditional" secondary parts, two more are singled out: the apposition and the objective predicative (Ttiey painted the door green). Objective predicative is co-referential with the object, subjective predicative is co-referential with the subject (The door is green). Both types are denoted by the term complement. This term may be also used to denote all verb-adjuncts.

So the model of parts of the sentence shows the basic relations of

notional sentence constituents. It does not, however, show the linear order of

, constituents. The order of constituents is shown by two models of analysis

worked out by the American school of srtuctural (descriptive) linguistics: the

distributional model and the model of immediate constituents (IC-model). These models analyse the sentence structure at the pre-functional level.

2. Methods of structural linguistics are based on the notions of position, co-occurrence aand substitution (substitutability).

Position, or environment is the immediate neighbourhood of the element.

Co-occurence means that words of one class permit or require the occurrence of words of another class.

The total set of environments of a certain element is its distribution. The tenn distribution denotes the occurrence of an element relative to other elements. Elements may be in:

1) non-contrastive distribution (the same position, no difference in meaning; variants of the same element): hoofs — hooves (see Topic III).

2) contrastive distribution (the same position, different meanings):

She is charming. She is charmed.

3) complementary distribution (mutual exclusiveness of pairs of forms in a certain environment; the same meaning, different positioas; variants of the same element): cows — oxen.

The distributional model, worked out by Ch.Fries ("The Structure of English"), shows the linear order of sentence constituents (see Topic XI). The syntactic structure of the sentence is presented as a sequence of positional classes of words:

The old man saw a black dog there.

1) D A1 N1 V D A2 N2 Adv

2) D 3a la 2 D 3b lb 4

Showing the linear order of classes of words (their forms may also be indicated), the model does not show the syntactic relations of sentence constituents. The sentence

/saw a man with a telescope

is ambiguous, but the ambiguity cannot be shown by the distributional model. This drawback is overcome by the IC-inodel.

A sentence is not a mere sequence, or string of words, but a structured string of words, grouped into phrases. So sentence constituents are words and word-groups. The basic principle for grouping words into phrases (endo- or exocentric) is cohesion, or the possibility to substitute one word for the whole group without destroying the sentence structure. Applying the substitution test, (or the dropping test, dropping optional elements) we define syntactic relations and can reduce word-groups to words and longer sentences to basic structures:

(1) NP—>N poor John —>John.

The phrase is endocentric, the adjunct poor is optional, the head-word John is obligatory.

(2)The old man saw a bisck dog there.

Word-groups are reduced to head-words and the sentence is reduced to the basic structure, directly built by two immediate constituents — NP and VP.

When we know the rules of reducing the sentence to the basic, elementary structure, it is not difficult to state the rules of extending/expanding elementary sentences:

S —NP + VP /NP—A + N /VP —V + D (Adv)

So the sentence is built by two immediate constituents (NP+VP), each of which may have constituents of its own. Constituents which cannot be further divided, are called ultimate (UC). The IC-model exists in two main versions: the analytical mode! and the derivation tree. The analytical model divides the sentence into IC-s and UC-s:

The derivation tree shows the syntactic dependence of sentence constituents:

The. old man saw a black dog there.

The sentence:

I saw a man with a telescope will have 2 IC-sltuctures:

(1) Prn V Т N Prp Т N

(2) Prn V Т N Prp N

So the IC-model shows both the syntactic relations and the linear order of elements.

3. Different sentence types are structurally and semantically related. So the syntactic structure of a given sentence may be described by making these relations explicit.

Sentences, in which all constituents are obligatory, are called basic structures, or elementary sentences, or kernel sentences. Linguists single out from 2 to 7 kernel sentences: 1) NV 2) NVN 3) NVPrepN 4) N is N 5) N is A 6) N is Adv. 7) N is PrepN.

The structure of all other sentences may be explained as a result of certain changes, or transformations of kernel structures. This analysis, showing derivational relations of sentences, is called transformational (TM). TM is based on IC-model and it goes further showing semantic and syntactic relations of different sentence types.

TM was first discussed by the outstanding American linguist N.Chomsky and it greatly influenced further development of linguistics, other models either developing TM or being reactions to TM. In the course of the development of the model the focus of attention shifted from syntax to semantics.

TM describes paradigmatic relations of basic and derived structures, or the relations of syntactic derivation. Kemel sentences, which serve as the base for deriving other structures, are called deep, or underlying structures, opposed to surface structures of derived sentence types, or transforms. So both the deep and the surface structure belong to the syntactic level of analysis.

Transformations may be subdivided into intranodel, or single-base (changing the kernel structure) and two-base (combining 2 structures).

Single-base transformations may be of two types: modifying the kernel structure and changing the kernel structure:

(1) She. is working hard — She is not working hard.

(2) She. is working hard —• Her working hard—Her hard work. Some basic types of intrainodel transformations: substitution, deletion (Have you seen him?—See.n him?); permutation or movement (He is he.re— Is he here?);

nominalization (He arrived-His arrival); two-base

transformations: embedding (I know that he has come), word-sharing (I saw him cross the street).

TM shows that sentences with different surface structures paraphrase, because they are derived from the same deep structure: He arrived—his arrival—for him to arrive -his arriving.

TM shows that some sentences are ambiguous, because they derive from distinct deep structures: Flyfing planes can be dangerous— (1) Planes are dangerous.

—(2) Flying is dangerous.

So TM is an effective method of deciding grammatical ambiguity.

A grammar which operates using TM is a transformational grammar (TG). In TG the IC-analysis is supplemented with rules for transforming one sentence into another. TG became an extremely infuental type of generative grammatical theory, also called generative grammar.

The status and classification of transformations varied a great deal in the 1960s and 1970s, and several models of generative grammar have been investigated, following N.Chomsky.

TOPIC XII