Towards a Contextual Grammar of English

The Clause and its Place in the Definition of Sentence

Eugene Winter

• Publisher: Routledge
• Year: 2020
• ISBN:
  9781000156126
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Grammar & Punctuation
• Index: Languages & Linguistics

This study, first published in 1982, attempts to show that the foundations of a contextual grammar of English must be firmly based on an adequate definition of the sentence. This book will be of interest to students of language and linguistics.

Part One

The Nature of the Clause

Section 1

Sentence and Clause in English

---

1.1 Sentence and Clause

The central problem of working towards a contextual grammar of English is our lack of an adequate description of the contextual role of clause in the notion of sentence, and especally of the contextual contrast between subordination and independence for the clause. This description is necessary as a preliminary to finding a definition of the sentence which will account for the real-world use of the clause in our daily communication, whether spoken or written. Familiar categories will have to be used in unfamiliar ways because they are being applied in a contextual description of clause and sentence. For instance, independent clause is seen as ‘sentence’ while question clause is not seen as ‘sentence’, but as a demand for ‘sentence.’

While many traditional and modern grammarians conflate the notions of clause and sentence, I follow Halliday (1961, pp. 253–4) and Pike and Pike (1977, p. 482), who distinguish them. In Halliday’s old rank analysis, a sentence consists of one or more clauses, and I wish to further specify that at least one of these clauses must be an independent declarative (including imperative and exclamatory) clause. In contrast, question clause is not regarded as ‘sentence’ but as a specially incomplete clause which demands propositional and grammatical completion as independent clause or part of independent clause. This follows from the complementary definition of sentence presented in the final part of this book. We ignore the single-clause sentence (simple sentence) where the sentence function is represented by a single independent clause such as (1) below, which can be seen as an answer to a question like: ‘What kind of claim was the claim of the councils?’

(1) The claim of the councils was a reasonable one. (Scheurweghs, 1959, p. 53)

The distinction between sentence and clause reflects the distinction between a whole and its parts. The simplest possible solution to the problem of conflating the terms ‘sentence’ and ‘clause’ is to examine how these two terms might be used in analysing those sentences where there is more than one clause.

What all clauses have, in common is that they have the well recognised syntactic constituents of subject and predicate with or without adjunct, or just predicate with or without adjunct. (We will take it for granted that the notion of Quirk et al. (1972, p. 42) of S V, S V O, S V C, S V O O, and SVOC requires supplementing for prepositional structuring, for example ‘He talked to her’, in which the verb talk has the fixed phrase structure of talk to, and ‘He questioned her about me’, in which the prepositional phrase about me is part of the structure of the verb questioned. Such considerations require the analysis of S V prepositional ‘object’ and S V O prepositional ‘object’). If we examine sentences, we find one or more clauses of different kinds of grammatical status, as in (2) below, where there are three:

(2) It is possible that the contrast between the classical drama of England and the classical drama of France, to which reference has already been made, can be accounted for by the differences of audience. (Scheurweghs, 1959, p. 275)

Before we analyse the three clauses, we note a refinement to the analysis of SVOCA by Quirk et al.: the structure of the main clause can be represented as Si V C Sii, where Si is the use of anticipatory It, V is the verb is, C is the adjective possible as complement, and Sii is the that-clause as real subject of the clause which ends the sentence. Looking at the three clauses, we find that they are all finite clauses; one independent clause is represented as (a), a subordinate declarative type clause is represented as (b), and a relative clause is represented as (c). Both (b) and (c) and contained within the clause grammar of (a), with (b) as the that-clause subject which in turn contains (c):

(a) It is possible (b)

(b) that the contrast between the classical drama of England and the classical drama of France, (c) can be accounted for by the differences of audience

(c) to which reference has already been made.

The difference in grammatical status between the three clauses can be seen in how they are grammatically signalled. First, according to the Meillet/Bloomfield/Jespersen definition of sentence to be discussed later in this section, only (a) can stand alone as sentence, provided that it has the that-clause of (b) as part of its clause structure which fulfils the anticipation of clause subject by the item It. Taking the clause (a) in some detail, the grammatical item It immediately followed by finite present tense verb is confirms that the item It is the grammatical subject of an independent clause whose predictable pattern is of the It is possible that-clause kind. As A. S. Hornby (1954) has shown, this pattern is one we can predict for the adjective possible, so that this adjective reinforces the prediction of a that-clause structure to follow it. Secondly, neither clause (b) nor clause (c) can stand alone as they are both signalled as subordinate clauses which are part of clause (a)’s structure. In clause (b), the subordinate that signals that there will be a declarative type clause as the real subject which fulfils its prediction by anticipatory It. Within the structure of clause (b) itself, we note that clause (c) interrupts the syntactic relation between its subject and its predication; more specifically, the subordinator (to) which signals that a non-defining relative clause will postpone the predication. Notice how this corresponds directly with the use of the punctuation by commas.

Taking Fries’s (1952, p. 56) point that a sentence is the synthesis of its structural and lexical meanings, there are three clauses here whose structural and lexical meanings are synthesised in the larger structure containing them all. It follows that, in discussing the grammar and semantics of parts of this sentence, we must take account of the particular clause each part is in and in turn relate this clause to the other two. For instance, the nominal group the contrast between the classical drama of England and the classical drama of France must be taken as S of the that-clause; the interrupting relative clause to which reference has already been made comes in between the nominal group subject and its predication can be accounted for, etc. Secondly, in analysing our clauses within their sentence, we must take seriously the function words when we work out our clause boundaries. It therefore makes sense to keep a clear distinction for ‘sentence’ as the finished whole and ‘clause’ for focusing on matters of grammar and semantics within the sentence.

However, in speaking of the grammar of the clause within its sentence, we have the paradox that the clause itself does not exist except as a generalisation that all clauses have subjects and predicates, with or without adjuncts; or simply predicates, with or without adjuncts. Taking a simplified form of clause (b) above as that the contrast can be accounted for by differences of audience, we note that this exhibits just one of the many kinds of grammatical status which its clause elements can take, that is, the subordinator that signals the enclosure of what might otherwise be taken as independent declarative clause: the contrast can be accounted for by the difference of audience. As independent clause, it would have a different context. Again, if we nominalised the elements of the clause as the differences of audience which can account for the contrast, we would have the same clause elements but yet another context meaning. Here the unique clause meaning is used to identify the nominal head the differences of audience. The conclusion to be drawn here is that, although we have the same clause elements throughout, we can only speak of the particular contextual meaning imparted to the clause by its grammatical status as clause.

We come closest to recognising the changes of grammatical status for the clause when a particular clause structure is repeated by substitute clause. Consider the change of clausal meaning for the clause elements they (the rugby tourists) retaliated in (3) below. Note the first appearance of the clause elements in the if-clause (if they had not retaliated) and the change of subordination to when-clause for the second appearance of these elements (when they did):

(3) It seems, then, that he must have been direly provoked, to lash out as he did. The reaction, unfortunately, was the opposite of what he had hoped: instead of civilising their methods, New Zealanders saw the attack as a challenge to their toughness….
As soon as it became clear that Neanderthal methods were the order of the day, the Lions showed themselves to be lively fighters. The trouble was that, once violence erupted, the tourists were bound to lose face; if they had not retaliated, they would have been branded as cissies; when they did, they were condemned as thugs. (Sunday Telegraph, 7 August 1966, p. 11)

The point of this extract is that the clause the tourists retaliated is presented as a fact which the paragraph is evaluating. It is presented as negatively hypothetical by the if-clause, and then re-affirmed by the when-clause as true (real). The contrast between the hypothetical action and the real action by the tourists is made explicit by the same clause elements according to the grammar of their subordinate clauses.

If we return to the text discussed on p. 8 (D), we can note that the relations we looked at there in terms of hypothetical and real and affirmation and denial can be seen as being realised by changes of grammatical status for the elements of the two clauses.

(4) Mr Baldwin promised to resign if the Cabinet refused his request. It did refuse, and he did not resign.

The signals of hypothetical in the first sentence, the verb promise with its concomitant non-finite verb and the if-clause, are replaced by the denial he did not resign. This is an independent declarative clause instead of the previous non-finite clause (Mr Baldwin) to resign; it answers the question: ‘What did he later actually do: did he resign?’ The answer is ‘No, he did not resign.’ The affirmation function of the clause can be seen in the change from the if-clause elements (if the Cabinet refused his request) to the independent clause It did refuse. Notice here that the clause is not the unmarked It refused but the marked It did refuse. (This is marked because the operator did is not grammatically required).

The point of the last example is to illustrate an important fact about the unique lexical elements of a particular clause. This is that, while we all accept the notion of subordination or downgrading (Hill, 1958, p. 357), we are not so familiar with the converse contextual process: the upgrading of the information of the clause. This is the change from the subordinate clause status of (Mr Baldwin) to resign to the independent clause status of he did not resign, and from the subordinate clause status of if the Cabinet refused his request to the independent clause status of It did refuse, where there is a deletion of the object his request. Both changes involve replacements of one grammatical status with another for the same clause. Only by noticing that we have the ‘same’ grammatical elements in contrast with each other in the clauses of succeeding sentences can we begin to account for the contextual semantics of this replacement. (See discussion of replacement in Winter, 1974, pp. 211–16.)

At this stage of our knowledge, it is only when an individual clause is significantly repeated within or without the sentence boundary that we, as linguists, realise the contextual meanings of independence (as in the upgrading discussed above) and subordination (downgrading). In studies of replacement relations between clauses, we study both upgrading and downgrading of the information of the clause within and without the sentence boundary. If we provisionally define the term ‘sentence’ as potentially the largest grammatical unit built around one or more independent declarative clauses which may or may not have one or more subordinate clauses, then we must use the term ‘clause’ for all the clauses within the sentence. However, as an abstraction the clause needs to be specified according to its grammatical status, for example question clause, independent clause and subordinate clause. These general terms are themselves further specified according to their contextual function, for example the distinction between wh-question clause and yes/no-question clause.

So far we have established a definition of sentence as consisting of one or more clauses, at least one of which is an independent declarative clause. The term ‘clause’ can be used to describe the minimal structure of the sentence, the simple sentence, but normal...