Traditional grammar

Within traditional grammar, the syntax of a language is described in terms of a taxonomy (i.e. classificatory list) of the range of different types of syntactic structures found in the language. The central assumption underpinning syntactic analysis in traditional grammar is that phrases and sentences are built up of a series of constituents (i.e. syntactic units), each of which belongs to a specific grammatical category and serves a specific grammatical function. Given this assumption, the task of the linguist analyzing the syntactic structure of any given type of sentence is to identify each of the constituents in the sentence, and (for each constituent) to say what category it belongs to and what function it serves. For example, in relation to the syntax of a simple sentence like: it would traditionally be said that the sentence consists of two constituents (the word students and the word protested), that each of these constituents belongs to a specific grammatical category (students being a plural noun and protested a past-tense verb) and that each serves a specific grammatical function (students being the subject of the sentence, and protested being its predicate). The overall sentence Students protested has the categorial status of a clause which is finite in nature (by virtue of denoting an event taking place at a specific time), and has the semantic function of expressing a proposition which is declarative in force (in that it is used to make a statement rather than, for example, ask a question). Accordingly, a traditional grammar of English would tell us that the simplest type of finite declarative clause found in English is a sentence like (1) in which a nominal subject is followed by a verbal predicate. Let’s briefly look at some of the terminology used here.

In traditional grammar, words are assigned to grammatical categories (called parts of speech) on the basis of their semantic properties (i.e. meaning), morphological properties (i.e. the range of different forms they have), and syntactic properties (i.e. word-order properties relating to the positions they can occupy within sentences): a set of words which belong to the same category thus have a number of semantic, morphological and syntactic properties in common. For example, nouns are traditionally said to have the semantic property that they denote entities: so, bottle is a noun (since it denotes a type of object used to contain liquids), horse is a noun (since it denotes a type of animal), and John is a noun (since it denotes a specific person). Typical nouns (more specifically, count nouns) have the morphological property that they have two different forms: a singular form (like horse in one horse) used to denote a single entity, and a plural form (like horses in two horses) used to denote two or more entities. Nouns have the syntactic property that only (an appropriate kind of) noun can be used to end a four-word sentence such as They have no . . . In place of the dots here we could insert a singular noun like car or a plural noun like friends, but not other types of word (e.g. not see, or slowly or up, since these are not nouns).

In contrast to nouns, verbs are traditionally said to have the semantic property that they denote actions or events: so, eat, sing, pull and resign are all (action-denoting) verbs. From a syntactic point of view, verbs have the property that only an appropriate kind of verb (in its uninflected form) can be used to complete a three-word sentence such as They/It can ... So, words like stay, leave, hide, die, starve and cry are all verbs and hence can be used in place of the dots here (but words like apple, under, pink and if aren’t). From a morphological point of view, regular verbs like cry (in English) have the property that they have four distinct forms: e.g. alongside the dictionary citation form cry we find the present-tense form cries, the past-tense/perfect participle/passive participle form cried and the progressive participle form crying.

Let’s begin by looking at the following set of sentences:

Sentence (2a) comprises the noun John which serves the function of being the subject of the sentence (and denotes the person performing the act of smoking), and the verb smokes which serves the function of being the predicate of the sentence (and describes the act being performed). In (2a), the subject is the single noun John; but as the examples in (2b–d) show, the subject of a sentence can also be an (italicised) phrase like the president, or the president of Utopia or the former president of the island paradise of Utopia.

Now consider the following set of sentences:

Sentence (3a) comprises the subject John, the predicate smokes and the complement (or direct object) cigars. (The complement cigars describes the entity on which the act of smoking is being performed; as this example illustrates, subjects normally precede the verb with which they are associated in English, whereas complements typically follow the verb.) The complement in (3a) is the single noun cigars; but a complement can also be a phrase: in (3b), the complement of smokes is the phrase Cuban cigars; in (3c) the complement is the phrase Cuban cigars imported from Havana; and in (3d) the complement is the phrase a specific brand of Cuban cigars imported by a friend of his from Havana. A verb which has a noun or pronoun expression as its direct-object complement is traditionally said to be transitive.

From a semantic perspective, subjects and complements share in common the fact that they generally represent entities directly involved in the particular action or event described by the predicate: to use the relevant semantic terminology, we can say that subjects and complements are arguments of the predicate with which they are associated. Predicates may have one or more arguments, as we see from sentences such as (4) below, where each of the bracketed nouns is a different argument of the italicized predicate:

A predicate like resign in (4a) which has a single argument is said to function as a one-place predicate (in the relevant use); one like feel in (4b) which has two arguments is a two-place predicate; and one like send in (4c) which has three arguments is a three-place predicate.

In addition to predicates and arguments, sentences can also contain adjuncts, as we can illustrate in relation to (5) below:

In both sentences in (5), smokes functions as a two-place predicate whose two arguments are its subject the president and its complement a cigar. But what is the function of the phrase after dinner which also occurs in (5a)? Since after dinner isn’t one of the entities directly involved in the act of smoking (i.e. it isn’t consuming or being consumed), it isn’t an argument of the predicate smoke. On the contrary, after dinner simply serves to provide additional information about the time when the smoking activity takes place. In much the same way, the italicized expression in his office in (5b) provides additional information about the location of the smoking activity. An expression which serves to provide (optional) additional information about the time or place (or manner, or purpose etc.) of an activity or event is said to serve as an adjunct. So, after dinner and in his office in (5a,b) are both adjuncts.

So far, all the sentences we have looked at in (1)–(5) have been simple sentences which contain a single clause. However, alongside these we also find complex sentences which contain more than one clause, like (6) below:

(6) Mary knows John smokes

If we take the traditional definition of a clause as a predication structure (more precisely, a structure containing a predicate which has a subject, and which may or may not also contain one or more complements and adjuncts), it follows that since there are two predicates (knows and smokes) in (6), there are correspondingly two clauses – the smokes clause on the one hand, and the knows clause on the other. The smokes clause comprises the subject John and the predicate smokes; the knows clause comprises the subject Mary, the predicate knows and the complement John smokes. So, the complement of knows here is itself a clause – namely the clause John smokes. More precisely, the smokes clause is a complement clause (because it serves as the complement of knows), while the knows clause is the main clause (or principal clause or independent clause or root clause). The overall sentence (6) Mary knows John smokesis a complex sentence because it contains more than one clause. In much the same way, (7) below is also a complex sentence:

(7)       The press clearly think the president deliberately lied to Congress

Once again, it comprises two clauses – one containing the predicate think, the other containing the predicate lie. The main clause comprises the subject the press, the adjunct clearly, the predicate think and the complement clause the president deliberately lied to Congress. The complement clause in turn comprises the subject the president, the adjunct deliberately, the predicate lied, and the complement to Congress.

As was implicit in our earlier classification of (1) as a finite clause, traditional grammars draw a distinction between finite clauses(which describe events taking place at a particular time) and non-finite clauses (which describe hypothetical or projected future events). In this connection, consider the contrast between the italicized clauses below (all three of which function as the complement of remember):

In (8a), the clause what pills he is taking is finite by virtue of containing present tense is: likewise, the clause what pills he took in (8b) is finite by virtue of containing past-tense took. However, the clause what pills to take in (8c) is nonfinite by virtue of containing no tense specification – take here is an infinitive form which is not inflected for tense, as we see from the fact that it could not be replaced by the past-tense form took here (cf. ∗‘John couldn’t remember what pills to took’ – the star indicating ungrammaticality).

Whether or not a clause is finite in turn determines the kind of subject it can have, in that finite clauses can have a nominative pronoun like he as their subject, but non-finite clauses cannot (as we see from the ungrammaticality of ∗‘John couldn’t remember what pills he to take’). Accordingly, one way of telling whether a particular clause is finite or not is to see whether it can have a nominative pronoun (like I/we/he/she/they) as its subject. In this connection, consider whether the italicized clauses in (9a,b) below are finite or non-finite:

The fact that students in (9a) can be replaced by the nominative pronoun they (as in ‘I didn’t know they have problems with syntax’) suggests that the italicized clause in (9a) is finite – as does the fact that the present-tense verb have can be replaced by its past-tense counterpart had in (9a). Conversely, the fact that students in (9b) can be replaced by the accusative pronoun them (as in ‘I have never known them have problems with syntax’) suggests that the italicized clause in (9b) is non-finite – as does the fact that we can optionally use the infinitive particle to in (9b) (as in ‘I have never known students to have problems with syntax’), and the fact that we can replace the have expression by one containing the infinitive form be (as in ‘I have never known students be worried about syntax’).

In addition to being finite or non-finite, each clause within a sentence has a specific force. In this connection, consider the following simple (single-clause) sentences:

A sentence like (10a) is traditionally said to be declarative in force, in that it is used to make a statement. (10b) is interrogative in force in that it is used to ask a question. (10c) is imperative in force, by virtue of being used to issue an order or command. (10d) is exclamative in force, in that it is used to exclaim surprise or delight. In complex sentences, each clause has its own force, as we can see in relation to (11) below:

In (11a), the main (asked) clause is declarative, whereas the complement (gone) clause is interrogative; in (11b) the main (know) clause is interrogative, whereas the complement (retired) clause is declarative; and in (11c), the main (tell) clause is imperative, whereas the complement (had) clause is exclamative.

We can summarize this subject as follows. From the perspective of traditional grammar, the syntax of a language is described in terms of a taxonomy (i.e. a classificatory list) of the range of different phrase-, clause- and sentence-types found in the language. So, for example, a typical traditional grammar of (say) English will include syntax of negatives, interrogatives, exclamatives, imperatives and so on. The subject on interrogatives will note (e.g.) that in main-clause questions in English like ‘Is he winning?’ the present-tense auxiliary is inverts with (i.e. moves in front of) the subject he, but not in complement-clause questions like the if-clause in ‘I wonder if he is winning’, and will typically not be concerned with trying to explain why auxiliary inversion applies in main clauses but not complement clauses: this reflects the fact that the primary goal of traditional grammar is description rather than explanation.