3.1.2.1
The contents of pedagogic grammars

Very often, pedagogic and textbook grammars focus primarily on verbs and verb tenses (see sections 3.3.2 'Audiolingualism' and 3.4 on 'Grammar in the syllabus'). However, it is important that, as language teachers, we are aware that grammar covers a wide range of features, from morphology through to text structure (for an example of grammar at the level of text, see Activity 10), and also a wide range of word classes.

Word Classes
Standard definitions of word classes or parts of speech are as follows:

noun naming word
verb doing word
adjective describing word
adverb a word which describes a verb

These definitions are fine as far as they go. Possibly the most striking fact about such definitions is that they describe the semantic function of words - what they contribute to meaning.

It should be obvious that a given word does not always belong to the same word class, so that, while most people would instinctively say that 'brick' is a noun, this seems true only in sentence number one below:

  1. I'll build my house of brick and stone.
  2. I'd like to own a red brick house.
  3. They bricked in the window.
  4. I'll brick you if you're not careful.
  5. I'll build the house *brickly.

In number 2, 'brick' seems to be an adjective, but the fact that we can say 'very happy', but not 'very brick' suggests that this is not such a simple matter; for the moment, it might be better to say that 'brick' in 2 is a noun doing the job of an adjective.

In 3, 'brick' is obviously a verb, and in 4 also, at least in the grammars of speakers of certain dialects of English.

If 1-4 suggest that a word can belong to any word class, 5 suggests that this is not the case. In 5, 'brick' is behaving like an adverb, a fact unacceptable (I think) to all speakers of English.

Simple definitions of word classes thus seem problematic, and it may be easier to look instead at what words are doing in phrases, sentences, and utterances (and, by extension, in texts).

In attempting to do this, rather than look at definitions we will examine for four word classes both the form (whether words change their shape and how) and the distribution of words (where different words can occur). (Note that the examples we will look at will show only typical members of different word classes, and you should easily be able to think of many exceptions to the generalizations I will be making.)

a) Nouns
Words that can fill the 'sheriff's' spot in 6 to 8 are very likely to be nouns:

  1. I shot the sheriff.
  2. The sheriff shot me.
  3. I gave my gun to the sheriff.

Nouns typically inflect for number and genitive case in English.

b) Determiners
A second look at 6-8 above shows that sheriff is always preceded by 'the'. It should be obvious that 'the' can be substituted by a number of other words, giving us the additional sentences below:

6. I shot the sheriff.
9.   a  
10.   this  
11.   that  
12.   our  
13.   which sheriff?

These words are determiners. Although traditional grammars often ascribe distinct terms and roles to many of the determiners, these distinctions are not central to our discussion. In English, the main determiners are:

articles: the, a
demonstratives: this, that, these, those
quantifiers: some, any, each, every
possessives: my, your, his, her, etc
Wh-determiners: what, which, whose

Determiners do not inflect. They are a closed set: we cannot create new determiners with the same ease we constantly create new nouns, verbs, and adjectives.

c) Pronouns
If we look at 6 one more time, it should be obvious that it is possible to substitute the sheriff by other elements, giving us the following:

6. I shot the sheriff.
14. him.
15. her.

These words are pronouns (often defined as words which substitute nouns). Pronouns are one of the few parts of the English language where inflection is still widespread. Pronouns in subject position are often said to take nominative case (eg he, she), while pronouns in the object position take the accusative case (eg him, her). Pronouns inflect for number as well as accusative and genitive cases. Like determiners, they are a closed class that cannot easily be added to.

d) Verbs
Following the system we have used so far, you can probably predict that verbs are just about anything that can fit into the 'shot' slot in 6.

6. I shot the sheriff.
16. Killed
17.

He
frequently
Kills

Sheriffs.

Typical verbs inflect for 3rd person singular and tense:

Apart from looking at word classes, it is important to note the interaction between grammar and lexis. Many words both carry meaning and perform a grammatical function. Pronouns, prepositions, and determiners, are essentially grammatical words, but also carry meaning. Grammatical words are closed sets. Verbs, nouns, and adjectives, on the other hand, are lexical rather than grammatical, and are open sets.

There is another sense in which grammar and lexis are interactive. Sinclair (1991) has maintained that much language knowledge and use is based, not just on the manipulation of grammatical structure, which he calls the 'slot and filler' model, but also on the production of fixed and semi-fixed phrases. In other words, rather than slotting lexical items into structural frameworks in order to communicate, much of the language we produce consists of fixed phrases and collocations (collocation is the way words tend to occur with each other).

As an example, Sinclair notes that the phrase 'of course' is equivalent to one word. Similarly, the expression 'by the way' is also really only one lexical item. The grammar of this expression is 'frozen' and cannot be changed to *'by a way', for example.

Sinclair also notes how semi-preconstructed phrases show levels of predictability. The phrase 'set eyes on', for example, generally takes a pronoun subject, and is often accompanied by either 'never' or a temporal conjunction like 'the moment' or 'the first time'. Similarly, the verb 'happen' collocates with unpleasant things such as accidents.

Hunston (2000) extends this idea and shows how lexical items show a range of collocational behaviours not covered in traditional grammars. We can see how this works in the following (slightly edited) lines taken from the COBUILD Bank of English database. The expression 'no reason' is followed by a wh-clause, while 'chance', 'danger', and 'no doubt' are followed by that-clauses. In other words, lexical items are often patterned with specific grammatical structures. These aspects of language are rarely covered in traditional pedagogic grammars.

years. Yeah. so there's no reason why it shouldn't happen in South Africa as
of catching it then there's no reason why you shouldn't talk to people with these qualities then there's a chance that you won't be out of a job. You know
I mean I know there's a chance that it won't actually but there is a
so there there's a danger that people who use
off. Right. There's always the danger that you might not know which one motivation is there. There's no doubt that he is one of our most talented
that then there's there's no doubt we will beat them. Absolutely

We can see here that grammar is a complex phenomenon that is not restricted to lists of verb tenses. It takes places at all levels of language, from the morpheme (smallest item of meaning) up to the level of text discourse (see Figure 3.1), and it is important that we as language teachers keep this in mind.

Diagram showing the scope of grammar

Figure 3.1: the scope of grammar