On Form and Function


By form here I mean the various form that a given lexical item takes in terms of inflection. By function it is meant what function does the term have in syntax?

If we use the above definition for form, there will be four parts of speech:

nouns: they occur with the inflectional suffix '-s.'

adjectives and adverbs: they occur with inflectional suffix '-er' and '-est'; they may also occur in paraphrastic constructions with 'more' and 'most.'

verbs: they occur with the third person singular ending '-s', the past tense ending '-ed,' the progressive participle '-ing,' and the non-progressive participle '-en.'

pronouns: they can be distinguished for the nominative and accusative case forms.

The remaining ones take no inflectional affixes: prepositions, conjunctions, complementizers, quantifiers.

While all this sounds nice, there are the determiners this/these and that/those which are like nouns in that they take a form of '-s' to mark the plural agreement forms. No one calls these nouns. The function of a noun is to designate an object in the most abstract sense possible. Events and most modifiers are not objects. However, some nouns while referring to objects, function as a modifier as we saw in Chapter 1 (a cotton shirt). An object would have to include resultatives and acts such as destruction, creation, remembrance, decision, and so forth.

So we then switch to function to identify categories. For example, D is a function which stands for anything marked [+Def]. Determiners would then be different from nouns in function. But this is going too far. Pronouns are marked as [+Def], which the exception of 'one' which is [-Def]. Pronouns have been analyzed as D (for determiner). We would have to distinguish parts of speech by both form and function. This isn't such a bad idea, actually. The function of a pronoun is to be a non-referential form that must be linked (coindexed) to an antecedent. See the chapter and my notes on pronouns and coindexation. A pronoun then could have the features [+Def, +Obj, -R] = [+Definite, +Object, -Referential].

Well, then, what is 'A'? Carnie goes all the way and claims that it includes adjectives, adverbs, and degree words. But he isn't clear on the distinction of function and form. But it seems to have a function: to modify something. If A is a function terms, then it should include all modifiers, which he nor any other known linguist does. That would include PPs which function as modifiers, relative clauses which function as modifiers, and perhaps other modifiers that I haven't thought of. Since adjectives and adverbs have been distinguished for ages and ages, we would then have to go one step further. We could differentiate them by what they modify. Hence an adjective is a form that takes '-er' and '-est', and it modifies nouns and sometimes adjectives as in 'bright yellow shirt.' The term 'adverb' is a major headache. Traditionally it has been used to denote a modifier that can modify everything except the kitchen sink and nouns. This definition just goes too far. I would like to restrict it to modifiers that modify verbs and adjectives. Most of these are modifiers that end in '-ly' but not always: rapidly, softly, slowly, quickly, barely, sincerely and fast (*fastly), well, bright (the sun shines bright) and brightly (the sun shines brightly), bright (a bright yellow shirt) but (*a brightly yellow shirt) and (*the sun bright shines). Confused? No need to be. If an adverb does not end in '-ly', it cannot precede a verb and modify it. If you think this is hopeless, read on.

The category 'P' has no inflectional endings. Its function is varied:

to mark location - 'the book is on the table.'

to mark space - 'the book is in the desk.'

to mark time - 'John will arrive in a week.'

to mark some abstract relations - 'He is talking about his graduation.'

to mark direction - 'Mary crawled up the latter.'

and it occurs with verbs to form phrasal verbs - Kelly looked up the number.

So how do we label categories? Right now it just plainly looks hopeless. The best thing I can suggest is to follow recent tradition and contemporary linguistics. Some labels we will simply have to argue over and feel frustrated. I wish I had a brilliant answer, but, alas, I do not. Woe is the world of linguists.

Go to top of page

This page last updated 13 MY 2004