Parts of Speech

In a response to a comment on my Kilpatrick blog, I found myself waxing on about parts of speech. That is a sort of boring subject, I know, but I think I can make it at least somewhat interesting by saying something you have never heard before and contribute to people's understanding of the issues involved in determining parts of speech -- something James Kilpatrick is totally clueless about.

Typically people comment on the parts of speech of the words in sentences without in any way making clear what the foundations of their claims are. There was a time when Latin grammatical classifications guided those analyzing English. This was a monumentally bad idea because Latin has properties English does not have and conversely.

So, English must be described independently of assumptions about other specific languages except in one respect. Linguists often use the term "universal grammar" meaning by that that there should be certain principles of language that are shared (but not necessarily in identical ways or to the same degree) by all languages. What is a "verb" in one language will have to be consistent with what we say a "verb" is in every other language. This is where it gets interesting.

Consider, for instance, the adjective-verb distinction. Let might begin by citing an especially mindless definition of "adjective":

Adjectives are words that describe or modify another person or thing in the sentence. The Articles — a, an, and the — are adjectives.

So, we are to understand that adjectives modify human beings per se. They may attribute properties to human beings but they don't modify them. Plastic surgery modifies human beings. An adjective might modify a noun that refers to a human being but that is something different. This one is a bit better:

Words which are used to modify nouns or pronouns are usually referred to as adjectives

What remains is to say what we mean by "modify." That, boys and girls, is a toughie.

In English, as these definitions suggest, we seem to have a clear distinction between adjectives and verbs, at least on the surface. But Chinese and some other languages do not make a sharp distinction. "He bigs," for instance, could mean "He is big" or "He is growing" depending on the context. The utterances of Asian languages generally, speaking, depend on context much more than English seems to and have less morphology even than English.

Words which are used to modify nouns or pronouns are usually referred to as adjectives

What remains is to say what we mean by "modify." That, boys and girls, is a toughie.

In English, as these definitions suggest, we seem to have a clear distinction between adjectives and verbs, at least on the surface. But Chinese and some other languages do not make a sharp distinction. "He bigs," for instance, could mean "He is big" or "He is growing" depending on the context. The utterances of Asian languages generally, speaking, depend on context much more than English seems to and have less morphology even than English.

So, if we make a sharp adjective-verb distinction for English, what are we to do with Chinese? Once one starts thinking about that and gives English a closer look, we find that perhaps the adjective-verb distinction is not as great in English as we thought. There are two prongs to our approach to resolving this conundrum.

It is widely agreed in linguistics that the major categories, of which the most important are noun phrase and verb phrase, have heads that share similarities with the phrase as a whole. Nouns are, of course, the heads of noun phrases. In some cases nouns exhibit number, as in the distinction between "boy" and "boys." This grammatical marker might seem to determine agreement in the verb for verbs that show it, as in

(1) The boy is happy.

(2) The boys are happy.

But what are we to do with:

(3) The boy and the girl are happy.

Here it is the noun phrase as a whole, not the constituent nouns, that is plural. It gets more complicated as language always does with a case like this:

(4) The boys or the girl is happy.

Nevertheless, both nouns and noun phrases are involved in number. So, determining parts of speech is a downward (from phrase to head) and an upward (from head to phrase) looking enterprise.

What are we to do with the following:

(5) John likes Mary.

(6) John is happy.

Arguably, both "likes Mary" and "is happy" are verb phrases. Suppose we adopt as our criterion for what is the head of a verb phrase that it bear number. Given that criterion, "likes" and "is" are our verbs. Now, in other contexts, "is" (i. e., "be" in one form or another) is a "helping verb." Examples are

(7) John is being arrested.

(8) John was arrested.

Here we would seem to want to say that the "main" verb or "true" head is "arrest" and "is" and "being" and "was" are helping verbs. But wait a minute. "Happy" in (6) sits right where "arrested" sits in (8), immediately after the verb "be." Moreover, the important content of the verb phrase in (6) is contributed by "happy" not "is." What is more, in many languages, and in some versions of African-American English Vernacular to boot, the verb "be" would not appear in a case like (6) unless it is negative or plural (but different languages handle these cases differently).

Now consider the fact that there are compound sentences in which we create ellipses in verb phrases. Consider

(9) John is going to the store and Joe is too.

(10) John was being arrested and Joe was too.

(11) John is happy today and Joe is too.

If we are to have a simple characterization of what is elided in these cases, we might want to consider the possibility that not only "is going to the store" is a verb phrase with "is" as its head but that "going to the store" is a verb phrase with "going" as its head. Similarly, "was being arrested" is a verb phrase with "was" as head; "being arrested" is a verb phrase with "being" as head; and "arrested is a verb phrase (note it could be extended with "by a policeman") with "arrested" as the head. This creates a hierarchy of verb phrases, one included within another, going down to the smallest verb phrase. This allows us to say that what is elided in every case is a verb phrase. I would argue, similarly, that "is happy today" is a verb phrase with "is" as head and "happy today" is a verb phrase with "happy" as the head -- as the head what? As the head verb.

I don't really expect all of you to believe this. I'm not sure I am 100% down with it myself, but the discussion should convince you not to take seriously or, at least, not to uncritically buy into the sorts of superficial grammatical analyses given by nitwits like James Kilpatrick or of grammarians that resort to silly or obtuse definitions of the sort we looked at above for adjectives.