It’s Not a Litmus Test


Last week I praised the late Michael Dummett for making an attempt at defining “noun” syntactically instead of relying on murky semantic intuitions about naming. Below I discuss a very different book (as American as Dummett’s book is painfully British) that does the same thing in a very different way. I will summarize both, and briefly draw an analogy with chemistry.

Dummett defined noun as “principal word in a noun phrase,” and noun phrase as an “expression that can serve as a subject,” and subject as an “expression preceding the verb in a simple sentence,” and verb as a “word that can show a tense contrast.”

This means noise is a noun not because it names a person, place, or thing (it doesn’t), but because it’s the principal word in phrases like the noise of the sirens. And the noise of the sirens is a noun phrase because it can be the subject in a sentence like The noise of the sirens was deafening. (Notice the verb agreement. The verb takes the form was rather than were because noise, not sirens, is the syntactically principal word. Its being singular suffices to make the whole noun phrase singular.) And the subject is the part before the verb in the sentence.

A very different presentation can be found in The Structure of American English by W. Nelson Francis (Ronald Press Company, New York, 1958), 237–252. Francis defines nouns by reference to four main criteria, all thoroughly structural:

  1. They occur with preceding occurrences of functions words of a class he calls the noun-determiners. These are basically what I call the determinatives plus the genitive forms of the pronouns (my, your, etc.). So when noise occurs after the, that’s a sign that noise is a noun.
  2. They have singular and plural forms. So the plural form noises is another sign that noise is a noun.
  3. They often exhibit various derivational suffixes specific to nouns: breakage, sadness, mathematician, communism, etc. So in noisiness, the suffix -ness is a sign that noisiness is a noun (and indirectly also that noise is a noun, because -y/-i forms adjectives like noisy, and -ness forms nouns from adjectives).
  4. Nouns often occur just before the verb in short sentences. So in The noise was deafening, the position of noise is an indication that it is a noun.

(For space reasons I set aside Francis’s discussion of stress contrasts, e.g. between the noun súspect and the verb suspéct.)

Francis and Dummett put things in very different ways, though they would hardly have disagreed at all about the data (provided we turn a blind eye to Dummett’s hostility toward a few “Americanisms”). There’s more than one sensible way to lay out the criteria for classifying words as nouns.

And you will have noticed that neither treatment provides a foolproof operational test. There are exceptions left right and center. Both offer partial syntactic definitions of the relevant cluster of terms, but from what they say you cannot derive a copper-bottomed, necessary-and-sufficient definition of the form “w is a noun if and only if you can take w and form the expressions f1(w) and f2(w) … and fn(w).” There isn’t one.

There are helpful general truths that have some exceptions, and tests that are useful if you deploy them sensibly, and rule-of-thumb guidance that depends on your already having a sensitive understanding of various other terms; but there’s no simple sure-fire test.

Once upon a time, before there were smartphones or Xboxes, children used to spend a lot of time with chemistry sets. I spent many hours messing around with glass test tubes, interesting chemicals, a cheap bunsen burner, and a little book of paper slips called litmus paper. Litmus paper provides a simple test for pH value. I verified that adding vinegar to water made it turn litmus paper pink, while adding washing soda turned it blue.

(I also secretly investigated deflagration in various mixtures of C12H22O11 and NaClO3, but we’ll draw a veil over that, since mercifully no one was injured.)

Litmus testing, though, is fourth-grade chemistry. Grown-up chemistry is vastly more complex, and the tests for properties and classification of substances are much more delicate and intricate.

The people who expect English grammar to be a simple matter that you can cover exhaustively in a few pages of memorable hints and tips are trying to keep it at a 9-year-old level. That’s not feasible. English syntax is an spectacularly complex system. It takes a a full semester of undergraduate work to get even a superficial grasp of it. That’s just the way things are. You know what Einstein is reputed to have said.

Return to Top