Linguists and students alike are taught that the verb agrees with the subject, but is this always true?

Evi Judge
Course: Semantics (LING 3540)
Advisor: Professor Zygmunt Frajzyngier
LURA 2021

Huddleston and Pullum (2005) write “the verb agrees with the subject” (p. 31), and for quantificational nouns “the form of the verb depends on the … NP [noun phrase] that is complement to the preposition of ” where “[t]he meaning of number is such that the embedded NP must be plural” (p. 89). Thus, the conclusion that Huddleston and Pullum draw is that the verb agreeing with quantificational NPs must agree with the plural complement of the quantificational noun.

Provided this claim about verb agreement, the following phenomenon becomes a puzzle:

a)     the number of X is

b)    the number of X are

c)     a number of X is

d)    a number of X are

Note, that the above X denotes any countable noun (e.g. people, books, cats, etc.). The aim of this article is to briefly explore the role of the copula and the articles used with the phrase “number of X”. In the phrases “the number of X” and “a number of X” both the singular and plural form of the copular verb ‘be’ seem to be used. The question then is whether or not one form of the copula is preferred over the other form. To study this “agreement preference”, the frequency of each phrase was recorded from the Corpus of Contemporary American English (COCA). Based on the frequency tokens of each phrase, in the following, I will propose a tentative hypothesis for copular agreement as an independent coding means. The following frequencies were recorded from COCA:

a)     the number of X is | 702 frequency tokens

b)    the number of X are | 54 frequency tokens

c)     a number of X is | 46 frequency tokens

d)    a number of X are | 624 frequency tokens

The evidence that (a) and (d) are the most frequently occurring forms contradicts the proposal that verbs agree with their subject. Furthermore, this contradiction of copular agreement raises another issue—namely, does the copula or the determiner code what the head and modifier of a quantificational NP is?

The preference for phrases containing ‘the’, as in “the number of X”, to take ‘is’ stems from the head being “the number”. This is due to the word ‘number’ being singular, indicating that the copula would also prefer to be singular (i.e. to take the form ‘is’). The modifier in the phrase “the number of X” is in fact “of X”. “of X” merely specifies what “the number” refers to. Without the modifier, one would be liable to ask, “The number of what?”.

This is the opposite from phrases containing the indefinite article ‘a’—as in, “a number of X”—where the head of such a phrase is “of X” and the modifier is “a number”. Since the object of the preposition will be plural, Huddleston and Pullum’s rule that the following verb will be plural applies. The impact of having the whole phrase "number of X" with either determiner 'the' or 'a' is essential for making sense of what is being said. If either part of the phrase "a/the number" or X-alone was stated, one may expect a wh-word question. For example, if one were to state, "Wow, the number has really shot up." The response may be, "What shot up? (Number of what?)" Likewise, if one were to say, "Testing sites have really expanded." Another person may ask, "How many testing sites have expanded?" These scenarios illustrate the dependency of the components in the phrase "number of X"—regardless of the determiner used, 'number' aims to describe something, that something being X. And X, while it can stand alone as a noun, needs the context of 'number' to describe a quantity.  

The above analysis contradicts Huddleston and Pullum’s proposal in two ways: 1) the complement of the NP dictates the (following) copula's form, and 2) the determiner codes the head and complement within an NP. I have addressed the issue posed in (1), as the preferred copular change witnessed between “the number of X is” and “a number of X are” disproves that X’s plurality will predict the copula's form. The proposition in statement (2) is more difficult to tease apart.

What then, in these phrases, is the role of the article? Examine the following:

a)     The number of testing sites is expanding.

b)    The number of testing sites are expanding.

c)     A number of testing sites is expanding.

d)    A number of testing sites are expanding.

If one were to omit the copula (‘is’ or ‘are’) in all of the sentences above, one would not be able to discern what is “expanding” in the sentence. The copula, not the determiner, provides this information. Sentences in the form of (a) and (d) are preferred because the copula is less ambiguous in those sentences; whereas, in sentences (b) and (c), I suggest that the meaning of the sentence (i.e. what is expanding) is less easy to decipher. Furthermore, I propose that the meaning in sentence (b) is more akin to sentence (d) due to the use of ‘are’, and it follows that sentence (c) is more akin to sentence (a) because of ‘is’.

Huddleston and Pullum (2002) state that the indefinite article, when preceding ‘number’, indicates number-transparency (p. 350); however, when the definite article is used before ‘number’, a non-transparent number (i.e. a precise number) is understood. Thus, one is able to substitute a precise number in for the phrase “the number of ”—and if performed on sentence (b), the sentence would become “Five testing sites are expanding” further support the idea that (b) codes the same meaning as sentence (d), only (d) has some indeterminate amount of testing sites expanding.

The conclusion: verb agreement is not solely driven by the preceding subject. The articles ‘a’ and ‘the’ do affect verb agreement. This analysis supports the claim that verb agreement is an independent coding means (Frajzyngier, 1997).

References

Davies, M. (2005). Retrieved November 15, 2020, from https://www.english-corpora.org/coca/
Frajzyngier, Zygmunt. 1997c. Pronouns and agreement: systems interaction in the coding of reference. Atomism and binding. ed by. Hans Benis, Pierre Pica, and Johan Rooryck. 1997. Dordrecht: Foris, 115-140.
Huddleston, R. D., & Pullum, G. K. (2005). A Student's Introduction to English Grammar. Cambridge: Cambridge University Press, 13-14, 37, 82-92.
Huddleston, R. D., & Pullum, G. K. (2002). Nouns and noun phrases. In The Cambridge Grammar of the English language (pp. 326-340). Cambridge: Cambridge University Press.