"Girl, 7, Seeking U.S. Flight Record, Dies in Crash," says a headline in the New York Times, April 12, 1996. The news story includes the following sentence:
In the preceding quotation, them refers to "a person" and he or she to "a licensed pilot." We speakers of English have never known which pronouns to choose when referring to nouns of undetermined gender. The sentence sounds correct, albeit illogical, to me. Perhaps the alternation of singular and plural pronouns actually makes the meaning clearer.
On the other hand, speakers of English used to know which pronoun to use with a plural antecedent. Unfortunately, some of us have forgotten. In the fall semester of 1996, a high school teacher taking a graduate linguistics class I was teaching wrote the following sentence in a term paper:
Former President Clinton used the pronoun his or her with the antecedent every American in his State of the Union message in 1996. That in itself should not draw our attention. The context, however, suggests that he was talking about fathers rather than about every American:
The use of they with singular antecedents was once common. The use of he or she as a compound pronoun has become increasingly frequent in recent years. There is reason to believe that this compound has also broadened its meaning and can now be used to refer to both singular and plural antecedents.
A total of 17 out of 52 students in three different classes of mine in December 1995 (a composition class for ESL students, an undergraduate linguistics course, and a graduate course in sociolinguistics) saw nothing wrong with sentences of the type: "All citizens have the right to express his or her opinion." The percentage was the same at all three levels, which is surprising. Two years later, in December 1997, 5 out 9 students in the undergraduate course and 10 out of 25 in the graduate course judged the sentence correct. The ESL class was not tested. Whether the students would have produced sentences of this sort is an independent question. There is a difference between saying a sentence is correct and using it in speech or writing.
My sample was small and my method was haphazard. Hypercorrections of this sort, however, have been found by others. Miriam Watkins Meyers reported the following construction: "The production, 'By the Sea,' is a zany new age comedy about a bizarre group of characters' mad quest for his or her sense of higher power" [emphasis in Meyers' article] (229). Michael Newman, referring to an earlier survey of pronoun use by W. Green, reports, "Eight percent made hypercorrections matching a singular pronoun to the plural antecedents" (14).
It may be that we are seeing the beginning of a linguistic change: he or she (which I will use as shorthand to include his or her and him or her as well) can be used as a plural pronoun.
The word they is always used with a plural verb and is usually plural in meaning. It is not plural, however, in sentences such as: Have they come to fix the leak? Some time before the Norman Conquest, in an Old English translation of the Bible, formally plural pronouns were used with formally singular antecedents. James D. Gordon cites the pronoun hyra 'their' (120) and explains in a note that he has translated it 'his' (122). Indeed, according to the OED, the earliest attestation of the word everybody occurred in the sentence "Everye bodye was in theyr lodgynges."
Furthermore, as I have argued elsewhere, they appears to be the underlying subject of agentless passives (201). Thus, When was the Verrazano Bridge opened? is identical in meaning with When did they open the Verrazano Bridge? (The answer is October 8, 1964.)
Traditional handbooks, until recently, advocated the use of he with indefinite antecedents. Feminists urged the use of he or she instead of he. The masculine singular pronouns, as it happened, had never really caught on. Instead, he or she replaced the word they, which had always referred both to singular antecedents and plural ones. For the first time in the history of the English language, singular they was threatened. Speakers of English, thinking they were avoiding the use of sex-marked he, began to replace they with he or she. Thus, it is not surprising, that, by analogy, plural they was replaced as well.
In the case of indefinite pronouns, words like everybody, which require a singular verb but are notionally plural, sex-marked he has always been a form that felt stilted and unnatural. Michael Newman came up with the following result:
Descriptive linguists have maintained that descriptivism and prescriptivism are mutually exclusive. Introductory textbooks in general linguistics, sociolinguistics, etc., often open with a discussion of prescriptive linguistics, not to be confused with real - descriptive - linguistics. I have selected an example by Ronald Wardhaugh.
This is a mistake; any descriptive analysis should note prescriptive tendencies - perhaps a linguistic universal - as part of its description. Indeed, Wardhaugh grudgingly recognizes as much:
Linguistic change is inherent in the fact that language is learned, that we are not born speaking any particular language. Furthermore, linguistic development is a prerequisite for human existence. We survive because it is natural for us to be unnatural - to invent and use tools, to develop specialized skills and consequently to divide labor, to do things that have not been done before and to communicate these innovations to our contemporaries and our posterity. Human language must be designed to produce sentences that have never been said before, which can happen not only because grammatical rules are recursive, but because languages have processes to augment the lexicon.
Nevertheless, the evolution of language, inevitable though it may be, is itself an obstacle to communication. The Tower of Babel story in Genesis (11:1-9) is evidence that people have always feared linguistic change. Societies everywhere have prescriptive rules to slow or prevent the development of barriers of unintelligibility caused by natural linguistic evolution. Any descriptive grammar is incomplete if it fails to note the prescriptive rules honored, if not always observed, by the speakers of the language.
Linguists study the rules of language: grammar, lexicon, phonology. They also study the dynamics of language: linguistic change and language in society. In either situation, one would expect an analysis of what the standard language actually is, and what is working to maintain or undermine aspects of this standard. For whatever reason, standard has long been a taboo. Handbooks on usage, for which there is a demand, have been written with little linguistic input. Some of the rules advocated by teachers and the texts they use are, in fact, not rules. The rules for they, he, and he or she are in fact more complicated and nuanced than what prescriptivists tell us. Newman cites an interesting example:
The sentence does not sound wrong, even though she would be the obvious pronoun to use in this case. Why doesn't it sound wrong? Linguists should have been exploring this issue decades ago. Perhaps it is the tradition of anti-prescriptivism that was at work.
When we were taught to use he with indefinite antecedents, the problem is not that we were being taught prescriptivism. What we were taught was nonsense. It doesn't follow that prescriptivism has to be nonsense.
An example of what prescriptivism ought to be is found in an article by Yuchi Todaka about the distinctions between the prepositions between and among. Todaka's research led to the following conclusion: "If the items in the NP objects are seen individually, between is used, if not among is used" (32). Other variables exist as well. Todaka lives in Japan and faces the very tangible problem of what to teach students who want to learn English. Here is a case where linguistic research uncovered the rule, the correct description of standard English, that explains how to use these prepositions. The rule about using among with three or more objects has never been true.
An analogous study by Thomas Nunnally explains that there is a difference between nominal-force gerunds and verbal-force gerunds. One may say, I was surprised at Ashley marrying Scarlett, but no one says *I was surprised at Ashley marrying of Scarlett (364-65). Once again, a scholar has explained the reality of a descriptive rule that had been masked by an incorrect prescriptive rule.
The issue of epicene or generic pronouns is more complicated because the linguistic and political issues have become intertwined. Activists are likely to be Whorfians, at the practical if not at the theoretical level. Practical Whorfians, whether seeking to undo gender bias or attempting to confront different issues, are activists. The need to be active can lead to contradictory tactics. A few years ago, if my memory is correct, people - activists and others - said gay men and women. Nowadays it is lesbians and gays. When gay became a noun, it became masculine. On the other hand, the noun actor has become gender neutral, at least for some speakers: Marilyn Monroe was a great actor.
Thus, he or she has a political advantage over they: it is new enough to draw attention to the problem of gender bias in language. Perhaps no one intended to eliminate epicene they. The attack against biased but marginal epicene he turned out to be an attack against unmarked they. Speakers simply replaced all occurrences of they with he or she. Hypercorrection takes place when overgeneralization occurs. Instead of replacing singular they, he or she simply replaced they in formal usage.
In speech, on the other hand, they survives when referring to indefinite pronouns.
An earlier version of this paper was read at the 41st Annual Conference of the International Linguistic Association, 14 April 1996. I am grateful to Franklin Horowitz and Dorothy Sedley for their bibliographical suggestions.
Clinton, William Jefferson. 1996. State of the Union Address. 23 January 1996. Printed in the New York Times 24 January 1996.
Gordon, James D. 1972. The English Language: An Historical Introduction. New York: Thomas Y. Crowell.
Green, W. 1977. "Singular Pronouns and Sexual Politics." College Composition and Communication 28: 150-53.
Jochnowitz, George. 1982. "Everybody Likes Pizza, Doesn't He or She?" American Speech 57: 198-203.
Meyers, Miriam Watkins. 1990. "Current Generic Pronoun Usage: An Empirical Study." American Speech 65: 228-37.
Newman, Michael. 1997. Epicene Pronouns: The Linguistics of a Prescriptive Problem. New York and London: Garland.
Nunnally, Thomas. 1991. "The Possessive with Gerunds: What the Handbooks Say, and What They Should Say." American Speech 66: 359-70.
Todaka, Yuchi. 1996. "Between and Among: A Data Based Analysis." Word 47: 13-40.
Verhovek, Sam Howe. 1996. "Girl, 7, Seeking U.S. Flight Record, Dies in Crash." New York Times 12 April 1996.
Wardhaugh, Ronald. 1995. Understanding Grammar: A Linguistic Approach. Oxford UK and Cambridge USA: Blackwell.