what speech do we like best?
toll booth operator

Verb Machine
Many irregular verb conjugations are based on fashion rather than logic

The Decline of Grammar
Geoffrey Nunberg makes an impassioned plea for civility among grammarians  when assessing language change

Additional Resources
Correct American Index

Correct American


Barring The Gates Of Language
“You say po-tay-toe, and I say po-tah-toe…” Our appearance, manner and the way we speak broadcast a social message. Language gatekeepers – often self-appointed – judge how we speak. John Fought explains how linguists try to help us be language-neutral. Read Summary

As we go about our lives, we move in and out of many social groups, each associated with specific places and characteristic activities — school, work, family, shopping, worship, soccer and more. For each of us, this dense network of associations formsour society. Every detail of our selves and our lives — our appearance, speech, activities and associations — is perceived by others as an emblem of who we were, what we have become and where we fit in. The social profile of each person is recognizable and unique, like a fingerprint.

The ways we speak also reflect this combination of social similarity and diversity in fine detail. Our speech reflects national, regional, occupational and educational differences (among others) via complex differences in accent, vocabulary and grammar. Our language is both a tool for communication and a part of our personal image, like our physical appearance, behavior and belongings.

Language helps signify who we are in society

Language has always helped to signify who we are in society, sometimes serving as a basis for exclusion. A Bible story tells how a password, shibboleth, was chosen because the enemy didn’t use the sh sound. “Shibboleth” has since come to signify an emblem of belief or membership, an identifiable sign of those who must stay outside the gate.

Societies use shibboleths in many ways. Indeed, speech is a convenient stand-in for other kinds of stigma that we recognize but do not openly acknowledge. For example, in our society, discrimination based on appearance, race, sex, religion or national origin is TABOO and often illegal, whereas discrimination based on particular details of language use by men or women, people of different religions, people from other countries and so on is often allowed. Thus, multiple negatives such as He ain’t got none are stigmatized as “illogical” when used in African-American Vernacular English (but not in standard French, Il n'en a pas, or Spanish, No tiene nada.). Double modals such as He might could are thought as an amusing "Redneck" trait, but similar formations such as I should do are upper crust British English. We can reject a person for the use of stigmatized language without feeling that we’re prejudiced against the person or their group. It’s just their language that’s unacceptable…right? We wouldn't want someone at our school/place of worship/workplace to sound like that, would we? It might leave a bad impression.

You Say Po-tay-to, I Say Po-tah-to

Differences in patterns of language use are normal, not evidence of social decline. Diversity is everywhere —in the most educated speech communities and in languages that have never been written down. In fact, we learn long before formal instruction that different speech situations call for different styles. For example, we have special family words and phrases that we don’t use with outsiders (Could you get that, poopsie?). Our jobs also call for certain ways of speaking; colleagues call me John or Dr. Fought, not dude or sweetie. Each form of address calls attention to the relationship between speaker and addressee. Mix them up and people will notice.

Your perception of each situation helps you decide what to say and how to say it

If you use a stigmatized way of speaking at work, that too will be noticed. However, what is stigmatized is different from person to person, from job to job, and from place to place. Using double modals “might could” make you an official "Redneck," but in the Mountain South, that’s a plus, not a problem. Saying running instead of runnin’ may not raise an eyebrow in some schools, but in others it makes you "Teacher’s Pet." It may even make you sound White. Your perception of each situation helps you decide what to say and how to say it.

A significant part of learning a job is learning its technical language. Ask any specialist. Often the biggest part of walking the walk is talking the talk. One person’s five-sixteenths offset wrench is another person’s twisty thing. What’s more, the mechanic asking for that wrench might not be a man. To me, your plum-colored raw silk mandarin-collared poet blouse with French cuffs is just a purple shirt — but to sell clothes, I’d better learn the distinctions.

Don’t Use No Bad Language

All social activities involving speech call for expert knowledge of when and how to talk. We know much more about this than we realize — much more than has been studied or described. Language gatekeepers have singled out some aspects of the many patterns of variation as grounds for action. Their choices — and the reasons given for them —  often don't stand up to inspection. For example, multiple negatives are said (wrongly) to violate logic.

However, in some groups and situations, they are expected — and everyone understands them perfectly. Natural languages never have been governed by logic. When Mick Jagger sings that he can’t get no satisfaction, he clearly doesn’t mean that he can get some.

A complex pattern of influences keeps the linguistic pot bubbling. Variation is everywhere. Change never stops. Language gatekeepers cannot control an ever-changing world of diversity. It’s hard on them, because in the gatekeepers' world, variation means error and change means decline.

What’s more, the very notion of a single standard of correctness in language is quite recent. “Correctness” is based solely on a purist’s own notion of what is socially or culturally correct: if it's not in, it must be out. A language purist works from a list of exceptions to the rule, ordinary speakers follow a hierarchy of patterns that reveal analogical similarities.

Just as no two people live the same life, so each of us speaks differently

Just as no two people live the same life, so each of us speaks differently. It starts with vocabulary and pronunciation. Some differences are regional, such as skillet vs. frying pan. Or Mary, merry, marry. Do they sound the same or different? What about cot and caught? Grammatical patterns are also diverse and constantly changing. Compare your irregular verbs with those of someone else, and you’ll find specific differences within general patterns. (Linguists who have traced the history of verbs in English and closely related Germanic languages have found evidence of such diversity and innovation dating back to the earliest surviving documents.) Modern speakers in England and in North America use so many competing patterns that the same verb stem shows different forms of past and perfect tense in different regions. Differences even show up within the same region. Given all this, it’s impossible to justify a purist’s list of ‘correct’ irregularities on linguistic grounds.

The great American linguist Leonard Bloomfield observed many years ago that the child who learns to say I seen it has learned just as much as the one who says I saw it. Both of these forms are irregular. The least common American past-tense form of see is the regularized form I seed it. These so-called mistakes shed more light on language than “standard” forms, because what’s going on is clearly not ignorance, laziness or poor schooling. The pattern of present, past and perfect of see, seen and seen in place of see, saw and seen reveals that speakers don’t put irregular verbs together just by combining a stem and a suffix, the way they form many thousands of English regular verbs. Among the roughly 180 ‘approved’ irregular verbs now listed in grammars of American English, there is no verb with an -en suffix in the past as well as the perfect form.

Where does I seen it come from?

So where does I seen it come from? It follows a more general pattern implicit in all the regular verbs and in many irregular ones as well. All of the regular verbs, such as need, needed,, and about 75 of the irregular ones, such as lead, led, led, have the same form in the past and present perfect, but a different form in the present. The see, seen, seen formation fits this more inclusive pattern, which can be stated as present differs from past and perfect; the past is like the perfect minus 'have.' [See English irregular verbs for more discussion and some experiments.]

To understand why gatekeepers feel they must judge these and other variants, think of overlapping social groups of all sizes as surrounded by boundaries, with members inside and everyone else outside. Many of these invisible boundaries subdivide large societies, and each boundary has two sides. Schoolchildren establish and maintain social boundaries by teasing and rejecting outsiders and favoring and accepting insiders. As we grow older, the groupings become more complex. Enforcement grows more subtle. Are these the right clothes? Is this a bad haircut? Did I say the wrong thing? You could get a STIGMA if you’re wrong…and every stigma has a cost.

Prescription Privileges

Gatekeepers want to keep insiders in and (perhaps even more important) outsiders out by opening and closing a real or imaginary gate. Many organizations have people or departments whose function is to let you (or lock you) in or out — ticket takers, prison guards, admissions officers, personnel managers and so on. Society also has freelance gatekeepers,who have decided —based on their own strong feelings — that some people or behaviors or beliefs or words are wrong and should be kept out. Gatekeeping is an exercise of authority, even if the authority is only imagined.

Gatekeepers want to keep insiders in and outsiders out

Linguistic gatekeeping is primarily exclusionary in purpose. It involves mostly a list of don't's.  Language gatekeepers typically do not engage in any systematic study of language structure. Instead, they issue judgments on specific items of usage that are structurally and functionally unrelated. Gatekeepers cite particular words or expressions as “wrong,” “incorrect,” “illogical,” “uneducated,” “awkward,” “substandard” or “ungrammatical” (out), and sometimes to alternative items as “right,” “correct,” “logical,” “educated,” “standard” or “grammatical” (in). These are terms of prescriptive rather than descriptive grammar.

Prescriptivists may also stigmatize a usage because they believe it is “lazy” or “foreign.” They frown on language associated with people or values they disapprove of (such as a person ending a sentence with a preposition). What's more, descriptivists, they often claim, advocate an “anything goes” policy. But that’s not so: descriptive linguists are devoted to finding out what actually does go on; that is, they discover how people actually use English or another target language. Their conclusions, however tentative, are based on the study of actual usage.

Linguists, through this painstaking work, see the many interlocking patterns of language form and social language use in a neutral way. They have the tools to take language apart into identifiable elements and to map their recurring relationships. They try to capture these intricate patterns in ways that help us understand how a particular language serves its communicative functions, as well as how usage differences match or interact with society’s other structural differences. Although many other social scientists look at the messages people send and receive, linguists also pay special attention to the forms that carry the messages.


African-American Vernacular English (AAVE): One of many terms for English spoken by African-Americans socially outside the middle or professional class, or who may temporarily wish to appear so for social reasons. Other terms include Black English, Non-Standard Negro English, and Ebonics.

double negatives: The use of more than one negative particle in an expression, such as “I ain’t got none” or “I can’t get no satisfaction” (compare Spanish No tengo ninguno. Against all experience, prescriptivists claim to believe these will be interpreted as if part of a formula of classical logic. Instead, they are interpreted by everyone as meant: they are negatives. If they weren’t understood as negatives, both the prescriptivists and Mr. Jagger would be satisfied. Anyhow, classical logic was modeled after the Greek language, not the other way around, and there are other types of logic, including many in which true or false is not the whole story, so that’s no help.

double modals: Used in some Southern speech varieties: “I might could” for “I might be able to,” or “I shouldn’t oughta,” etc. Stigmatized, but used anyway, even by College English speakers from the region. A similar formation is in limited use by the uppermost classes in Britain: “Are you going to visit your uncle?” “I should do.”

gatekeeping: controlling access to goods, services or information, usually applied by individuals or groups in hierarchical organizations.

stigma: A wound or scar; here, any social blemish, such as poverty, unfashionable clothes, bad grammar, or the wrong parents.

taboo: An action or word under strong social prohibition.

Suggested Reading/Additional Resources

  • Bauer, Laurie, and Peter Trudgill.  Language Myths.  London & New York: Penguin Books, 1998. [A collection of useful and easily understood articles, each dealing with a common myth about language structure, use and history. A good way to sample the linguists' perspective.]
  • Bloomfield, Leonard (1887-1949). Much of his work is very hard to approach, but a 1927 paper, “Literate and Illiterate Speech” (American Speech 2.432-39) is a clear, easily understandable illustration of usage differences among speakers of Menominee, an “unwritten” Algonkian language spoken in rural Wisconsin, that Bloomfield studied and described over a period of many years. The paper was reprinted in Hymes 1964.391-96.
  • Hymes, Dell, ed.  Language in Culture and Society (New York:) A still-useful anthology, 1964.
John G. Fought, now an independent scholar, was an Associate Professor of linguistics at the University of Pennsylvania and has also taught at the University of Innsbruck, Austria, and at Pomona College. He has written on modern Mayan , Chorti (Mayan) Texts I, University of (Pennsylvania Press, 1972) and on the history of American linguistics Leonard Bloomfield: Assessments of Leading Linguists, (Routledge, 1999), and, with Dell Hymes, American Structuralism, (Mouton, 1981). He has studied and taught American regional and social dialectology for many years.

Back to Top

Sponsored by:

National Endowment for the Humanities Hewlett Foundation Ford Foundation   Arthur Vining Davis Foundations Carnegie Corporation

National Endowment
for the Humanities

William and Flora Hewlett


Rosalind P.

Arthur Vining
Davis Foundations

Corporation of New York