Alex Haley is the author of "Roots," in which a character exclaims "Freedom am won!"AP

When I was a tot, my mother made me read Alex Haley’s Roots all the way through. Even though my linguist days were far ahead of me, I was struck by one sentence: At the end of the Revolutionary War, a slave exclaims, “Freedom am won!” That seemed an off rendition of black speech to me then, and I assumed that Haley had innocently concocted that am usage.

Yet Haley was hardly alone in putting into black mouths an overgeneralized usage of am beyond the first-person singular. The Alexander in Irving Berlin’s pop hit “Alexander’s Ragtime Band,” who’s supposed to be a black man, sings, “That’s just the bestest band what am!” As late as the 1950s, the cartoon character Buzzy the Crow, rendered as an African American in the vein of the crows in Disney’s Dumbo, was given to saying things like, “That cat am a tobacco-smokin’ fiend!”

Overgeneralized am sounds so gratingly unnatural to a modern ear that even experts on Black English have long assumed that this usage was created by white minstrels. After all, minstrel dialogue was dripping with these clumsy ams; one routine included the query “You’se a man with a heap of intellectualities, am you not?” Zora Neale Hurston, who did extensive ethnographic study of poor rural black people and painstakingly reproduced their speech in her literary work, memorably wrote: “If we are to believe the majority of writers of Negro dialect and the burnt-cork artists, Negro speech is a weird thing, full of ‘ams’ and ‘Ises.’ Fortunately we don’t have to believe them. We may go directly to the Negro and let him speak for himself … Nowhere can be found the Negro who asks ‘am it?’”

But a great deal of evidence suggests that black Americans actually did use am in this way. The minstrels vastly exaggerated it, even as they savagely distorted black speech overall. However, amid this injustice lurks an example of how stereotype can be based in fact. Historical am usage also demonstrates something counterintuitive, given the racial conflict so deeply embedded in this nation’s history: a great many of the roots of Black English reach back to the speech of rural white folk in the British Isles.

In the 1930s and 1940s, thousands of ex-slaves were interviewed under the auspices of the Works Projects Administration. One of the oddest things about their narrations is that they are replete with overgeneralized am, or what a linguist might call invariant am. “And people says now dat Aunt Harriet am de bes’ cook in Madisonville,” says one slave, while another says, “Him am kind to everybody,” and another, “The truth am, I can’t ’member like I used to.” Reminiscent of Haley’s “Freedom am won!” one slave actually describes Emancipation Day with, “Then surrender am ’nounced and massa tells us we’s free.”

These and numerous other examples led one scholar, Jeutonne Brewer, to describe invariant am as black speech in a Ph.D. dissertation back in the 1970s. However, scholars studying the WPA interviews since have largely settled on a consensus that the ams were artificially inserted into these transcriptions by people expecting to hear am when the subjects were using standard forms of be such as is and are. That’s not an unreasonable argument. For one, in actual recordings, as opposed to transcriptions, none of the ex-slaves use invariant am. Then, also, one written transcription exists in two versions, one with the ams and one with the ams “corrected” to standard forms.

The case cannot realistically rest there, however. It may have been just a matter of chance that a mere 15 people did not use invariant am amid a corpus of thousands of other interviews. Plus, the person who “corrected” that one transcription, perhaps hearing the ams as unfamiliar, may have hastily assumed they were a mistake—when they were not. Was the mistake this one person’s, or was it that of legions of other interviewers and transcribers all writing am when they were hearing is and are? The latter seems implausible, especially given that is and are don’t sound much like am.

Besides, evidence exists beyond the slave interviews suggesting that invariant am was a real historical phenomenon. At the height of the Harlem Renaissance, Claude McKay published a hit novel titled Home to Harlem, which lovingly depicts black migrants from the South adjusting to life in a northern metropolis. McKay’s unfiltered rendition of black speech often includes invariant am: “Oh, these here am different chippies, I tell you,” “How the brown-skin babies am humping it along!” and so on. If this usage was a mere minstrel distortion, would McKay have put it in the mouths of characters meant to be taken seriously as demonstrations of the full humanity of black people?

The question becomes more urgent in that McKay was hardly alone among black writers in showing black characters using invariant am. In Sutton Griggs’s Imperium in Imperio, a black mother describes her son: “His name am Belton Piedmont, arter his granddaddy.” Note the arter for after, an inheritance from rural Britain. If Griggs’s ear was keen enough to catch this unequivocal linguistic reality, then why would he suddenly slip in a minstrelism? Or why would the NAACP’s sober house organ, The Crisis, edited by W. E. B. Du Bois, publish a poem in 1912 that includes lines such as “When de earf it am a-quakin’ / An’ de dead dey am a-wakin’ / An’ de people am-a-shakin’ in de knees”?

Either am was a part of Black English, or sober-minded black writers couldn’t tell a stereotype from a fact. Adding to the former side of the ledger is that invariant am has been transcribed from the mouths not only of black Americans, but also of rural Brits, who in the plantation era taught African slaves much of their English. (I thank the linguists Juhani Klemola and Marianne Hundt for bringing this to my attention.) People from southwestern England, for example, as late as the 1950s said things such as “Because you’m a lot older than I.” And people in the West Midlands have been documented as saying, “We drink water when we am thirsty.” Such evidence suggests that invariant am must join the long list of things that sound “black” to the American ear but that actually emerged across the Atlantic, such as ain’t, the use of be in sentences like “We be showin’ it all the time,” and the use of it in sentences like “It’s somebody at the door.”

How, then, to explain opinions such as that of Hurston? One reading is that she did not mean to suggest that invariant am was pure invention, but objected to the distorted way that black people were often depicted in using it. People perhaps did not specifically use am in questions like the “am it?” she used as an example. Hurston also mentioned Ise. Ex-slaves indeed said “I’se,” but not things like “I’se isn’t” and “I’se don’t know ’zactly,” as the servant Hannibal does in the novel What Can She Do? by E. P. Roe. (Roe apparently thought of Ise as a pronoun itself rather than a contraction.) It may have been depictions like this that disturbed Hurston.

Invariant am, then, seems not to have been a fiction. Its documentation in the historical record is too rich and unequivocal for such a verdict to make sense, as peculiar as it sounds to the modern ear, and as ticklish as its association with minstrelsy can make us. Invariant am, so saliently different from standard usage, likely came to be as stigmatized in its day as aks for ask is today, and seems to have receded into extinction after the 1920s. But the evidence suggests that it was once a thriving part of the dialect’s grammar.

Invariant am is, in the grand scheme of things, a useful indication of Black English’s coherence and legitimacy as human speech. Before people spoke Modern English, they spoke Old and Middle English; before Modern Chinese, they spoke Old and Middle Chinese. Why would Black English remain static over hundreds of years, with only slang changing up from time to time? Rather, Black English has ever been in transformation, just as all living human languages have always been, to the point that speakers from very long ago would sound peculiar to us. Despite how odd I found Alex Haley’s “Freedom am won!” in 1977, Haley got the language of the era just right.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.