During the Great Depression, the Federal Writers' Project, a program of the Works Projects Administration (WPA), collected 2,300 interviews with formerly enslaved people throughout the U.S. South. The WPA was a New Deal program designed to put people to work, but the program gave historians one of our most valuable resources for learning about the experiences of enslaved people. Before the Civil War, the only enslaved people who could write down and publish their experiences were those who escaped, and after the Civil War, most freed people wanted to focus on the future. Without these "slave narratives," the only records of slavery that would exist would be those produced by white enslavers-- and they would tell a very different story than that we learn from the enslaved people themselves.

As valuable as the interviews are, they are not easy to use. Because of the way they were transcribed, they can be difficult to read. And as with any primary source, we have to understand why and how it was created -- but an interview has two creators, the interviewer and the interviewee. This means that we have to understand not only what is said but what was asked. With the WPA narratives, this can be difficult, because we don't have the questions. When discussing the relationship between interviewers and interviewees, the Library of Congress highlights how the issues of race and politicization confused the results of the WPA's project:

"The interviewers were writers, not professionals trained in the phonetic transcription of speech. And the instructions they received were not altogether clear. 'I recommend that truth to idiom be paramount, and exact truth to pronunciation secondary,' wrote the project's editor, John Lomax, in one letter to interviewers in sixteen states. Yet he also urged that 'words that definitely have a notably different pronunciation from the usual should be recorded as heard,' evidently assuming that 'the usual' was self-evident.

In fact, the situation was far more problematic than the instructions from project leaders recognized. All the informants were of course black, most interviewers were white, and by the 1930s, when the interviews took place, white representations of black speech already had an ugly history of entrenched stereotype dating back at least to the early nineteenth century. What most interviewers assumed to be 'the usual' patterns of their informants' speech was unavoidably influenced by preconceptions and stereotypes."

Historians face similar problems in working with any oral history interview, but with the WPA slave narratives, those problems are magnified. Most of the interviewers were amateurs with little experience, and they had little understanding of (or concern about) the possibilities of distortion. (Today, even students doing oral history interviews for a class usually have access to better guidelines for interviewing than the Writers Project was able to provide.)

Key questions

What questions did the interviewer ask?

The folklorist Alan Lomax developed a questionnaire for the interviews, but some interviewers followed it more clearly than others. When the interviews were transcribed, the questions were left out, so we have to infer the questions from the answers.

You can often infer, though, what prompted a particular story or answer. Most interviewers asked specifically whether the interviewee had been beaten or sold, and most asked whether they remembered Union troops coming through and about their experiences when they were freed.

How well did the interviewee remember his or her experiences?

By the time these interviews were conducted in 1937–38, more than seventy years had passed since emancipation. In addition, most of the interviewees were describing events that took place when they were children or teenagers. By now, they were at least eighty years old -- sometimes more than one hundred -- and they may not have remembered clearly the events they were describing. (Think about how accurately you remember what happened to you when you were five or six years old.) In addition, the interviews were conducted during the Great Depression, and most of the interviewees were living in great poverty. That fact, along with the possibility of nostalgia, may have made them look on the past with rose-colored glasses. Their childhoods may have seemed happier in retrospect than they were at the time.

Did the interviewee answer truthfully?

Some interviewees mistook the interviwer for a government representative who could get them financial assistance -- many ask about their "pension." As a result, they may have exaggerated their answers to please the interviewer. Some may also have exaggerated their answers simply out of pleasure at being the center of attention.

At the same time, some interviewees may have avoided topics they found unpleasant. Cornelia Andrews, for example, claimed never to have been beaten until her daughter, present at the interview, told her to show the interviewer her scars.

Finally, the interview may have been very different depending on the race of the interviewer. Nearly all of the interviewees were living in the South and were subject to Jim Crow laws that subjected them to discrimination and allowed them little economic or political power. As a result, some interviewees were probably more likely to tell a white interviewer what they thought he or she wanted to hear -- whether that was good or bad.

Was the interview written down and transcribed accurately?

Unfortunately, we don't have any way to know how accurately interviewers wrote down what interviewees said. We do know that they didn't write down every word that was spoken, and we also know that some writers and editors revised or censored the interviews. In some cases, editors added words that may have substantially changed the tone, impact, or meaning of the narratives. They may have done this to make the narratives more readable or interesting, or to remove information they didn't think was particularly valuable (but which historians today might like to have!). They may also have done it for political reasons -- either to highlight the abuses of slavery or to suggest that slavery was not all that bad.

The interviewers were given specific guidelines in how to transcribe what was said, and as a result, the transcriptions are often frankly racist. For example, an interviewer, following Writer's Project guidelines, might have typed "Dey said dat der wuz..." instead of "They said that there was..." In some cases the choices reflect the dialect of the speaker, but not always -- everyone pronounces the word was as wuz. The transcription guidelines may have made formerly enslaved people sound quite different from the way their interviewers had heard them.

How reliable is this narrative?

After considering all of the above questions, ask yourself how reliable the narrative is as a historical source. If the interviewee mentions only the best or worst experiences, what might he or she not have mentioned? What might have been edited out? How might the answers have been different if the interviewer had asked different questions, or asked the same questions differently? What questions would you like to ask the formerly enslaved people that weren't asked? And finally, what do you believe -- and disbelieve -- from this narrative, and why?