We all owe thanks to our new Lounge Colleague Steve Lubet for his exceptionally interesting and thoughtful string of comments here and elsewhere on Alice Goffman’s On the Run, an “immersive” sociological study of an impoverished Philadelphia neighborhood. (See here (Lubet’s original book review in The New Republic), here (Goffman’s reply), here and here (Lubet’s response to Goffman’s reply), here (UCLA sociologist’s Jack Katz’s comments), and here (Lubet’s rejoinder to Katz).) I’d like to offer another perspective on the Goffman book that may provide a useful complement to those expressed so far.
This additional perspective arises out of what I agree with Lubet is the striking and disturbing inconsistency between Goffman's original presentation in her book of her actions and motivations on the night she drove around with her armed friend Mike looking for their mutual friend Chuck’s killer in the hope of exacting revenge, and her later restatement of the same events after Lubet had pointed out that, as Goffman had originally portrayed her own conduct and intentions, she had not only committed a felony, but had endangered human lives. (Neither of these practices, it is fair to say, is or ought to be a common feature of sociological fieldwork, “immersive” or not. That may not be a very limiting constraint, but even if it were, I would hope that it could command a broad consensus.)
Specifically, Goffman relates in her book that, Mike having earlier affirmed that “somebody gon’ die regardless,” she volunteered to drive Mike around one night in search of their friend’s killer “because, like Mike . . . , I wanted Chuck’s killer to die.” Goffman says she waited in the car with the motor running, “ready to speed off as soon as Mike ran back and got inside,” as Mike got out with his gun to go after someone he thought might be the killer (it turned out he wasn’t). After Lubet raised serious and legitimate questions about what the hell an assistant professor of sociology was doing engaging in conduct that appeared to comprise a conspiracy to commit attempted murder, Goffman posted an explanatory response on the Internet asserting that “I had good reason to believe that this night would not end in violence or injury”; and that “Talk of retribution was just that: Talk.” As Lubet explains in more detail, you can shuck and jive all you want, but the “before” and “after” versions of Goffman's story can't realistically be reconciled. The discrepancy raises important and serious questions about which narrative more accurately presents Goffman's actual actions and intentions. As Lubet points out in one of his posts, we don't know, and that in turn raises troubling issues about the reliability of Goffman’s entire book.
That prompted me to think about the uncomfortable (and unwelcome) erosion of the distinction between memoir and fiction, which Lubet’s commentary began to persuade me might be traced to, or at least compared with, developments in social science research implicated here. It is my superficial understanding that, as originally conceived, sociological and anthropological ethnography was an observational science in which the investigator as dispassionately and “objectively” as possible described the behavior of a culture's members, and through observation and interview inferred their beliefs and motivations. (Readers should supplement or correct this understanding in the Comments if it’s incomplete or wrong; I freely disclaim any depth of expertise.) “Immersive” study allowed the investigator to step into and become to some degree a part of a culture or its practices: Instead of depicting Yaqui Indians engaging in ritual ingestion of hallucinogens, describing their gestures, words and post-hoc recounting of their experiences, Carlos Castaneda actually takes peyote with his brujo, Don Juan, and tells us what he saw and felt. (Castaneda’s first three books were part of his undergraduate and graduate studies in anthropology at UCLA.) More generally, “immersive” observation allows the investigator to observe her own feelings and reactions as well as describe those of the subject population, and hopefully thus gain additional understanding and perspective on the culture being investigated that is meaningful, and meaningfully different from dispassionate “objective” observation.
But whether you're watching from the outside or the inside, there will inevitably be questions about the integrity or verisimilitude of your portrayal. Margaret Mead's seminal and deeply influential 1928 study of Samoan culture, for example, has become a battleground over whether she got many important features right, and if not, why not. I don’t know that anyone has accused Dr. Mead of fictionalizing or embellishing, but you don’t have to. The problem is that any observed portrait of a culture necessarily simplifies it to make it accessible as an object for study and understanding. And it necessarily does so by emphasizing certain details and omitting or downplaying others; finding patterns in some concatenations of circumstance but not in others; and so on. Ironically, observation becomes characterized not by its accuracies, but by its inevitable inaccuracies—what the observer necessarily simplified, or emphasized, or left out to render us an observation that is more meaningful, organized, and accessible than the impossibly messy and inaccessible mass of countless human details the observer was watching and trying to describe.
And once you concede that your observations are necessarily inaccurate and incomplete, it is a much shorter and seemingly easier step to consciously and deliberately tailoring your observations to your conceptions in the service of the deeper "truth" that you have "found" embedded out there. Carlos Castaneda admitted to narrative license in his work “to heighten some dramatic sequences,” while at the same time insisting that his stories were absolutely true to life. Challenged about discrepancies in a Time magazine interview, he said, “To ask me to verify my life by giving you my statistics . . . is like using science to validate sorcery. It robs the world of its magic and makes milestones out of us all.” Boy, is that ever a long strange trip from Emile Durkheim’s imperative of observational objectivity less than a hundred years before. (I use Castaneda as an example here because he was so prominent in popular culture—his books were widely read in nonacademic circles, and he was featured on the cover of Time magazine in 1973. By publishing her study as a general-interest trade book, Goffman apparently aspires to similar popular notice and influence.)
Ethnographic observation, especially the “immersive” kind, is really a memoir of sorts. And it would be fair to say that the slide from motivated perception to self-deception to outright misrepresentation in some corners of ethnography saw parallel development in the literary memoir and, a bit later, even journalism. The 1990s and early 2000s brought us James Frey and his Oprah-endorsed memoir of recovery, A Million Little Pieces, significant parts of which turned out to be made up; Stephen Glass, who repeatedly fabricated sources and incidents in The New Republic and elsewhere to substantiate his perspectives on modern culture, and then manufactured fake reporter's notes and otherwise confabulated to try and cover it up, and Jayson Blair, who plagiarized and invented sources for his reporting in the New York Times. All of these people, and others like them, were using contrived "facts” to paint a picture they believed was ultimately accurate in some real and important sense, just as Castaneda’s narrative license created stories he felt were true to life (if not literally true) and consistent with the purposes of his ethnographic studies. This entire concept is brilliantly illustrated in Stephen Colbert’s notion of “truthiness”—made-up “facts” presented as reality because they demonstrate something important the speaker earnestly and sincerely believes to be true.
Given the discrepancies between Goffman’s “before” and “after” stories, it may be that she is just reacting to the shock of recognizing the moral and mortal danger in conduct that felt perfectly natural to her at the time and accurately mirrored the thoughts and feelings of the study subjects who had become her friends. But it may be that Goffman fits uncomfortably into the unfortunate if narrow tradition of ethnographer as fabulist. Was her stint at the wheel that night a focused and premeditated mission of revenge or, as she now insists, just participation in a ritual of grieving that threatened no one? How much of each version of the story, factually and emotionally, actually happened? Can she even tell anymore? The only thing that seems clear on the current state of the record is that both stories—asserted with equal certitude by the same observer—cannot comfortably coexist in the same reality. Something’s gotta give.
If you think about it, good fiction is just a skillful narration of invented facts resembling some of the ones we live with that, one way or another, illustrates truths about our culture or condition. Really good fiction may illustrate such truths more evocatively or indelibly than the accidental vagaries of reality. But that doesn’t mean that there’s no difference between reality and fiction. And it certainly doesn’t mean that you’re right to claim something fictional really happened in the service of a deeper "truth" you earnestly believe.