Liars, Damn Liars, and Sociologists

Friday 1st April, 2016

Erynn Masi de Casanova

In the sixth part of our special section on Sociology and Fiction, Erynn Masi de Casanova reflects on the fictional aspect inherent in social research and how we might negotiate it in a creative and helpful way.

We like to draw firm boundaries around things. A bold line usually separates fiction writing and sociological research reports. Rather than taking this line between fiction and sociology as a starting point, I want to see what happens when we trouble it.

I write some fiction. So do you and so does every other sociologist. Quantitative, qualitative, it doesn’t matter. We are humans studying other humans. Deception and self-deception are part of human interaction, including the interactions that produce the data we analyze.

Surveys, interviews, and (to an extent) observations focus on people’s accounts: of their life experience, attitudes, opinions, and knowledge. Any time we analyze and report data that consists of research participants’ accounts of their lives, we are dealing in the realm of fiction. And we know it. We can’t afford to take the details of a life history as unassailable facts, as we might ignore rather than make sense of the contradictions they contain. And it’s often the contradictions that reveal participants’ discomforts and uncertainties, the silences and nervous laughs letting us know we are on to something. Only inexperienced or careless researchers take participants’ accounts at face value one hundred percent of the time. 

Why are the accounts of participants in social research potentially fictitious? 

Reason one: they may be consciously lying to us. We all teach our students about social desirability bias, but then hesitate to apply this insight to our own data. Social desirability bias and prestige bias reside in the farcically leading questions in the what-not-to-do examples of research methods textbooks. Despite what we know about bias, we often present data in a way that assumes of course our participants were sharing their true and honest opinions, unaffected by what they thought people should say or what a researcher with a fancy degree might want to hear. We glide through the research process without this bias tainting our findings.

Reason two: research participants may be unconsciously lying to themselves about their lives, and then passing those stories on to us. (Writer Chuck Klosterman discusses this problem with documentary filmmaker Errol Morris in an essay called “Something Instead of Nothing,” which I assign to my research methods students, who giggle at the cuss words.) Some of the most potent truths about our own existence are things that we can’t accept and thus would never say to ourselves or others. Self-deception may be the unreachable bedrock of our participants’ accounts.

Reason three: research participants may not give us much to work with, and in conscientiously filling in the gaps in their accounts, we may end up writing fiction. This well-intentioned embellishment could mean adding meat to skeletal survey data through imaginative interpretation, or inferring the reasons for mismatching actions and words in ethnography. The pressures on social scientists to present neat, coherent narratives of our research and participants spur us on in this fiction-writing.

So we all write some fiction. Now what do we do with this realization? Do we throw up our hands, say that you can never really know anything about anything, that all social life is made up, is fiction? Then we all go become accountants. As any Enron executive could tell you, there’s no gray area or creative storytelling in the world of accounting.

But we already know that, in social settings, truth is provisional, contingent, constructed through interaction. This is what Erving Goffman taught us with his theory of frontstage and backstage behaviors. Our aspiration to scientific rigor doesn’t have to keep us from acknowledging the fictional elements of the social life we study and the texts we craft to report on it.    

Rather than throwing out the baby with the cloudy epistemological bathwater, we can accept the indeterminacy of people’s survey responses or oral histories. Embracing provisional truths, truth with an asterisk, can be a way out of paralysis. We can admit that our texts are fictional; as fictional as the memories we share with our therapists or the biographies we sketch in online dating profiles. As fictional as the accounts on which we base our work. Critical ethnographers, novelists, and neuroscientists agree that even observers of the same event may view it differently. This is not a problem we have to resolve. We can realize that, as surely as the words “Call me Ishmael” signal an unreliable narrator, so too does hitting the record button to begin a research interview. We can make peace with the limitations of social science and aim for a good, thoughtful, study rather than overgeneralizing. If we reject the idea of a single absolute truth, then we can begin to dwell—at first awkwardly and tentatively, but then perhaps more comfortably—in the ambiguity that fiction captures so well. 

Here's more you may be interested in: