Last summer, I pointed out a peculiar reference in James March's The Ambiguities of Experience (2010: 63). He notes that "stories interpreting observed experience are among the earlier and more respected contributions to research in organizations" and then offers this puzzling sentence: "Indeed, some of the more artful practitioners of the craft resist making their own explicit interpretations, leaving meanings to be elaborated by others (Chekhov 1979; Krieger 1979)." While Krieger 1979 is indeed a piece of social research (Hip Capitalism, the story of a real San Francisco radio station), Chekhov 1979 is a collection of short stories by Anton Chekhov. It is a work of fiction, not a "respected contribution to research in organizations".
The other day, I noticed a similarly strange reference in Thráinn Eggertsson's contribution to Alston, Eggertsson and North's Empirical Studies in Institutional Change (1996: 15). "As for the historical evidence," he says, "it indicates that slaves assigned to semiskilled or skilled work were frequently granted better living conditions than slaves engaged in unskilled physical labor, such as mining, stonecutting, and agricultural work with low care intensity (Solzhenitsyn 1968; Fenoaltea 1984)." Here, again, while Fenoaltea 1984 is indeed an article in the Journal of Economic History, Solzhenitsyn 1968 is a novel, The First Circle, by the famous Russian dissident Aleksandr Solzhenitsyn. It is not "historical evidence".
The problem with these references is not just that they provide inappropriate support for their claims. It is also that they confuse two very different kinds of sources, treating them as though they were the same kind of text. After all, it is not impossible to cite novels and short stories in social science writing. Indeed, it is quite easy to rewrite March's and Eggertsson's sentences to avoid the mistake they have made:
Indeed, like Chekhov (1979) in his short stories, some of the more artful practitioners of the craft resist making their own explicit interpretations, leaving meanings to be elaborated by others (e.g., Krieger 1979).
As for the historical evidence, it indicates that slaves assigned to semiskilled or skilled work were frequently granted better living conditions than slaves engaged in unskilled physical labor, such as mining, stonecutting, and agricultural work with low care intensity (Fenoaltea 1984). Aleksandr Solzhenitsyn (1968) has vividly illustrated this difference in his novel The First Circle[, which was based on his personal experience of a prison that also served as a research lab]*.
It would be interesting to know whether this isn't more or less what an earlier draft of these chapters actually said. Somewhere in the editing process, perhaps, it was deemed more economical to assume that the reader would know who Chekhov and Solzhenitsyn are and why they are being cited in this context. I don't think this economy is justified here.
The problem, like I say, is that the reference implies that Chekhov has made a "respected contribution to research in organizations" and that Solzhenitsyn's novels present "historical evidence". Sometimes this confusion is achieved simply by citing two very different kinds of text in the same reference. For example, March (2010: 22) cites Gladwell's The Tipping Point on "threshold effects"; in the same breath (the same parenthesis), he cites an article in the Journal of Mathematical Sociology, thereby implicitly putting a work of popular non-fiction on the same level as a paper in a scientific journal. It is a nice irony that March's book is about "ambiguity" and that Eggertsson has been writing about contracting as "essentially a theoretical fiction". But I don't think irony is enough here.
*As Thrainn points out in the comments, there is an argument for using this novel as a kind of "historical evidence". I have here added his argument in square brackets. This argument has to be made explicitly, in my opinion, and would need to cite also the biographers or literary critics who have vouched for the historical accuracy of Solzhenitsyn's account. And they would presumably have based their judgments on less fictional documentation of prison life in the Soviet Union as well, which we might then much better cite as historical evidence instead. If it were up to me, we would always only cite historical novels as illustrations, never as evidence.