The Rashomon Posture January 24, 2009Posted by Will Thomas in Methods.
Recently, I have been advocating a deeper consideration of the importance of maintaining “chronological problematics” in history—in short, the idea that important historiographical topics are revealed and consolidated by arranging them in a coherent and sequential order. I have taken chronological problematics to be confronted by a predominance of “epistemological problematics” wherein what is necessary is to theorize categories of epistemological practices (observing, communicating, representing, etc.) and then researching and presenting different historical approaches to these practices, creating a “gallery of practices”.
Today, I would like to consider the idea that the relationship between the gallery of practices and any sort of problematics is purely implicit. Being implicit, there is no sense of historiographical responsibility to resolving problematics. In place of this commitment is what I like to call the “Rashomon posture” (Christopher prefers “Ghost Dog”).
In Rashomon (see the Kurusawa film trailer here), conflicting, self-interested perspectives in a murder trial present no possibility of reconciliation, leading those who hear of the trial into despair about the possibility of truth and the human capacity for goodness. For examples of the Rashomon posture on history of science blogs, or at least a nod toward it, see Michael Gordin’s Hump-Day History post on Dmitrii Mendeleev; or Michael Robinson’s post “The Smudgy Window of History”, especially the last paragraph; or History of Economics Playground where we can replace Rashomon with Lawrence Durrell.
The idea behind the Rashomon posture is to sanction a vision of historiographical responsibility wherein study can be layered on top of other studies, each espousing a different perspective on the past, none of them advocating a set explanation of events. It is, essentially, a defense mechanism against possible accusations that the author is reifying ideas about the past into a definitive vision that should not be questioned. It uses any or all of the following resources to justify itself:
- The inevitable subjectivity of the historian’s own interests in looking at the past.
- The poverty (or overwhelming wealth) of the historical record.
- The inevitable intervention of the interests of the creators and preservers of the historical record.
- The minuscule contingency of the past, and thus the impossibility of encapsulating its dynamics in a final set of principles.
The worth of the Rashomon posture is that it advocates a free play of “suggestions” for possibly fruitful perspectives. Because these suggestions are not definitive, they cannot in good sport be heavily criticized. The Rashomon posture suggests that responsibility for interpreting the contradictions between perspectives resides with the audience, or possible future research, not the suggesting historian who humbly but valiantly struggles against the profound difficulties of 1-4.
The presentation of a new perspective often takes the form of representing a neglected historical constituency, a neglected disciplinary methodology, or both. Thus people in literary studies can contribute an understanding of the “scientific public”. Art historians can analyze methods of scientific representation. Feminist scholars can contribute an understanding of women contributors to science, or the science of women. Sociologists can contribute an understanding of the normative structures of scientific institutions. We can highlight the relationship between “science” and laboratory assistants, artists, medical patients, journalists, and on and on.
Worthy projects all.
But these are worthy projects that require reconciliation into a coherent-if-complex picture. It is not sufficient to layer perspectives and assume that the historiography will somehow automatically be advanced by the effort, and it is not true (nor do I believe it is genuinely believed to be true) that we live in a Rashomon-like world where perspectives cannot be reconciled.
History has “grains” that can be identified by continuities in ideas held by distinct self-identified constituencies, as embodied in their rhetoric and practices. Querying the historical record about these ideas across these constituencies results in an incoherent response from the record. The Rashomon posture advocates an interpretation of this incoherence as an indicator of the “contingency” of the past. The summation of these ideas (as collected in the gallery of practices) represents different manifestations of the idea subject to different historical contingencies and contexts.
I make a more radical and difficult proposal. If a query of the historical record produces unrelated or only loosely related manifestations of the idea, the query is not valid. The “idea” cannot be said to have existed. The historian is forced to redefine the query, or the constituency being queried. The accumulation of manifestations does not produce new knowledge about the idea, meaning when we go to tell the history of an idea, we need to be exceedingly cautious when we select what actually constituted an idea.
“Observation”, I would argue, is an example of a research interest. Crucially, it is not a topic that can be historicized. While the term “observation” and different modes of observation can be said to have existed, I would argue that there has not been a coherent “idea” of observation. Therefore the mere accumulation of “perspectives” on observation in a gallery of practice can generate no new knowledge about, say, the “17th-century idea of observation”: it is an incoherent query. It is the historian’s responsibility to isolate and taxonomize rhetoric and practices concerning observation and to identify specific constituencies employing them. Only once this is done can the history of ideas be recounted, and, crucially, their consequences evaluated. There must be a consolidation of gains, which requires the use of a problematic. (See also Peter Galison’s “problem 7 and 8” on “locality and globality” from Isis last year, which I think is a related critique).
Studying the past according to a research interest, and collecting galleries (or Foucauldian “archaeologies”) is legitimate and useful. However, using the Rashomon posture to curtail inquiry at that point—just where the responsibility of the professional historian to reconcile the historical data through revised conceptualizations of the past becomes most useful—is not legitimate. Attributing the inherent incoherence of non-existent ideas to the contingency and complexity of contexts surrounding an incoherent “idea” without mapping those contexts is not legitimate. It is incumbent upon historians to go beyond pointing out and accumulating examples of the non-fixity of ideas and the difficulties of the historical record, and to develop complex and nuanced explanations for changes through history.
Insofar as the Rashomon posture excuses (or in some cases even celebrates) the failure to provide complex chronological accounts, it must go.