Monday, April 3, 2017
"Hindsight bias" and doing automotive history
Hi folks -- I read Michael Lewis' Undoing Project over the weekend. Lewis is the author of Moneyball and The New New Thing, among other titles. The author focuses on two Israeli psychologists and their theories of learning and perception. In this engaging narrative, a few pages ( pp. 202-8) cover an episode during which one of the main characters addresses a group of historians in 1973. Amos Tversky often said "It is amazing how dull history books are, given how much what's in them must be invented."
According to Tversky, historians are prone to distort the intuitive interpretation of data. These judgments of past events are subject to an inherent bias, and perhaps this is one reason why well-worn reconstructions of automotive history keep coming back to us time after time. Errors and interpretations are perpetuated. And as a further consequence, writing history is essentially shaped by irrationality.
The problem that historians face is that we write from the present, and thus we know (or better, think we know) the outcome of an event. Knowledge of that outcome influences how we reconstruct it. Our occupational hazard then, involves "the tendency to take whatever facts that were observed (neglecting the many facts that were not or could not be observed) and make them fit neatly into a confident-sounding story.
In sum, a flaw in our reasoning leads us to believe there is a less uncertain world than there actually is ( i.e. "the lessons of the past"). Thus, we revise our narratives so the stories seem to fit what happened. Most histories then, grossly simplify a far more complex and random world.