#alsodata

I remember a thing, several years ago, when Border’s was still extant.

I was walking with a friend in the northern part of the city, before the hills begin, and we stopped into the Border’s on the corner of the main street.

The front part of the store, where they used to keep “the good stuff”--the most popularly attractive material--was choked with memoirs.

It was an unusual thing, because, in my experience, this section was generally choked with mysteries and thrillers, popular journalism maybe, the most recent set of dry and interminable attempts at “contemporary literary fiction”, and maybe some fantasy or sci-fi from a best-selling author. Not stories about people talking about themselves. A memoir is a strange thing--compelling because it’s mostly true, limited by that same factor. I thought it would pass, in a year or two, like literary trends do--and join mysteries involving single, middle-aged women with unusual professions, grim stories about apartment-hunting and not being in college anymore in New York, and hyper-pop social criticism in the dustbin of history--or at least the back-shelves. It didn’t.

Sometimes, it seems like we live in the age of anecdote--where personal experience determines is both the boundary of intellectual possibility and the sine qua non of empirical authority.

You can hitch a point to your personal narrative and have it carry, if you want to know about a thing, you go to ask the people who’ve done it or who were nearby--if you work in a quantitative area of social analysis, which is where I am usually located, you build datasets out of what people say they think, or what they say they do, which is anecdote writ large. Memories bound together by math. It is a thing I am ambivalent about. Inasmuch as the models we work with are fundamentally incomplete, the presence of someone else’s expert knowledge about their own experiences and activities can blow an analysis wide open. It prevents you from repeating what you’ve heard in general and saying it is both what you think and true.

On the other hand, it bounds possibility--the number of things you can think, the ways you can think--so stringently. Anecdote is biased towards some mixture of now and then, was and is, but it is hard to leverage towards will be. When I run a model, I can only tell you, with any authority, what happened before. Unless I do some slightly weird things, and take a much larger guess than is usually acceptable, it is very hard to say what will happen when things change and even harder to say what the range of things that might have happened. And I cannot tell you, working from the data only, what was not considered, when it comes down to it. You need theory for that--alternative possibility. Conclusions drawn in this way have some authority, for what they are, but also an illusory sense of completeness. This narrative mode--in the big or in the small--is often very bad at talking about things that are not itself, or delineating its own boundaries. It’s concreteness, and its firm linkages to the world that created it, make it very difficult to question, also.