My first impressions of Jockers’ work is that he struggles to articulate what makes his method so appealing. Couching everything in economic terms, in my view, is a weird decision. It’s bad enough we have to borrow methods and tools from silicon valley; let’s try and keep ourselves as clean as possible from the rest of the biz-techno-world salvos. (28 & 29).

Plus, he goes back and forth on whether the “macro” approach he’s advocating for is capable of identifying details. Or is it themes? I think that it sort of gets lots in its own abstraction and would have been better suited to just starting off with examples (from his own work, rather than just list of new-ish examples of DH initiatives across the world).

He really takes a hammer to close reading; “Close reading… totally inappropriate as a method of studying literary history”. (7) — and i guess if you’re interested in timelines, development, etc, of course you should be as zoomed out as possible. But there’s no substitute for actually interrogating the documents in front of you (because not everything about a book can be boiled down into a data frame).

One question I had, with his advocacy, is about the way that normal people read books. Our computers behave very differently than readers. And there are details that they catch that readers don’t. But do we lose anything when we totally replace the way that normal people consume this media? Nobody reads like a computer. We’re not supposed to be able to read distantly.

Jockers then begins to summarize actual work he’s carried out. This portion of the book has been much more enjoyable so far. His chapter on Irish-American writings was interesting because he employed a relatively simple classifier scheme but made some pretty robust material out of it. His next focus is on genre, which, while cool, I don’t have too much of an interest in — I wouldn’t consider myself a literary historian in the slightest. He does a good job of explaining how the day to day work of a data science type project actually works, which was helpful. However, I’m struck by the fact that his “macroanalysis” doesn’t really seem that different from distant reading. Perhaps he thinks he’s doing something different because he is a little less adventurous than Moretti — barely ever bringing the text back up, trying his hardest not to make conclusions of any sort (which, he thinks are impossible for some reason), but the bread and butter techniques of distant reading still shine through in his work… and how could they not?