Epistemic status; attention conservation warning.

by waxbanks

Reading Scott Alexander’s Slate Star Codex (one of the best blogs out there, no question), I’m reminded of a feature of his blog that I wish were more widely adopted: the epistemic status note at the top of a post.

A recent example:

Epistemic status: idea for one’s toolbox of ideas; not to be followed off a cliff


Epistemic status: So, so speculative. Don’t take any of this seriously until it’s replicated and endorsed by other people.

You might think this is humourlessness, or the author assuming his readers’ humourlessness or poor reading comprehension, and some idiot is probably getting ready to use the phrase ‘Swiftian satire’; please don’t. What Scott is doing is suggesting one or more reading frames for his readers, in order to shape both their approach to the posts and the discussions that follow. But crucially, this isn’t about content — it’s just additional information about how strongly certain claims are intended to be taken.

This is the most important thing, I think, and the strongest indication that Scott’s site is ‘grownup’ in a way most modern-USA ‘intellectual’ discourse simply isn’t: he assumes that the point of his writing is to generate and contribute to robust adult communication, and avails himself of the right tools for that job. Moreover, he does not assume that his readers will agree with him (and they often don’t) — only that they’re willing to read in good faith and assume that he’s writing in the same spirit.

This isn’t quite the same as a content note: if you look at, say, shakesville.com, the ubiquitous content notes often (usually?) function as neutral guides to topics under discussion, but surprisingly often serve as editorial prefaces, e.g. a (hypothetical) discussion of rates of gender-detransition might be framed with a ‘transphobia’ content note. The purpose of such notes isn’t to increase reader flexibility, and they don’t assume readers’ good faith — they’re there in part to shape the readers’ attitude toward the content itself. They aren’t just warnings to stay away, of course: most readers will read the posts regardless of the content notes. For those readers, the content notes are just guides to reception posture at the level of content.

Scott’s ‘epistemic status’ warnings guard against unproductive forms of argument but are agnostic as to reader perspectives; Melissa’s, I’d argue, militate subtly against specific perspectives. Both are intended inclusively, I think, but my sense is that they don’t both function that way, at least not to the same degree.

The great and knowledgeable Cosma Shalizi includes ‘attention conservation notices’ atop his long posts, which are somewhat more complicated (or at any rate pretentious) than normal content notes/trigger warnings.

In theory, credentials serve as persistent epistemic status warnings: ‘I have a PhD in area XYZ, so I can be expected to know A, B, and C.’ But life is complicated and dumb.

But again: why would you take my word for any of this?