Research

At bedrock I’m interested in three things: (1) collective cognition, (2) knowledge, which almost certainly works through (1), and (3) trust, which is plainly the dark matter that supports (2).

Why? Consider the inherent frailty of an individual’s epistemic position: short of where your car is right now, what you had for breakfast, and some basic, physical, cause-and-effect relationships, almost everything you know came from someone else’s testimony rather than direct observation. I believe that smoking increases the risk of lung cancer and the earth is round, but I’ve never attempted to verify these beliefs and would scarcely know how to if I tried. I could dig up the evidence trails, but they’re yet more testimony. (“Testimony that you have good grounds to believe,” you may say—true, but those good grounds are themselves testimony.) Critically, the picture is not so different at advanced levels of expert knowledge, as John Hardwig vividly argued: the sheer number of authors on the average physics paper makes it nearly impossible that all of them know exactly how each number in the paper was derived. And yet, despite this fragile system, humans do amazing things: conducting physics experiments, developing antibiotics, building bridges. How?

My preferred beginning to an answer is the community of knowledge hypothesis, the idea that knowledge is distributed over minds in a meaningful sense. This hypothesis explains strange results like mere access to knowledge making people feel like they know stuff even though they don’t, or mere belief that others possess knowledge conferring a sense of understanding. But these proofs of concept are just the tip of the iceberg. What determines whether people trust each other’s reports in the first place, much less the reports of increasingly complex tools like AI? What is trust, anyway? How much do ingroup/outgroup dynamics influence these processes, as they almost certainly do? And just how different are we from animals that also do incredible things without complex language (hence arguments and evidence)? Sometimes this all makes me nervous, since the community of knowledge hypothesis implies that separate communities with their own cadres of experts could arrive at opposite conclusions even for matters of life or death, especially when information travels quickly and outgroup perceptions sour. But I’m also optimistic because this unprecedented connectivity affords incredible new possibilities for the study of collective behavior, and hopefully, the discovery of new ways to patch up our very old system of distributed cognition.

above: Jackson conducting research

(Okay, strictly speaking I’m interested in more than three things, like my kids, the bleak underground, the nuclear option for aesthetics, Nasreen Mohamedi…)