In email to a friend, December 2 (2016)

A book I love is “Proofs and Refutations” (Imre Lakatos). A big point of that book is that people tend to think of definitions & categories as starting points, from which we derive theorems or scientific results about the world. But the process really works the other way round. Definitions and categories are things which emerge _after_ we’ve understood phenomena in the world, and as a consequence of that understanding. And so while it’s worth making very tentative definitions, it’s a mistake to do so too prematurely, or to hold too strongly to them. Those tentative definitions are, at best, scaffolding to help arrive at an understanding, which may eventually result in good definitions.

(Lakatos's particular examples are extraordinary, if you haven’t read him: he explains the 150 or so years it took to arrive at the modern definition of polytopes, simplicial complexes, homology etc, all basically as a consequence of trying to understand Euler’s idea for a proof of his formula for a polyhedron: E - H + V = 2. He also has an appendix where does the same thing for notions such as differentiable and continuous functions.)

All of which is a way of saying that I tend to be skeptical of definitions & categories early on, and reluctant to commit. My book on open science and collective intelligence doesn’t define either (perhaps I carry this perversely too far, to be honest). Rather, my approach was to gather examples and ask “How can I best understand these? What unifying principles and governing mechanisms are there? What differences” Having made some progress on those questions, I’d actually now be somewhat comfortable defining open science, though not yet collective intelligence.

The same is true of interfaces, media, symbol systems, etc. We don’t yet know what an interface is. What seems clear to me is that your examples - Venn diagrams, causal graphs etc - do provide very interesting vehicles for thought, at least in the hands of experts. In fact, the particular Venn diagram I use in this essay is something that recurs through all my writing (only the labels change): I find it one of the most generative tools I have for thinking (and feeling) about the world.