A Map of Recent Philosophy
October 16, 2018 6:20 AM Subscribe
"This graphic is my attempt to give a data-driven representation of the structure of recent philosophy. ... For this map I parsed 55327 papers in philosophy from the Web-Of-Science-Collection."
A previous version of the map, with code samples and explanations.
A previous version of the map, with code samples and explanations.
The "Spinoza & Descartes" blob appears to be undergoing mitosis, as does "Aesthetics & Art". It sounds like that's a consequence of the cluster size he selected. I suppose a knowledgeable philosopher would know what the two distinct nuclei inside of "Marxism" represent.
posted by paper chromatographologist at 7:22 AM on October 16, 2018
posted by paper chromatographologist at 7:22 AM on October 16, 2018
The ways in which the final and previous versions of the map differ look interesting, though I don't know enough about the discipline to understand the differences.
posted by clawsoon at 7:37 AM on October 16, 2018
posted by clawsoon at 7:37 AM on October 16, 2018
i've thought about doing something similar in my field but I'm worried it will just confirm what I already fear: that neuroeconomics has ruined everything
posted by logicpunk at 11:03 AM on October 16, 2018 [3 favorites]
posted by logicpunk at 11:03 AM on October 16, 2018 [3 favorites]
the author uses HDBSCAN, which is an incredibly good clustering algorithm, with a Python/C implementation that is both fast and easy to use -- it should be the first thing you try if you want to cluster things. it requires way less tuning than anything else I've seen.
posted by vogon_poet at 11:28 AM on October 16, 2018 [2 favorites]
posted by vogon_poet at 11:28 AM on October 16, 2018 [2 favorites]
wow, I want to subscribe to a "ten minute a day" newsletter that explains each branch.
Also, I looked for nazis and randites but didn't see any. So there's that.
posted by rebent at 12:25 PM on October 16, 2018 [2 favorites]
Also, I looked for nazis and randites but didn't see any. So there's that.
posted by rebent at 12:25 PM on October 16, 2018 [2 favorites]
It's more than a little strange to me that the following are all distinct clusters, some of which seem to be very widely separated: philosophy of science, philosophy of biology, theory of science, philosophy of psychology, QM, spacetime, mechanistic explanations, and causality.
The division between epistemology and philosophy of science doesn't surprise me at all. The bits of epistemology that usually strike philosophers of science as interesting are essentially all in "decision theory," which is a distinct cluster that is right next to phil sci in the diagram. Analysis of knowledge, brain-in-vat skepticism, internalism v externalism, ... those usually strike philosophers of science as pretty worthless. (I think that's a mistake, but it's a mistake that I also find really intuitive.)
posted by Jonathan Livengood at 1:12 PM on October 16, 2018 [1 favorite]
The division between epistemology and philosophy of science doesn't surprise me at all. The bits of epistemology that usually strike philosophers of science as interesting are essentially all in "decision theory," which is a distinct cluster that is right next to phil sci in the diagram. Analysis of knowledge, brain-in-vat skepticism, internalism v externalism, ... those usually strike philosophers of science as pretty worthless. (I think that's a mistake, but it's a mistake that I also find really intuitive.)
posted by Jonathan Livengood at 1:12 PM on October 16, 2018 [1 favorite]
Well, he got that it’s a mess, at least.
posted by Segundus at 2:54 PM on October 16, 2018 [1 favorite]
posted by Segundus at 2:54 PM on October 16, 2018 [1 favorite]
I mean projecting down into two dimensions, you sort of can’t help but have some things be far apart that should be close.
Probably something like d=10 would give more reasonable clusters, but the advantage of 2D is you can view the infographic without first smoking a lot of DMT.
posted by vogon_poet at 3:24 PM on October 16, 2018 [3 favorites]
Probably something like d=10 would give more reasonable clusters, but the advantage of 2D is you can view the infographic without first smoking a lot of DMT.
posted by vogon_poet at 3:24 PM on October 16, 2018 [3 favorites]
Maybe sorta related: The History of Philosophy, summarized & visualized
posted by Bron at 7:07 AM on October 17, 2018 [2 favorites]
posted by Bron at 7:07 AM on October 17, 2018 [2 favorites]
Maybe sorta related: The History of Philosophy, summarized & visualized
That's very nice! I have some quibbles with how it scrolls but it's fun to play with.
posted by thelonius at 10:00 AM on October 17, 2018
That's very nice! I have some quibbles with how it scrolls but it's fun to play with.
posted by thelonius at 10:00 AM on October 17, 2018
« Older "just whose side was Virgil on?" | Apiarists Gone Rogue Newer »
This thread has been archived and is closed to new comments
posted by Made of Star Stuff at 7:03 AM on October 16, 2018 [13 favorites]