How Scientific Paradigms Relate

Here is a giant chart mapping relationships among scientific paradigms, as published in the journal Nature. This map was constructed by sorting roughly 800,000 published papers into 776 different scientific paradigms (shown as pale circular nodes) based on how often the papers were cited together by authors of other papers. Information Esthetics, an organization founded by map co-creator W. Bradford Paley, is giving away 25" x 24" prints of the Map of Science (you pay postage and handling via PayPal). There are also links to a 3000+ pixel wide jpg of the chart. It would be all one long spectrum except for Computer Science, which makes the connection (via AI) between the hard sciences and the soft sciences.

Comments

Torrent

[Ankur Dave]

The server is just asking to be Slashdotted, with a 5.3MB file, so here's a torrent.

Re:An obvious hoax

[be-fan]

There are lots of links. For example, there are lots of connections between the development of syntax and grammers in linguistics and the work on syntax and grammer in computer languages.

Map is itself an example of CS & social scienc

[Geof]

So an algorithm generates this map from journal articles, then lays it out as a network - and I see people on here arguing about whether the categories are biased. What more proof do you need?

Or, take a close look at social science - there's economics in there. I see asset allocation; I'm sure game theory is there too (Prisoner's Dilemma, Tragedy of the Commons, public goods theory).

What's really surprising here is not the strength of the connection between computer science and the social sciences; it's the scarcity of connections elsewhere. Where are the connections between ecology and social science, ecology and computer science? I see infectious diseases - where are the links to network theory? What about the social and communication basis for physics and the other hard sciences?

Habermas has a fascinating analysis of this. He argues that science depends on a prior consensus about how the validity of evidence is evaluated. That consensus cannot itself be scientific. In other words, scientists can't agree about the value of each other's work until they first achieve a certain level of agreement on a social and communicative level.

If that sounds suspect to you, remember that the use of the word "paradigm" debated elswhere in this discussion originates from Thomas Kuhn's The Structure of Scienticific Revolutions, which is about the (significantly nonrational) process by which science is conducted, and is grounded in philosophy, history, and social science.

Perhaps the biggest missing links here are philosophy (including mathematics) and history. But then, they aren't sciences. At least not now: there have been scientific theories of history; science itself was once a branch of philosophy. Hurrah for computer science closing the circle, but the circle shouldn't be in need of closing.

Re:I bought one..

[wbpaley2]

I apologise profoundly for this. (I was one of the co-creators of the poster, and the Information Esthetics organization distributing the print is my responsibility.)

We are using a standard Drupal shopping module and I have received two reports of this. I am sure others have seen the problem and not reported it.

We have a Drupal guru looking at that code, and hundreds of orders have cleard fine, but for now I suggest people do exactly what gammaxy did: if someone else's information show up, wait until tomorrow.

I will remain personally responsible for any mis-charged or undelivered prints. You may find me by Google-ing "Brad Paley": e-mail addresses are available on my various Web sites.

Thank you for the interest! Sorry about the glitch.

Kind Regards, Brad

Application of Graph Theory

[ct1972]

Rather ironic, the whole thing is an application of a branch of mathematics : graph theory, and yet seems to suggest that mathematics makes very little contribution to the whole thing. It really isn't believable that maths could have so few connections, this just proves that people don't see it when it is everywhere.

Re:It's terribly biased

[timeOday]

I was about to buy 3 of these, but when I actually looked closely at the graph I realised how biased it is toward the biomedical/health sciences. Math is a puny cluster of small dots, there's no area labeled Engineering and Chemistry looks like it has more lines than all the hard sciences put together.

What are your expectations based on? The chart is based on scientific publications, and IME it is representative. Federal research budget in 2004:

Life Sciences: 54%
Engineering: 17%
Physical Sciences: 10%
Environmental Sciences: 7%
Math, Computer Science: 5%
Social Sciences: 2%
Psychology: 2%
Other: 2%

So the only mismatch here is Engineering. But it's the "Map of Science," so I'm not sure Engineering really belongs in there. Math, I'm afraid, really is a puny cluster.

Medicine is where all the money is. It's 16% of the GDP! (I realize most of that's not research, but still...)