Printable Version of Topic

Posted by: ustrax Mar 22 2007, 12:59 PM

"This http://www.seedmagazine.com/news/uploads/scimaplarge.jpg 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".

From http://www.seedmagazine.com/news/2007/03/scientific_method_relationship.php.

Just see the new branches of the tree of knowledge expanding...
Very interesting and quite a beautiful work of art!

Posted by: ngunn Mar 22 2007, 03:59 PM

Very nice, that! Interesting that it should take up a ring-like configuration. What, I wonder, is the significance of the hole in the middle? Connections we are not making?

Posted by: ustrax Mar 22 2007, 04:33 PM

It is isn't it?
Have you seen http://mapofscience.com/? It doubles the number of paradigms!
I would see that hole maybe as ourselves, establishing a web of knowledge all around us, that will eventually grow larger and larger, with new areas of study and consequently new paradigms...

Posted by: nprev Mar 23 2007, 02:20 AM

Hmm...very perceptive observation, ngunn.

To me, the hole indicates a sort of serial (spiral?) approach to science (yet another artifact of our tendency to categorize nature al a the Pluto debate?... )...physics begets chemistry, chem begets biology, biology feeds biochemistry and thereby medicine, and the social sciences are hooked in via the psychiatric branch of medicine. Maybe there are indeed missed connections, but a direct reachback to physics from the social sciences seems unlikely...perhaps via chaos theory and/or nonlinear systems behavior? If this is considered as a separate discipline, it may well occupy the center.

I find it interesting that engineering seems to 'ride' along physics. Although definitions may vary, it seems surpising that engineering is under-represented in biology and medicine. Maybe this explains why, for example, artificial heart technology seems to be lacking. I've always wondered why a variable-response pump should be so hard to build...

EDIT: Re the center: Shouldn't this at least symbolically be occupied by mathematics?

Posted by: AndyG Mar 23 2007, 08:54 AM

I'd rather see mathematics as the quantum vacuum underlying this whole structure. ;-)

Andy G

Posted by: ngunn Mar 23 2007, 09:42 AM

Looking again at what closes the ring - it's a very modern connection, via computing. Otherwise we would be looking at a linear spectrum from maths/physics to social sciences. Computing provides the link because it can model systems that are too complex (or poorly defined) for explicit, rigorous analysis. The surprising thing is that the social sciences seem to have stronger links to computing than do subjects like, say, meteorology or organic chemistry, both of which use computer modelling routinely (but perhaps don't cite the papers)? I wonder which way the citations go. Do computer scientists cite sociologists? We need vectors on that diagram.

Posted by: nprev Mar 23 2007, 02:54 PM

Probably vice versa. Statistical modeling/linear programming to solve problems with large numbers of variables is really taking off, so I suspect that the social sciences are beginning to mine & analyze their databases in a rigorous way.

Posted by: ngunn Mar 23 2007, 03:12 PM

That's what I was hinting at. I think if you put arrows on the whole diagram most of them would point in one direction - ultimately towards the maths/physics end of the spectrum. One can imagine new sociology emerging from mathematics, but hardly the reverse.

Posted by: edstrick Mar 25 2007, 09:00 AM

I'd *LOVE* to see a similar map of pseudoscience nonsense... Where would phrenology, numerology, homeopathy, visiting space aliens, cattle mutilations, and bigfoot plot?

Oh.. and the we-never-went-to-the-moon morons.

Posted by: RJG Mar 25 2007, 12:23 PM

And to see what can be done, Hans Rosling takes some beating:

http://www.ted.com/tedtalks/tedtalksplayer.cfm?key=hans_rosling

Posted by: nprev Mar 25 2007, 10:43 PM

I suspect that the map would be far less orderly, if there were any real connections at all. Perhaps that fact serves as a good discriminator between science & pseudoscience. Science advances & diversifies based on proven facts obtained by critical examination and testing of alternative theories, while pseudoscience generally embraces a single hypothesis regardless of alternatives and thus becomes quite narrow.

However, what's kind of interesting (and revealing) is that you could superimpose the pseudos over actual scientific disciplines & draw connections. Some of these 'disciplines' were ancestral to real sciences, like astrology was to astronomy. Remember, we've only had some actual idea of how things in the Universe really work for a pitifully small fraction of human history, and before that everything was <clinking> magic...no wonder the pseudos still have such a powerful grip on a significant fraction of the world population.

Posted by: CosmicRocker Mar 26 2007, 05:13 AM

Wow, this is really interesting. Thanks for posting that, ustrax. Ages ago I thought it would be interesting to try to draw a picture of science. I had a vague idea of sort of a scaffolding structure that reached out and linked up with other centers of knowledge, but it never occurred to me that this could be done in a somewhat rigorous way. Having no artistic talent whatsoever, the thought never went beyond that.

It wasn't apparent to me at first, but the map at http://mapofscience.com/ is interactive, allowing one to highlight it's parts. Computer science is apparently lumped in with engineering as an applied science. That whole segment provides a lot of scaffolding for the overall structure, and that makes abundant sense, since engineers and computer scientists are tasked with solving a wide range of problems. It is interesting to highlight the various scientific disciplines and see their individual contributions. There is a strong tendency to make close connections that reach out to neighbors, but sometimes you see outliers where a node from one discipline sprouts up from unfamiliar territory, in a manner reminiscent of the panspermia theory.

I think there might be more interesting details in the database they tapped. It sounds like it would be fun to slice and dice it with that Sandia software they cite.
"Visualization uses VxInsight, a proprietary software package from Sandia National Labs."

In the end, I guess one could argue that it is all physics.

Posted by: MizarKey Mar 26 2007, 05:16 AM

I want to start out that I don't believe in astrology as it is practiced today. But I am curious about it's origin in the 'mists of time'. And whenever I see or hear the phrase "how things 'really work'" I have to kind of smile because I don't think we have any difinitive answers, just a lot of theories that fit current observation. But as more observations are made we are constantly finding things we didn't anticipate in the theories.

We know that there are complex interactions between the sun and the planets and the planets and their moons and the planets and the other planets. For instance, the documented disturbance in radio communications due to Saturn and Jupiter's position in relation to Earth.

Nelson, John H. "Planetary Position Effect on Short Wave Signal Quality." Electrical Engineering 71, no. 5 ((May 1952): 421-24.

——. "Shortwave Radio Propagation Correlation with Planetary Positions." RCA Review 12 (March 1951): 26-34.

Of course this is not the astrology as practiced today, but does infer that there are interactions we may not fully understand. Just as we may not fully understand what effects may arise from the position of the sun relative to the galactic center.

Back on topic, the graphic of the interconnectedness is cool but I'll have to read more about it to understand that it's not biased in some way by how the data was input or how they chose to create the connections.

Posted by: AndyG Mar 26 2007, 08:39 AM

That's excellent - many thanks!

Andy