From: Anders Sandberg (asa@nada.kth.se)
Date: Tue Jul 23 2002 - 02:23:29 MDT
Kardashev classes are too crude to be really useful; they are better to
categorize supercivilization A from supercivilization B ("The A:s have
built Dyson spheres around all the stars in their galaxy, the B:s have
converted half of them into computronium and are using small black holes
to convert the other half into energy"). They don't tell us much about
humans, and even less about scientific output.
Counting publications is better, but assumes 1) that publications have a
constant scientific content, 2) that their number is not affected by
other factors such as economics. A look at
http://www.nsf.gov/sbe/srs/seind02/c5/c5s3.htm#trends shows that the
number in the US seems to be actually *decreasing*, for a number of
factors detailled in the article. So this measure seems to be hopeless
too :-(
What we really need to have is some kind of measure of how many new
facts are discovered, and how many new theories can explain the facts.
Facts are essentially data about the world; we likely need some kind of
chunking not to get distortion from a few huge databases like Hugo and
Hipparcos. I'm not sure how to do that. Maybe a fact can be defined as a
paper reporting it, so the total amount of facts would simply be the
number of papers reporting stuff rather than explaining them. This runs
into the old problem of paper economy.
The power of a theory is the amount of known facts it can explain given
a short list of assumptions and parameters, essentially the ratio of
facts / theory size. What would be really useful is some plot of total
theory power per year over time. Maybe one could do a crude estimate by
taking a few issues of major interdisciplinary journals from a given
year and estimate the power of each article not just reporting a fact by
dividing the number of references by the length of the theory part. Not
sure if it would work, but it might be worth trying. Just a quick look
at a journal on my desk suggests there are rather few theory papers in
most journals, most are simply fact reports.
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:15:38 MST