lpetrich

That's the way that science is often portrayed as working, both in much of the news media, and by advocates of various heretical theories. The cleverer advocates of this opinion often approvingly invoke Thomas Kuhn's "The Structure of Scientific Revolutions", which presents the progress of science as new paradigms overthrowing old ones.

But is that really the case? I think otherwise. Let us consider some notable cases.

Over the twentieth century, theories in fundamental physics have been developed that are drastically at variance with Newtonian mechanics, which had proven very successful for over two centuries. However, relativity and quantum mechanics have the important property that they subsume Newtonianism instead of replacing it; Newtonian is a special case of these theories, and usually an excellent approximation under familiar circumstances. Subsumption extends further; both relativistic mechanics and nonrelativistic quantum mechanics are subsumed by relativistic quantum field theory, a sort of meta-theory for all the elementary particle theories devised so far.

And on elementary particle theories, a critical requirement of Grand Unified Theories and would-be Theories of Everything is that they successfully subsume the Standard Model of elementary particle physics.

Going back to the mid-nineteenth century, James Clerk Maxwell composed his famous equations, which accounted for the behavior of electric fields, magnetic fields, and light -- thus subsuming all the previous theorizing about their nature and behavior.

Turning to chemistry, there had been a real paradigm shift in early modern times when the "elements" or "fundamental substances" had been redefined away from Aristotle's famous four (earth, air, fire, water). I doubt if a definite date can be set, but Lavoisier's classic work had clearly stated the modern conception of element -- and most of Lavoisier's elements are either those recognized at the present day or their oxides.

The law of chemical-reaction definite proportions led to the revival of the atomic theory of matter and the elaboration known as valence theory (each atom has a certain number of possible bonds). This, of course, subsumed Lavoisier's concept of elements. And modern-day quantum chemistry has subsumed these concepts also -- and explained various oddities.

In geology, an important paradigm since the early 19th cy. has been the dating of rocks by their order in rock formations; younger rocks are always above older rocks unless the older ones were pushed over the younger rocks or else the younger rocks were deposited inside of the older rocks. Both of which can be recognized by looking for evidence of discontinuity and the like. This paradigm was subsumed by more recent paradigms of rock dating, like radioisotope dating.

Likewise, the paradigm of plate tectonics subsumed the old paradigm of continental drift.

In biology, Charles Darwin had essentially founded the field of evolutionary biology. And ever since, the study of the history of Earth life has built on his work. The Modern Synthesis of half a century ago was essentially a subsumption of it that incorporated developments in genetics.

And in genetics itself, the paradigm that genetic information is carried by coded nucleic-acid molecules has subsumed earlier versions of Mendelism.

There are numerous other examples, and I think that one good indicator of mature science is whether new paradigms that seemingly replace old paradigms really subsume them, incorporating them as special cases or limiting cases.

This suggests that sciences with a lot of competing paradigms and no successful ones that have subsumed others are immature sciences -- or even pseudosciences. The "social sciences" often seem to be like that, though in fairness, cutting-edge research is often like that.

crocodile deathroll

I can recall a lecturer who told his class that "in 10 years from now some 50% of I am teaching you is rubbish; the problem is I do not know which 50%."

[ November 13, 2002: Message edited by: crocodile deathroll ]</p>

The Lone Ranger

I always thought that Kuhn was somewhat lacking in his understanding of the accomplishments of modern science. Many of his followers, however, have gone far beyond anything that Kuhn ever suggested.

There are those who would have us believe that scientific "progress" is simply a case of lurching from one paradigm to another, and that there's no reason to believe that science's ability to describe the world around us is getting any better. Personally, I think the notion that science is not providing us with an increasingly-accurate understanding of the world around us is very close to madness.

Today, we can send probes to distant planets. We can build machines that fly through the air three times faster than sound. We build computers that can perform millions of calculations per second. We can literally re-write a fetus' genetic code, to eliminate genetic defects. How could any sane individual consider what we can do today, compared to what we could do only 100 years ago, and conclude that we don't have a much better understanding of the world around us than even our most recent ancestors did?

In any event, there have been very few true revolutions in the sciences during the past 300 years. I think that it's clearly the case that a "new paradigm" is far more likely to subsume older ones, as lpetrich puts it, than to replace them.

I think that, in the future, our descendents will not be wondering how foolish we had been to "believe in" things like evolution through natural selection, and quantum mechanics -- but how very much we had been able to discover about the world around us in such a short time and with such limited tools.

Cheers,

Michael

Undercurrent

I find that most people who assert this have some kind of ulterior motive -- usually to debase science from its (earned) privledged epistemological position to argue that something they believe strongly in for religious (creationism/bible miracles) or egotistical (timecube) grounds should be taken as seriously as scientific progress.

The simple reason for this is that, even if a scientific theory is replaced, the original theory was believed for good observational reasons. A theory of relativity which had the planets not obeying Kepler's laws, at least approximately, would be immedately contradicted by all of those previous observations.

Similarily, even though modern thermodynamics doesn't really subsume caloric fluid theory, it still explains the observations that caloric theory explained, and explains why people used to believe in caloric. The observational evidence for caloric (put a hot object next to a cold object and the cold object will get hotter and the hot object will get colder) is a subset of the observational evidence for modern thermodynamics. We can see how someone working with only that subset can conclude the caloric theory.

Contrast this with creationism, which purports to be able to supplant the theory of evolution. Of all of the things creationism doesn't explain, the most baffling is the origin of the theory of evolution. No restricted subset of "evidence for creationism" consists of "evidence for evolution". If evolution was rejected in favour of creationism, students 50 years from now would legitimately be asking, "If there was only ever really evidence for creationism, why did people see evidence for evolution?" Short of a massive Satanic conspiracy, creationism provides no answer.

m.

Oxymoron

Overturning theories? - occasionally maybe. More commonly, it is establishing regions of validity of the existing ones and creating new theories outside those regions.

For example, Quantum Mechanics doesn't invalidate Newtonian mechanics, it just says that as things get small there's a change in the rules. Similarly Special Relativity doesn't invalidate Newton's laws, it says that things are different as you start to move fast.

New theories may radically change the way we look at the world (Kuhn's paradigm shifts), but they don't "invalidate" existing theory.

Hugo Holbling

Which "heretical theories" do you mean here, Ipetrich?

lpetrich

Various crank and crackpot theories -- their advocates often present themselves as persecuted Galileos. And the more sophisticated ones of recent decades brag about what great paradigm shifts their views are; they often indulge in the continual-overturn stereotype of science.

In fact, they sometimes seem to think that because their theories are currently heretical, that that means that they are right. But as Bertrand Russell has pointed out, there are infinite possibilities of error, and more cranks come up with unfashionable errors than unfashionable truths.

[ November 15, 2002: Message edited by: lpetrich ]</p>

Starboy

How would the theory of evolution fit into this scheme of gradual change?

Starboy

lpetrich

What, precisely, do you mean?

Guessing what you might mean, I wish to note that the progress of science has not been absolutely uniform; big paradigm shifts alternate with less-spectacular work, what Kuhn had called "normal science".

Starboy

So I guess you are saying it is mostly gradual except when it isn't.

Starboy