Are Kuhn's novelties still timely?

Daniel Kuehn recently wrote here about Ryan Murphy's reading of Thomas Kuhn, drawing from a couple passages in the scientist's groundbreaking work The Structure of Scientific Revolutions. Here I respond briefly to Murphy and then I launch into my own - and likely highly idiosyncratic - examination of Kuhn as I am in the process of reading him (for the reader's sake I have defined the break in topic). I wrote my personal response to Kuhn first, and my response to Ryan Murphy followed.

On the appearance of tautologies in the sciences, for Ryan Murphy

It may be useful to consider the immediate history of quantum investigation which provided the backdrop for Kuhn's work as a scientist. It is not clear to me that many in the Copenhagen School of quantum theory were preoccupied with a case for dividing universal and proximate explanations; they thought of themselves as practitioners of "hard science" and empiricists through and through. Neither is it clear (I went on this topic in more depth at a post on my blog) that they were convinced of a possible gap between achievable and unachievable explanations. If there is a distinction in that vein, I believe explanation may depend upon and correlate to the type of explanatory framework utilized - whether it be mainly based on empirical work or pure reasoning (if indeed Kuhn respects this distinction) may determine whether you can assert a result, and empiricism might be thought of as merely a more proximate kind of result.

Newton certainly respected the difference between the two. He thought that there may be a sound mechanism for creating gravity - if only God. While untestable by empirical methods, which Newton recognized, it is not necessarily a tautological explanation for gravity. What is interesting is that the reasoning process used to divine the Divine, for Newton coming to the conclusion that God might be behind gravity (I fear I am sorely abusing Newton's argument here) is a tautology or circular argument, but one which involves empirical evidence at a point in the chain. It would be a disservice to Newton to argue that he didn't understand that allusions to God or anything empirically unproven could be less than useful in science - he had a profound aversion to what he called "mere hypotheses."

The appearance of validity of this mode of inference, even given the logically formalized assertion of its invalidity (which appears in some sense, to me, arbitrary and limited to our own sphere of influence) which denies it, is an open question as far as I am aware.

I haven't read the whole paper yet, but Dr. Timothy McGrew makes some sound arguments here about the appearance of relativism in Kuhn's work. I haven't finished Kuhn either (via anthology) but it is fairly striking that Kuhn seems to admit no strong foundation for knowledge, and that is the basic observation from which most of the criticisms extend.

A personal response to The Structure of Scientific Revolutions

What puzzles Ryan Murphy puzzles me as well. Kuhn's writing reminds me of how I often attack subjects - trying to delineate a problem by categorization, and finding implications for the shape of the enterprise in the relationships of parts, especially long historical passages.

I was aware when I was still very little that there was the appearance of a progression from one theory to another - I knew that there was something the ancients thought, then there was Galileo, then Newton, then Einstein, and now there were newer things still. Kuhn argues there is not a smooth progression here - he even says that in many respects Newton's laws may have retarded the course of science, albeit in practical and useful ways, along with ways that were perhaps not so. However, there are some distinct possibilities that I don't see Kuhn grasping with. One is that, as I thought when I was ten, maybe empirical knowledge would keep getting more finely and finely grained, as one answer suggests ten questions; the structure of truth ends up being like Descarte's atomism without voids, where increasingly smaller spheres fill the space between other spheres - more finely and finely grained "truths" to find add smaller and smaller refinements to the structure of a model. This is not a revolution. If true, Kuhn's process of revolutions would not necessarily hold more sway in a highly mature system, especially if the nature of the incremental and diminishing theoretical improvements became apparent.

Another possibility is that Newton's laws, or something much like them, are not merely a paradigm to be overthrown, but something actually useful as a starting point for a better approximation of the universe. (I interpret one of Quentin Smith's recent lecture points, in his "Religion and Cosmology" course, as suggesting that Newtonian physics with some modifications might better explain phenomena previously thought solely the province work by Einstein and since.)

In Kuhn's Introduction he writes of "normal science," which I take to be the process of trying to refine existing "paradigms" or the structure of science. He speaks, for example, of trying to box the facts into the paradigm - a classic complaint hailing back at least to Aristotle.

At this point he mentions two examples of signal events which usefully break from this paradigm - the refusal of a puzzle to be solved, or (this second is an area he seems to return to more often) the result of an experiment failing to match the known prediction.

The first area is worth more elaboration, whose implications Kuhn apparently does not treat to my satisfaction (at least in my cursory examination) when he concludes abruptly in this passage that "normal science often goes astray." I hope to see how he treats the following problem I pose later in his work: What happens when normal science does not go astray, but instead we have reached the end of theory and experiments? This scenario is suggested by his first signal event.

He leads us with this passage to expect that the theory or paradigm is wrong and that there is a further answer - in the history of scientific revolutions. However, since at least the 1930s, Kurt Godel's incompleteness proofs provide a basis for suggesting that it may well be impossible to prove all the axioms of a usefully complicated self-consistent system (i.e., a system of arithmetic, in Godel's proof), and for thinking that there may be created a theory as perfect as can be obtained. Improving it only requires some perspective not afforded us; i.e., a perspective outside our seat in the midst of dimensions, a perspective not reined in by limitations in the scale and energy of experiments that may be conducted, perhaps a perspective outside time and space.

Kuhn's method of inquiry is necessarily constrained by one's time period, and by the availability of a list of all the possibilities (Baconian inductivism). Kuhn was perhaps lead on by the remarkable history of science at the time he was an active participant. His time in quantum mechanics was exemplified by new discoveries in quantum events leading to refinements of quantum theory. Today, on the other hand, we have the unique distinction of wondering where to go for a falsifiable theory as the Higgs Boson appears to have been found - to the chagrin of many scientists (and science followers) hoping for the reconstruction of the theory of the universe. We have a mechanical "explanation," by one measure, of events (in the case of the Higgs boson, the particle contributing the small remainder of the mass of objects in the universe); by another, we do not have an understanding of the causes of those events. This is a point on which I understand Kuhn attracts criticism. (My discussion of Newton vis-a-vis Kuhn for Ryan Murphy explains this criticism above.)

I need to finish up the section before I can say much more. I am impressed (as always) by the thoughtfulness and general usefulness of Kuhn's writing. However, he appears to suppose some things which may not still be supportable. We are left with something with an appearance like Goodman's new riddle of induction: What if the facts only hold so far, until they are replaced abruptly by other facts?

Kuhn deals more or less directly and repeatedly with this theme, not so named, throughout the introduction and the first sections of his work. Instead of absolute facts changing, he treats relatively known facts, the hidden or "layered" complexity that unravels with closer and persistent scrutiny - not the actual values of the world changing. That will be my foremost question for the reading: How do you discern the two, and is this a fair point to make? I think that the application of the riddle can have some surprising consequences - not only might you say (as he does) that science only moves to a more useful analysis on the discovery of (absolutely or relatively) new facts and a closer-fitting paradigm, but you might also say that the appearance of a fact might be merely contingent on a fact - this is absurd and I think Kuhn treats it as such, but how then do we determine whether (as per Goodman's riddle) whether a fact is absolute or relative to our temporal (or some other) reference? This is merely another complication on top of that which I already mentioned, the lackuse of revolutions when the facts are actually known.

And, following Wittgenstein on helio- and geocentrism, how would the world have looked as if the progress of science was not by revolutions? Even more strongly than the appearance of the sun going around the earth (rather than otherwise), the appearance of scientific revolutions may be the product entirely of our historical perspective. On the other hand, why should the theoretical virtues and other methodological tools not share in this?