Kuhn 4: What Is "Normal" Science?

[koganbot]

Further guidance for reading Thomas Kuhn's essay "What Are Scientific Revolutions?" which you can find complete here, pp 13 to 32. I'd say try to finish the essay by Tuesday, which is when we'll start talking about it unfettered, though if you've got questions and quotations to post before then, feel free to do so in the comments.

You could say that "scientific revolution," "paradigm shift," and "incommensurability" are mutually defining: Q: When is a change in a scientific discipline enough to be considered a revolution? A: When the change involves a paradigm shift. Q: What's a paradigm shift? A: A paradigm shift occurs when a set of interrelated concepts and terms and theories are replaced by a different set of interrelated concepts etc. that are incommensurable with the first. Q: In what sense are these sets incommensurable? A: The sets are incommensurable when the difference between the new and the old is so revolutionary that there's no way to map new concepts etc. onto the old or vice versa in a way that would make the resulting science intelligible.

OK, but must the fact that there are paradigm shifts necessarily mean that there are periods of scientific "revolution"? I think that Aristotle's concept of motion is unquestionably "incommensurable" with Newtonian physics (as Kuhn is using the term "incommensurable"), and if you were trained in one you'd have to shift paradigms to understand the other. But Newton and Aristotle lived a couple of millennia apart. So, placing Aristotle right next to Newton, you see a revolution-sized gap, but couldn't that gap be due to the accumulation of a lot of small evolutionary adjustments over 2,000 years? Imagine that the earth undergoes a lot of tiny tectonic shifts and readjustments but never has an earthquake; nonetheless the tectonic plates could have moved the same distance over a period of time that they would have were there periods of little movement interrupted by large earthquakes.

Similarly, the alternative to there being scientific revolutions would be that words and concepts and theories are always being contested in little ways and are always being adjusted in the face of anomalies and difficulties. So over time you get the same "incommensurability" that Kuhn observes, the same holistic and noncumulative change (noncumulative in that, e.g., later scientists didn't merely build on Aristotle's achievements: some of his core concepts disappeared and were supplanted by other concepts). And intellectuals who want to get all excited or upset by incommensurability and theory-dependent facts and such still have something to get excited or upset by. But we don't necessarily have to have periods of "normal" science (i.e., nonrevolutionary science) interspersed with periods of "revolutionary" science. We can have evolution rather than revolution, little shifts always happening rather than periods of no shifts interspersed with periods of big shifts. We can say that yes, some of the evolution happens more suddenly and in bigger increments than others - a big one would be Copernicus deciding in the early 16th century that earth was a planet or Ehrenfest and Einstein independently figuring out in 1906 that it is impossible to derive Planck's black-body law without positing discontinuous energy, i.e. energy jumping from level to level rather than increasing or decreasing continuously - but overall, big or small, the shifts and adjustments in language, concept, and theory never stop happening.

Whereas Kuhn is arguing that we get periods of "normal science" during which a field undergoes "normal change" and that we get contrasting periods of "revolutionary change." In "normal change," none of the major terms and concepts are being contested, scientists solve problems by applying the (small-p) paradigms they've got to new situations, and change is cumulative in that it builds on the work of predecessors rather than overthrowing some of that work. A revolutionary period of change is where paradigms compete with one another, scientists are creating new concepts and models, but no set of concepts has achieved enough coherence to take over the field yet. So Copernicus and Planck/Einstein/Ehrenfest changed the way their respective games were played, the language and concepts going into years of significant flux and not coming to rest until words, concepts, theories etc. were integrated into a largely coherent whole - after which, normal science returned, using new concepts and models.

Despite its name, I want us to think of Kuhn's idea that there is "normal change" and "normal science" as strange and challenging. We should ask: (1) Does normal science as Kuhn conceives it actually exist? (2) If it does exist, does the equivalent exist anywhere other than in science? --I'll say flatly that it sure doesn't seem to exist in music criticism or in art. What's normal in science may not be normal elsewhere.

In thinking about what Kuhn means by "normal science," get rid of the idea of a majority versus a minority, a dominant (big-p) paradigm* versus marginal ones, the dominant one being the one that a majority or a strong plurality of people in the field adhere to. Rather, think of everyone in the field working within the dominant paradigm (though this doesn't mean that there can't be failed "revolutions," where someone tries to make a major conceptual switch and it doesn't catch hold).

*Remember that "paradigm" has two meanings: it can be a specific model and it can be an overall practice. I'm distinguishing them by calling the former a "small-p paradigm" and the latter a "big-p paradigm."

Comments

[skyecaptain.livejournal.com]

To take a smaller example, like diabetes research, I think we're approaching something like a big-P paradigm shift in understand the nature of the disease -- until recently, it's been understood as a malfunctioning of the pancreas itself, whereas now there are schools of thought that have been fairly persuasively painting it as an autoimmune disease that can not only be stopped (after whatever damage has been done to beta cells, which are crucial to insulin production and are attacked by the body in diabetes mellitus, Type 1) but reversed. It goes against everything that is understood about the disease, which has been treated (assuming that the beta cells are dead 'n' gone) as something for implanted islet cells or, in some cases, an actual organ transplant -- something that inevitably fails within a year or two, precisely because the nature of the problem hasn't been understood.

I'm not sure if this actually qualifies as a big-P paradigm shift. But if it does, it still raises a question that might still apply: couldn't it be possible for there to be a major adherence, for a significant amount of time, of some in the scientific community to the old paradigm? At some point, everyone will accept the science, but wouldn't there necessarily (potentially anyway) be periods in which it really is a matter of dominant versus marginal (but incommensurable) paradigms being followed?

Would this be at all applicable to, e.g., history, if for instance new evidence were to erase our general concept of how events of the past happened? Or would it not actually challenge the very practice of history, since taking in new evidence and recalibrating based on it is the standard practice of historiography etc. anyway?

(Off to actually read some Kuhn...)

[koganbot]

couldn't it be possible for there to be a major adherence, for a significant amount of time, of some in the scientific community to the old paradigm? At some point, everyone will accept the science, but wouldn't there necessarily (potentially anyway) be periods in which it really is a matter of dominant versus marginal (but incommensurable) paradigms being followed?

I'd say quite certainly, I'd expect this to happen more often than not. But we can draw a distinction between (a) a period of time after which the "winning" model has been put forth but contrary models still have some plausibility, (b) the winning model has been nailed down but some members of the field still won't accept it (and may not even understand it), and (c) all the holdouts have either died off or for practical purposes they're no longer part of the discipline.

I'm not sure if I'm remembering this right (having been young at the time), but I think that after Newton's Principia, which pulled everything together in regard to gravity and the laws of motion and the motions of the planets, there was a period during which some people still challenged it, because they found the concept of gravity problematic, an occult type of force of no explanatory value. Newton himself found the concept problematic, and I'd guess that Cartesians would have had trouble considering "gravity" a part of the physical world, Descartes a couple of decades earlier having defined physical bodies as those that have extension and are divisible whereas thoughts and selves and minds are neither extended nor divisible, and where the hell does that leave gravity? So no doubt there was a time lag, but eventually it would have come about that if you weren't a Newtonian you simply weren't a physicist. (Compare to fine art, where there's never a situation where if you're not part of the dominant school then you're just not a painter.)

I'm hardly the one to judge the diabetes example; it would make sense that someone would call it a paradigm shift, but that may be giving "paradigm shift" a weaker meaning than Kuhn would give it. Thinking by analogy, let's say you're a police inspector who's investigating what you think is a murder and all of a sudden convincing evidence comes forth that (a) the bullet was fired into the body after the fellow died accidentally, (b) making this an altogether different kind of crime (perhaps now you're investigating an attempted frame-up rather than a murder). We could say your understanding of the case has undergone a shift, gone from one model to another. But your understanding of crime hasn't changed. Murder and frame-ups are already part of your world. Whereas before Copernicus, "planets" as we now conceive them didn't exist as a concept, even if the word "planet" existed. And over the next century and a quarter, scientists became aware of all sorts of entities and forces - satellites, pendulums, gravity - that they hadn't formerly even conceived of.

So I might not want to call the shift in the understanding of diabetes a paradigm shift even though it does seem to jump models, from pancreas dysfunction to autoimmune disease. Both those concepts already exist, just as the concepts murder and frame-up already exist (unless of course the study of diabetes brings out a wholesale change in our understanding of autoimmune disease and related concepts). But nonetheless, I could say that your diabetes example gives a sense of what a paradigm shift feels like.

(I don't know much about medicine, but I'd guess that there are whole areas within it where one can question whether what's going on has yet coalesced into a science. No dominating and convincing big-P paradigm has even formed - e.g., addiction treatment, chronic fatigue syndrome, post-traumatic stress, pain control by way of acupuncture, just to name a few. But the absence of an agreed-upon paradigm doesn't necessarily mean that the various treatments lack value.)

[koganbot]

But anyway, I don't think the difference I think I'm seeing between "normal science" and "normal lots of other things" in regard to whether or not there's majority vs. minority is affected by the fact that there's a time lag between the time when the dominant paradigm has "won" and the time when dissenters are no longer any part of the enterprise.

[koganbot]

The fact that I'm neither a physicist nor a historian of science (nor a historian at all) is a drawback here, so I wonder if there's anyone around who knows the answer, or knows someone who might, but...

I wonder how good an example of a scientific revolution the switchover from the Aristotelian concept of motion to the Newtonian actually is. As I said above, I think it's a good example of incommensurability and of a paradigm switch. One definitely has to go through a paradigm shift to go back and understand Aristotle's concept, and if via a time machine one pulled Aristotle forward to the late 17th century, Aristotle would have to go through a paradigm shift to comprehend Newton's concept.

In Copernicus's time, astronomy was an ongoing concern. Copernicus started with the Aristotelian-Ptolemaic cosmology and modified it to come up with his own cosmology. And of course in doing so he had to assert that the earth moved, rotated every 24 hours around its axis and revolved yearly around the sun. And he had a precedent for not going along with Aristotle's concept of motion, in that Aristotelians of the late middle ages had already criticized Aristotle's concept and were coming up with their own impetus theory in order to understand, say, why a spear traveled in the way that it did when you threw it.

But my question here is: was Aristotle's concept of motion ever in effect as part of an enterprise that people spent much time engaging in? Were there ancient Greeks who thought a lot about Aristotle's concept, who were convinced that motion is a quality and that this was something worth thinking about and elaborating on? Were the medieval Aristotelians who challenged Aristotle's concept and came up with impetus theory part of an ongoing project where people tried to understand motion and put their understandings to use, with Aristotle's concept the prevailing concept? Or were they merely paying attention to Aristotle's idea because it was Aristotle's, and poring over Aristotle was what you did if you were an Aristotelian - but as for their own ideas about motion, were they in effect starting from scratch? I'm so ignorant of the history here - haven't read the Kuhn chapter on impetus theory etc. in several years, and hardly know a thing about either medieval or ancient physics. So as far as I know the answer to my question could be "yes, Aristotelian motion is a big thing both amongst ancient scholars and the medieval guys, and your starting point in thinking about motion would be 'place' as a quality not just as a position." Or it could be "No, Aristotelian motion wasn't really part of an ongoing scientific endeavor." In which case, dealing with motion isn't a case of a paradigm shift in a science as the invention of a scientific paradigm, an invention that gets going in the late middle ages and that a couple centuries later becomes urgent for anyone who takes Copernican cosmology seriously.

As I said, I really don't know.

(Not that how we answer this question would have much bearing on our understanding or evaluation of Kuhn's ideas.)