Kuhn 20: Paradigms, what good are they?
Jul. 7th, 2009 08:33 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
koganbotThe term "paradigm shift" has made it into the common language as "a fundamental change in the way of seeing or doing something." Of course, many people's threshold for what counts as "fundamental" or "change" is really low, and "paradigm shift" is usually what people want the other guy to undergo so that his ideas will come to match the ones we've already got. But that's not the term's fault, that a lot of people become posturing dumbasses when they employ abstract intellectual terminology. A more damaging problem is that people are interested in paradigm shifts but not in paradigms; that is, the idea of breaking through restrictions is appealing, whereas the idea of creating new restrictions and being supported and nurtured by those restrictions is less appealing. But Kuhn came up with the idea of shifts back in the 1940s, along with the idea that the shifts were between incompatible modes of thought. It wasn't until the late '50s, however, that he developed the notion "paradigm," and he did so not only to understand how shifts occurred, but to understand how it was that the hard sciences were so much better at asking and answering questions and at creating fundamental shifts in thought than were the social sciences, which seemed to continually be reverting to square one.
If Kuhn is right, this is what paradigms do for the sciences:
1. Paradigms organize and focus a science's activity by telling the scientist what questions to ask and how to go about answering them. Which is to say that paradigms restrict and specialize the scientist's attention. And as a science develops and undergoes scientific revolutions it breaks into more and more subsciences that are ever more specialized.
2. Paradigms create expectations that are sufficient enough in their precision that anomalies can occur. And when anomalies can't be explained away as equipment failure or scientist error and can't be worked into the paradigm, then they become impetuses for a scientific revolution, i.e., a paradigm shift, and for the ultimate creation of new restrictions and new specialization.
(I recommend that you look back at Kuhn 8½, which contains excerpts from "The Essential Tension," the 1959 article in which Kuhn first uses the word "paradigm.")
If Kuhn is right, this is what paradigms do for the sciences:
1. Paradigms organize and focus a science's activity by telling the scientist what questions to ask and how to go about answering them. Which is to say that paradigms restrict and specialize the scientist's attention. And as a science develops and undergoes scientific revolutions it breaks into more and more subsciences that are ever more specialized.
2. Paradigms create expectations that are sufficient enough in their precision that anomalies can occur. And when anomalies can't be explained away as equipment failure or scientist error and can't be worked into the paradigm, then they become impetuses for a scientific revolution, i.e., a paradigm shift, and for the ultimate creation of new restrictions and new specialization.
(I recommend that you look back at Kuhn 8½, which contains excerpts from "The Essential Tension," the 1959 article in which Kuhn first uses the word "paradigm.")
no subject
Date: 2009-07-08 07:26 am (UTC)no subject
Date: 2009-07-09 05:01 am (UTC)The term "scientific method" means different things to different people, and some people question whether there is such a thing. But I'd say the crucial feature of "scientific method" as most people are likely to conceive it is that you test your hypotheses and theories against the facts, and if your theory doesn't fit the facts, then either you modify it or you come up with a new one. We can add the idea that results need to be repeatable and that when necessary there need to be control groups, etc. etc., but those are basically to validate facts and assure their relevance. The fundamental idea remains that theories are to be grounded in fact; facts have sovereignty.
A number of questions and problems arise, especially these two:
(i) You'd think that the desire to test one's ideas would be a widespread human trait, not limited to physicists, chemists, and biologists. If what it takes to do science is to practice the scientific method, why haven't history, sociology, psychology, political science, and philosophy developed into sciences, and why haven't literary criticism and music criticism developed a scientific component? Are sociologists et al. worse people, essentially unable to attend to facts or notice contradictions? Doesn't seem likely.
(ii) "Scientific method" assumes a stable body of facts - not that all the facts are here at hand, but that with the right equipment and the right experiments and enough ingenuity and effort, we can go out and get whatever facts are there to be gotten, correct our mistakes, make new discoveries. Kuhn challenges this, and I find him 100 percent convincing. If two theories are different in fundamentals, they will give you different facts. Facts are not independent of theory. This is because the different theories (or different paradigms or whatever you want to call them) will differ in what crucial words refer to, will have different concepts, different vocabulary, will be a different nomenclature. Newton didn't just disagree with Aristotle about the laws of motion, he disagreed about what motion was. (This doesn't mean that there weren't good reasons to choose Newton over Aristotle, or that no theory is better than any other, just that "adjusting the theory to the facts" hardly encompasses what's going on in a paradigm shift.)
What I think happens in sociology et al. is that the practitioners never find their way to a basic agreement as to what they're talking about. Maybe that's neither necessary nor desirable. I don't know. I can see no reason for music criticism ever to remotely resemble physics, but I do wish that music critics would spin their wheels less and would figure out how to sustain an intellectual conversation.
(I'll try to find time in the next few days to link to fuller discussions of these issues, and the punctuated equilibria thing.)
no subject
Date: 2009-07-09 04:36 pm (UTC)ETA: Also, I don't quite agree that the scientific method assumes a stable body of facts? Maybe just a matter of semantics but you attempt to make sense of the collection of facts you have through theory and hypothesis, but there's no guarantee that those data belong together or for that matter are actually representative of a consistent reality. I do agree though that theory directs you to collect a certain type of data and that the data you collect is dependent on theory, but I don't see this being as conceptually all that different from the standard idea that all experiments should be hypothesis-driven.
no subject
Date: 2009-07-09 02:22 pm (UTC)Of course the question can arise as to whether Kuhn is right about the existence of "normal science." One can argue that conceptual change is ongoing in a science. But that's a different matter.
no subject
Date: 2009-07-09 04:41 pm (UTC)no subject
Date: 2009-07-09 07:48 pm (UTC)Which is what? And what's unsatisfactory about Kuhn's description? (I'm not a scientist so I wouldn't know.) I take it that "descriptions given" refer to those in "What Are Scientific Revolutions?"
What was on my mind in my post up top was more the constructive aspect of "normal science" (of "paradigms" and "theories" and "hypotheses") - the constructive role of restrictions, that is (in comparison to, say, music criticism, which is too feckless to develop creative restrictions) - so what I wrote in number (2) was truncated, something of an afterthought, my going "Oh yes, and Kuhn would point out that it's the restrictions themselves that generate a lot of data including the anomalies that inspire the overthrow of the paradigm and its replacement by a new one." But my paragraph is very inadequate, and that's my doing, not Kuhn's. The crucial point I left out is that, according to Kuhn, normal science itself is chock full of anomalies - it's the anomalies that provide normal science with its puzzles - and almost never do the anomalies lead to the overthrow of a paradigm. Of course the rare instances when they do are what Kuhn calls scientific revolutions.
In "What Are Scientific Revolutions?" Kuhn was intent on explaining what made a revolutionary shift revolutionary, but he wasn't talking about what inspired the revolutions. Chapters 6, 7, and 8 of Structure are where he sketches his views about when anomalies lead to crisis, and when crises lead to a revolution. My physical therapist tells me that it's OK for me to resume typing with both my hands, so maybe in a while I'll type up an excerpt or two and see if those answer your questions.
In actuality it might be Kuhn's concept of "normal science" that's the greatest challenge to standard notions of "scientific method" (assuming I know what those standard notions are). He disputes the notion that for a scientist a counter-instance invalidates a fundamental theory.* He says that instead, most scientists most of the time have faith that the counterinstance will be explained in a way that gets rid of its being a counter-instance, but that doesn't necessarily mean that scientists will feel any urgency in making such an explanation or ever get around to doing so. Chapter 8, "Response to Crisis": "No one seriously questioned Newtonian theory because of the long-recognized discrepancies between predictions from that theory and both the speed of sound and the motion of Mercury. The first discrepancy was ultimately and quite unexpectedly resolved by experiments on heat undertaken for a very different purpose; the second vanished with the general theory of relativity after a crisis that it had had no role in creating."
By the way, the notion of "punctuated equilibria" is a good example of normal science, Gould and Eldredge (i) taking seriously the fossil record's showing that species spent most of their existence without changing fundamentally (rather than assuming that explaining this away by saying that the fossil record is incomplete [though that still might be a good explanation]) and (ii) coming up with an explanation that doesn't contradict the basic tenets of natural selection. So they take care of an apparent anomaly by explaining how it's not an anomaly, and potentially expand our knowledge by doing so.
*A fundamental theory would be something along the lines of Newton's Laws of Motion, not something like "there are a bunch of planetoids around and beyond Pluto; in fact, Pluto is one of them"; nothing fundamental or conceptual rides on whether that latter statement is true or not.
no subject
Date: 2009-07-09 07:52 pm (UTC)Are you saying that in normal science basic concepts are continually being overthrown? Or are you saying that in Kuhn's descriptions of motion and batteries and black body radiation/quantum discontinuity you don't see a fundamental shift in concepts? (How can you not see it for the Aristotle-Newton example? The example of the quantum discontinuity may be harder to see, since Kuhn doesn't take the idea of the quantum through all its reverberations, and the difference between the term "resonator" and "oscillator" doesn't seem all that big to me, but the point is that it was important to Planck.) Or is it my description of something or other (what I've got up top is barely a description) that you're having trouble with?
no subject
Date: 2009-07-09 08:32 pm (UTC)I think a part of the problem is that I am focusing on what inspires the revolutions because to me, the crux of the scientific method isn't in its scope but in, well, the whole methodology of having a hypothesis, collecting data and revising the hypothesis when the data doesn't fit. In the case of Planck, his paradigm shift was inspired by the phenomenon of blackbody radiation; without the observations there would be no new concept of quanta.
This quote though clarifies a lot of the confusion:
In actuality it might be Kuhn's concept of "normal science" that's the greatest challenge to standard notions of "scientific method" (assuming I know what those standard notions are). He disputes the notion that for a scientist a counter-instance invalidates a fundamental theory.* He says that instead, most scientists most of the time have faith that the counterinstance will be explained in a way that gets rid of its being a counter-instance, but that doesn't necessarily mean that scientists will feel any urgency in making such an explanation or ever get around to doing so.
That seems to suggest that at least going by my definition of scientific method, Kuhn's implying that "normal" science doesn't really follow through on the scientific method because anomalous data doesn't really ever lead to falsification, just revision. Which I will have to think on. I mean there are plenty of concepts that get overthrown in biology*, but whether or not they count as "basic" is another question. They are certainly not paradigm shifts.
* E.g. one of the big "anomalies" in my field is that it was found that the genes essential for surviving a particular stress are not the same genes that undergo changes in expression in response to that stress, which probably sounds very trivial to the layperson but is a huge anomaly that has required a rethinking of our understanding of the role of gene expression.
A side note but I also feel there's a fundamental difference between the Aristotle/Newton example and the classical/quantum mechanics example in that we still consider the classical mechanics paradigm to be a valid approximation for certain frames of reference whereas almost no one does physics based on Aristotle's idea of motion. Does Kuhn ever talk about this? Are there different types of paradigm shifts?
no subject
Date: 2009-07-14 12:21 pm (UTC)(1) I would love it if - for its own sake - you could expand on what you say about genes that are essential to responding to stress versus genes that change their expression in response to stress. (Not that I'll necessarily understand.) Do you ever use your own lj to comment on such stuff?
(2) Yes, Kuhn does talk about whether classic (Newtonian) physics can still be considered a valid subset of modern physics, though he does so in a part of Structure that I haven't arrived at yet in my reread. (And he's talking about it in regard to Newton versus relativity not Newton versus the quantum.) The answer would probably be that "valid approximation" means that it's the approximation that's valid (and the math is a lot simpler, which is why it's still used) but the underlying theory is not. An analogy would be that when people talk about the time of "sunrise" and "sunset" what they're saying is perfectly valid, and the two words are necessary, even if the sun isn't literally rising or setting. (Iirc, Kuhn also mentions that the earthcentric rather than heliocentric system still can play some role in navigation.)
no subject
Date: 2009-07-15 02:53 pm (UTC)no subject
Date: 2009-07-15 04:03 pm (UTC)no subject
Date: 2009-07-15 05:09 pm (UTC)My thought here is that the core question you're asking is this: In "What Are Scientific Revolutions?" Kuhn is differentiating between normal science and revolutionary science on the basis of their results (Boyle's law is neither a conceptual nor a holistic change, whereas the Copernican Revolution, culminating in Newton, is both) but not on the basis of their procedures. But does revolutionary science differ from normal science in its methods and procedures as well as in its results, and does one use "scientific method" more than the other? That is, Kuhn has shown that there are paradigm shifts but has he shown that there is a difference between the conduct of normal and revolutionary science? My quick answer is that Kuhn would say "Yes, there is a difference in methods/procedures, but 'scientific method' itself isn't particularly at issue." (Or anyway, I'm guessing that Kuhn would have said something of the sort.) Which is to say that, as far as "scientific method" goes, when something that some scientists believe contradicts something else they believe, they'll always want to do something about it, whether they're conducting normal science or revolutionary science, and they'll always see the value in coming up with ideas that can be tested and contradicted (though that latter statement opens up a can of worms). But in revolutionary science they'll more likely be making tradeoffs that they wouldn't have had to have made in normal science. That is, they'll adhere to positions that there are good objections to when there are nonetheless other advantages in adhering to it, so an idea can be embraced that reduces incoherence in one area even though it increases incoherence in another, or one can be embraced that increases simplicity but not accuracy, or one that seems to have a wider range of application, even if it doesn't reduce incoherence, etc. And in testing ideas, the more "fundamental" an idea is, the less likely it will be tested by how it stacks up against "observation," and the more likely it will be tested by way of, "Let's try another idea and see if something better happens."