Kuhn 2
As promised, here's some guidance for your reading of "What Are Scientific Revolutions?" a Thomas Kuhn essay that Google Books is giving you in its entirety here, running uninterrupted from p. 13 to p. 32, at least for the time being (though my experience with Google Books is that what pages it gives you can change without warning). First, a brief explanation of why I'm interested in Kuhn in the first place (beyond the fact that I think this sort of thing is fun).
Why Kuhn? Well, for one thing, I find the following passage inspiring, in that it makes me go, "I wish I were more like this fellow":
Those questions troubled me. I could easily believe that Aristotle had stumbled, but not that, on entering physics, he had totally collapsed. Might not the fault be mine rather than Aristotle's, I asked myself. Perhaps his words had not always meant to him and his contemporaries quite what they meant to me and mine.
Also, Kuhn's clarity and his willingness to explain himself and give examples makes him much more potent than, e.g., Nietzsche for the present day. When Nietzsche says "truth is a matter of perspective" or "truth is an illusion" (two incompatible statements, by the way), someone can retort, "So you say, but so what?" whereas when Kuhn says that for Aristotle motion is a change in state while for Newton and Descartes it's a state itself - so for Aristotle, calling motion a "state" would be like calling a circle "square" - this raises inescapable questions about how astronomers and physicists came to choose Newton's concept of motion over Aristotle's and about how we should go about choosing among incompatible premises.*
I've been saying that I consider the whole issue of "relativism," whether you're pro or con, to be the wrong conversation. Studying Kuhn is a good way to test this and explain it, since if you do think there are any "relativist" conclusions to be drawn that are neither banal nor vacuous, Kuhn's specificity can help you not be banal or vacuous yourself in drawing them (not that Kuhn drew any relativist conclusions himself, and not that people who call his work "relativist" know what they mean); and contra that, if I want to say what we should be talking about instead of "relativism," Kuhn is a good starting point.
(Btw, don't be shy about clicking on my tags and scrolling all the way down and then working up if you want to read previous conversations on some subject or other.)
Kuhn is best known for the concepts "paradigm shift" and "incommensurability." I'm starting us with "What Are Scientific Revolutions?" because it does a good job of presenting both. Interestingly enough, the piece uses neither term. "Paradigmatic" occurs once, and "incommensurable" and its variants not at all.
So a good exercise would be to try to figure out from this piece what Kuhn means by "paradigm." As background, I'll say that "paradigm" is just a fancy-shmancy word for "model," and "model" is the concept Kuhn began with, but then, throughout The Structure Of Scientific Revolutions, and without being altogether aware he was doing so, he broadened the term, the result being that he was using "paradigm" in two basic ways (think of this as being similar to how "basketball" is both the name of a ball and the name of a game). "Paradigm was a perfectly good word until I messed it up," he later said, and tended to drop it in favor of other terms. I'm fine with the word as long as we remember to differentiate the meanings: In a paradigm shift, paradigms shift, among other things - old models are replaced by new but also, when an old model hangs on, how it models might no longer be the same. So, after the shift, the model calls forth a different species of behavior among the people who are trying to make their work resemble it than it had before the shift.
In the essay, Kuhn gives three examples of scientific revolutions, though in the first, he discusses what was overthrown (Aristotle's concept of motion) but neither the overthrow nor the new order (Newton's concept). He's assuming we already have a concept similar to the latter - that motion is a state, so an object can be in motion without otherwise changing, and its changing location is merely that: a change in location - and that once he explains Aristotle we'll readily see the difference between Aristotle's concept and ours.
By the way, if you truly understand Aristotle's physics (example one), or understand batteries (example two), or understand how the idea of the quantum developed (example three), you're doing better than I am, but I don't think you need to in order to understand how the old differs from the new and why Kuhn considers the changes revolutionary.
Here are a couple of quotes from the next thing I'm going to have you read, "Planck Revisited" (both from p. 363):
paradigms were the concrete examples needed - since definition in words was impossible - to acquire the language of the older mode
Huh? What? Why does he think a definition in words was impossible? It's not obvious to me that such definitions are impossible, though maybe he's differentiating more sharply than I normally do between definitions and explanations, which he himself gives plenty of, quite successfully, in words.
Boltzmann's probabilistic derivation of the entropy of a gas is of particular interest, for it illustrates the problem to which the concept of paradigm was a response. The derivation was not reduced to rules but instead served as a model to be applied by means of analogy.
"Not reduced to rules" is not a phrase that explains itself. What's the distinction he's trying to draw between following a rule and using a model, between rules and analogies? I think there's probably a good one, the broader point being that science is an analogic rather than a methodical activity - hence there's no scientific method unless you define "scientific method" real loosely - but simply saying there's a distinction leaves a question unanswered: Can how you follow a rule be specified in words any more than how you apply a model? Imitating Wittgenstein, when I say that I can make a rule that instructs everyone to look wherever I point with my index finger, one can always then ask, "What's the rule that says I'm to look forward from your finger and not back along your elbow when you point, or in a line ahead rather than perpendicular to the finger, and what would be the rule that explains how one 'looks forward,' and what would be the rule that explains how I follow that rule, and on into an infinite regress?" Wittgenstein uses such an example not to say that there is an infinite regress but, to the contrary, to show us that we don't need underlying explanations or rules unless there is an actual lived problem that a further rule or explanation can clear up; also, that any rule or explanation that clears up a problem is adequate in itself, without needing the support of some deep, ultimate rule or explanation. But my purpose here is to show that when a child learns what pointing is, he's doing the same type of thing that a scientist like Planck is doing who models a new derivation on a previous one (e.g., Boltzmann's): the child is using an instance of pointing as a paradigm for potential future acts of pointing. Following a rule seems as analogic as using a model. (As I said, I think Kuhn's distinction between analogies and rules is on the right track. But the distinction is fuzzy and undeveloped, and it probably will never not be fuzzy - which may or may not turn out to be fine, the fuzziness.)
That's it for now. Incommensurability to come in a future post. But a question to keep asking yourself, for the time being, is "What is a model? What is going on when people model their behavior on previous behavior, their ideas on previous ideas, something such as pointing on previous acts of pointing, and so on?"
Well, one further thing. Give some thought to how we're going to alert each other when someone has made a new post on a thread, or that a thread is still active, given that lj doesn't have a "new answers" page and given lj's irritating feature of nesting subthreads.
*Not to denigrate Nietzsche, who was working more than a century ago and in isolation without good models for how to formulate his own doubts. Also, he was over-attached to the ideas he was challenging, since overestimating the importance of the ideas allowed him to overestimate the importance of the challenge.
Why Kuhn? Well, for one thing, I find the following passage inspiring, in that it makes me go, "I wish I were more like this fellow":
Those questions troubled me. I could easily believe that Aristotle had stumbled, but not that, on entering physics, he had totally collapsed. Might not the fault be mine rather than Aristotle's, I asked myself. Perhaps his words had not always meant to him and his contemporaries quite what they meant to me and mine.
Also, Kuhn's clarity and his willingness to explain himself and give examples makes him much more potent than, e.g., Nietzsche for the present day. When Nietzsche says "truth is a matter of perspective" or "truth is an illusion" (two incompatible statements, by the way), someone can retort, "So you say, but so what?" whereas when Kuhn says that for Aristotle motion is a change in state while for Newton and Descartes it's a state itself - so for Aristotle, calling motion a "state" would be like calling a circle "square" - this raises inescapable questions about how astronomers and physicists came to choose Newton's concept of motion over Aristotle's and about how we should go about choosing among incompatible premises.*
I've been saying that I consider the whole issue of "relativism," whether you're pro or con, to be the wrong conversation. Studying Kuhn is a good way to test this and explain it, since if you do think there are any "relativist" conclusions to be drawn that are neither banal nor vacuous, Kuhn's specificity can help you not be banal or vacuous yourself in drawing them (not that Kuhn drew any relativist conclusions himself, and not that people who call his work "relativist" know what they mean); and contra that, if I want to say what we should be talking about instead of "relativism," Kuhn is a good starting point.
(Btw, don't be shy about clicking on my tags and scrolling all the way down and then working up if you want to read previous conversations on some subject or other.)
Kuhn is best known for the concepts "paradigm shift" and "incommensurability." I'm starting us with "What Are Scientific Revolutions?" because it does a good job of presenting both. Interestingly enough, the piece uses neither term. "Paradigmatic" occurs once, and "incommensurable" and its variants not at all.
So a good exercise would be to try to figure out from this piece what Kuhn means by "paradigm." As background, I'll say that "paradigm" is just a fancy-shmancy word for "model," and "model" is the concept Kuhn began with, but then, throughout The Structure Of Scientific Revolutions, and without being altogether aware he was doing so, he broadened the term, the result being that he was using "paradigm" in two basic ways (think of this as being similar to how "basketball" is both the name of a ball and the name of a game). "Paradigm was a perfectly good word until I messed it up," he later said, and tended to drop it in favor of other terms. I'm fine with the word as long as we remember to differentiate the meanings: In a paradigm shift, paradigms shift, among other things - old models are replaced by new but also, when an old model hangs on, how it models might no longer be the same. So, after the shift, the model calls forth a different species of behavior among the people who are trying to make their work resemble it than it had before the shift.
In the essay, Kuhn gives three examples of scientific revolutions, though in the first, he discusses what was overthrown (Aristotle's concept of motion) but neither the overthrow nor the new order (Newton's concept). He's assuming we already have a concept similar to the latter - that motion is a state, so an object can be in motion without otherwise changing, and its changing location is merely that: a change in location - and that once he explains Aristotle we'll readily see the difference between Aristotle's concept and ours.
By the way, if you truly understand Aristotle's physics (example one), or understand batteries (example two), or understand how the idea of the quantum developed (example three), you're doing better than I am, but I don't think you need to in order to understand how the old differs from the new and why Kuhn considers the changes revolutionary.
Here are a couple of quotes from the next thing I'm going to have you read, "Planck Revisited" (both from p. 363):
paradigms were the concrete examples needed - since definition in words was impossible - to acquire the language of the older mode
Huh? What? Why does he think a definition in words was impossible? It's not obvious to me that such definitions are impossible, though maybe he's differentiating more sharply than I normally do between definitions and explanations, which he himself gives plenty of, quite successfully, in words.
Boltzmann's probabilistic derivation of the entropy of a gas is of particular interest, for it illustrates the problem to which the concept of paradigm was a response. The derivation was not reduced to rules but instead served as a model to be applied by means of analogy.
"Not reduced to rules" is not a phrase that explains itself. What's the distinction he's trying to draw between following a rule and using a model, between rules and analogies? I think there's probably a good one, the broader point being that science is an analogic rather than a methodical activity - hence there's no scientific method unless you define "scientific method" real loosely - but simply saying there's a distinction leaves a question unanswered: Can how you follow a rule be specified in words any more than how you apply a model? Imitating Wittgenstein, when I say that I can make a rule that instructs everyone to look wherever I point with my index finger, one can always then ask, "What's the rule that says I'm to look forward from your finger and not back along your elbow when you point, or in a line ahead rather than perpendicular to the finger, and what would be the rule that explains how one 'looks forward,' and what would be the rule that explains how I follow that rule, and on into an infinite regress?" Wittgenstein uses such an example not to say that there is an infinite regress but, to the contrary, to show us that we don't need underlying explanations or rules unless there is an actual lived problem that a further rule or explanation can clear up; also, that any rule or explanation that clears up a problem is adequate in itself, without needing the support of some deep, ultimate rule or explanation. But my purpose here is to show that when a child learns what pointing is, he's doing the same type of thing that a scientist like Planck is doing who models a new derivation on a previous one (e.g., Boltzmann's): the child is using an instance of pointing as a paradigm for potential future acts of pointing. Following a rule seems as analogic as using a model. (As I said, I think Kuhn's distinction between analogies and rules is on the right track. But the distinction is fuzzy and undeveloped, and it probably will never not be fuzzy - which may or may not turn out to be fine, the fuzziness.)
That's it for now. Incommensurability to come in a future post. But a question to keep asking yourself, for the time being, is "What is a model? What is going on when people model their behavior on previous behavior, their ideas on previous ideas, something such as pointing on previous acts of pointing, and so on?"
Well, one further thing. Give some thought to how we're going to alert each other when someone has made a new post on a thread, or that a thread is still active, given that lj doesn't have a "new answers" page and given lj's irritating feature of nesting subthreads.
*Not to denigrate Nietzsche, who was working more than a century ago and in isolation without good models for how to formulate his own doubts. Also, he was over-attached to the ideas he was challenging, since overestimating the importance of the ideas allowed him to overestimate the importance of the challenge.