Guide to Thomas Kuhn’s The Structure of Scientific Revolutions

Malcolm R. Forster: March 19, 1998

Note:
I have tried to let Kuhn speak for himself whenever possible. The make is easier to distinguish the quotes from the paraphrases, I have written the quotes in boldface. All references are to the 1970 edition of The Structure of Scientific Revolutions.

1. A Paradigm is ...?
Kuhn baptizes his famous notion of a scientific "paradigm" as originating from the "great works" of science, like Copernicus’s De Revolutionibus or Newton’s Principia. These great works became paradigms because they were "sufficiently unprecedented to attract an enduring group of adherents away from competing modes of scientific activity," and "sufficiently open-ended to leave all sorts of problems for the redefined group of practitioners to resolve." (p.10) The activity spurred by such great books goes by the name of "normal science." "There are ... only three normal foci for factual scientific investigation." (p.25.)

1. Determination of significant fact: Attempts to increase the accuracy and scope with which facts like the specific gravities and compressibilities of material, electrical conductivities and contact potentials occupy a significant fraction of the literature of experimental and observational science. This kind of activity is largely independent of the reigning paradigm - the Newtonian measurement of specific gravities, for example, was unchanged by the advent of Einstein's theory of mechanics.

2. Matches of fact with theory: (p.26) "A second usual but smaller class of factual determinations is directed to those facts that, though often without much intrinsic interest, can be compared directly with predictions from the paradigm theory." In support of the claim that these cases are rarer, Kuhn mentions that (p.26) "No more than three such areas are even yet accessible to Einstein's general theory of relativity."

3. The articulation of theory, which is "... empirical work undertaken to articulate the paradigm theory ..." This is broken up into three sub-categories. (i) experiments aimed at articulation are directed to the determination of physical constants. (ii) experiments aimed at quantitative laws: E.g. Boyle's Law relating gas pressure to volume. (iii) experiments to choose among alternative ways of applying the paradigm to new areas of interest. (p.29)

     This is one of the many places in which Kuhn might do well to make a distinction between theory and models. Let us stipulate that models are the sets of equations derived from theory plus auxiliary assumptions such that estimates of all adjustable parameters may be obtained from extant data (Kuhn uses the term ‘model’ in a different sense). Thus, the set of Newton’s equations of motion is not a model because it is impossible to make a precise quantitative comparison with data without (approximate) solutions of those equations. When Kuhn talks about the articulation of theory as a part of normal science, he is really referring to the development of models.
     While normal science is a highly determined kind of activity, Kuhn is quick to deny that he is slipping back to naive inductivism: (p.42) In a footnote, Kuhn acknowledges the work of M. Polanyi, Personal Knowledge, Chicago, 1958, in which scientists’ success is said to depend on tacit knowledge acquired through practice, which cannot be articulated explicitly. The point, expanded in the postscript, is that scientists gain tacit knowledge of a theory through working through textbook or laboratory examples (called exemplars).

2. Normal Science Does Not Aim at Novelty: Contrary to a popular picture of science, Kuhn insists that (p.52) "Normal science does not aim at novelties of fact or theory and, when successful, finds none." This striking view challenges the critical rationalism of someone like Popper, who sees the heart of scientific rationality in the constant critical scrutiny of accepted scientific belief. Kuhn is concerned to dispel the idea that the common occurrence of scientific discoveries disproves his thesis. For if normal science aims at discovery, and discoveries are novel, then normal science aims at novelty. Kuhn claims that discoveries are always accompanied by changes in the prevailing paradigm. If he is right, then the existence of scientific discovery does not show that normal science aims at novelty, but only that novelty signals the end of normal science. Kuhn therefore views such discoveries as ‘small’ revolutions.
     In summary, Kuhn’s argument is something like this:
1. All novelties of fact (discoveries) or theory lead to the end of normal science.
2. Normal science does not aim at its own demise.
Therefore, normal science does not aim novelties of fact or theory and, when successful, finds none.

3. Discoveries are Rare Because Expectations Obscure our Vision: The fact the normal science does not aim at novelty, as Kuhn has argued, cries out for explanation. Briefly, Kuhn’s response is that scientists are entrenched within a certain way of seeing things, and this clouds their vision (they tend to see what they expect to see). As an instance of what Kuhn thinks is a general psychological phenomenon, he cites a study by J. S. Bruner and Leo Postman, "On the Perception of Incongruity: A Paradigm," Journal of Personality XVIII (1949) 206-23. In this psychological experiment, subjects are shown ordinary playing cards mixed up with some anomalous cards, like a black four of hearts. Roughly speaking, the results show that subjects initially see what they expect to see (either the four of spades, or the four of hearts).
     In sum, Kuhn seeks to explain the difficulty of discovery as an instance of the general psychological fact that our expectations cloud our perception of the world.

4. No Paradigm Change without Crisis: Nevertheless, a paradigm (though resisting change) is playing an essential role in allowing a scientist to recognize something as anomalous, as contrary to expectation, and this is an important precondition for discovery (p.65). However, the process of improving fit between fact and theory is a part of normal science, so an anomaly, a failing of expectations, presents just another puzzle to be resolved by the construction of improved models. That is the standard fare of normal science. The point is that an anomaly is not by itself sufficient for paradigm change (that is the falsificationist’s folly).
     For example, Ptolemy’s system of astronomy certainly faced discrepancies, but it was only when those discrepancies built up to crisis point that the conditions were ripe for change (p.68): "Given a particular discrepancy, astronomers were invariably able to eliminate it by making some particular adjustment in Ptolemy’s system of compounded circles. But... astronomy’s complexity was increasing far more rapidly than its accuracy and that a discrepancy corrected in one place was likely to show up in another."  Notice that Kuhn mentions complexity as requisite condition for paradigm change in this example. It is a necessary part of what defines a crisis in normal science. For, a sufficient number of compounded circles would provide perfect fit with the data at any one time. That Ptolemaic model would provide no discrepancies with existing data. But the crisis is revealed by the way it changes over time, for as Kuhn puts it, the "astronomy’s complexity was increasing far more rapidly than its accuracy and that a discrepancy corrected in one place was likely to show up in another."
     On the other hand, Kuhn presents the Copernican revolution as an example which the crisis in the reigning Ptolemaic paradigm was almost the only reason for the change  (pp.75-6): "Copernicus’ more elaborate proposal was neither simpler nor more accurate than Ptolemy’s system. Available observational tests, as we shall see more clearly below, provided no basis of a choice between them."

5. In Normal Science the Theory is Not Questioned: The fact that anomalies are the driving force behind theory change does not mean that scientists are following a falsificationists’ methodology. "Though they may begin to loose faith and then to consider alternatives, they do not renounce the paradigm that has led them into crisis. They do not, that is, treat anomalies as counterinstances, though in the vocabulary of philosophy of science that is what they are."  (p.77) Rather: "The decision to reject one paradigm is always simultaneously the decision to accept another, and the judgment leading to that decision involves the comparison of both paradigms with nature and with each other." (p.79): "To reject one paradigm without simultaneously substituting another is to reject science itself." For Kuhn, anomalies are only treated as counterinstances by supporters of the competing paradigm.
     For example, "Copernicus saw as counterinstances what most of Ptolemy’s other successors had seen as puzzles in the match between observation and theory."  (p.79) As Copernicus complained (quoted from Kuhn (1957), p.138), in his day astronomers were "so inconsistent in these [astronomical] investigations ... that they cannot even explain or observe the constant length of the seasonal year." "With them," Copernicus continues, "it is as though an artist were to gather the hands, feet, head and other members for his images from diverse models, each part excellently drawn, but not related to a single body, and since they in no way match each other, the result would be monster rather than man." Again, we may note that Copernicus is not complaining about discrepancies per se, but about the fact that discrepancies in Ptolemy’s system are only removed ad hoc adjustments to the model.
     From within a paradigm, from the viewpoint of normal science, anomalies are not seen as testing the theory. Yet Kuhn concedes (p.80) that "Normal science does and must continually strive to bring theory and fact into closer agreement, and that activity can easily be seen as testing or as a search for confirmation or falsification." "But science students accept theories on the authority of teacher and text, not because of evidence." (p.80) Hence the standards of critical rationality ascribed by Popper are not present.
     There is a need to clarify what Kuhn is saying here. While uncritically taking the background theory for granted, normal science does subject the models derived from the theory to severe critical scrutiny. None of Kuhn’s examples undermine that idea that models are falsified (although there are independent reasons for thinking that such a view is too simple). So normal science strives to bring theory and fact into closer agreement by calling its models into question without ever criticizing the background theory itself. Kuhn does make use of a distinction between "approaches to problems" (models?) and the paradigm later in the book. Thus, while scientists will  (p.144) "try out and reject a number of alternative approaches, rejecting those that fail to yield the desired result, he is not testing the paradigm when he does so."
     The same point that is now often used in criticism of Popper’s falsificationism: Theories are never tested in isolation. And when they are tested, as in the Leverrier-Adams discovery of Neptune, it is the auxiliary assumptions (the assumption that Uranus was the outermost planet) that was criticized, and not Newton’s laws of motion.

6. New Paradigms Place New Relations Amongst the Data: When normal puzzle solving fails to resolve anomalies, crisis looms, and the nature of scientific activity gradually changes. (p.91) "The proliferation of competing articulations, the willingness to try anything, the expression of explicit discontent, the recourse to philosophy and to debate over fundamentals, all these are symptoms of a transition from normal to extraordinary research."
When the transition is complete, the profession will have changed its view of the field, its methods, and its goals. One perceptive historian, viewing a classic case of a science’s reorientation by paradigm change, recently described it as "picking up the other end of the stick," a process that involves "handling the same bundle of data as before, but placing them in a new system of relations with one another by giving them a different framework." [Herbert Butterfield, The Origins of Modern Science, 1300-1800 (London, 1949), pp.1-7] Others who have noted this aspect of scientific advance have emphasized its similarity to a change in visual gestalt: the marks on paper that were first seen as a bird are now seen as an antelope, or vice versa. (p.85) 
     Butterfield’s description of theory change is an important clarification of Kuhn’s poetic reference seeing a duck and then seeing a rabbit. For there is a danger that we might take Kuhn’s duck-rabbit analogy too seriously, and view theory change as literally involving a change in perception, and therefore a change in the data themselves, rather than just a change in the relations amongst the data.
     Another noteworthy feature of the above quote is the idea that the goals of science change when the paradigm changes. Presumably, this is because the goal of science is problem-solving, and this changes as the problems change. Yet one might also define problem-solving in a paradigm independent way (e.g., as the goal of finding predictively accurate models), in which case the goal would not change. But Kuhn does not appear to subscribe to such a view of science.

7. Science is Non-Cumulative Because Terms Change their Meanings: In the meantime, Kuhn mounts an attack on the common idea that scientific knowledge is accumulative. He begins with the often quoted idea that most often, "the new paradigm, or a sufficient hint to permit later articulation, emerges all at once, sometimes in the middle of the night, in the mind of a man deeply immersed in crisis." (p.90) If paradigms change in such a sudden way, how can they simply built on prior knowledge? For Kuhn (p.92), this indicates that scientific revolutions are "non-cumulative developmental episodes in which an older paradigm is replaced in whole or in part by an incompatible new one." Moreoever, paradigm choice in science are "Like the choice between competing political institutions, that between competing paradigms proves to be a choice between incompatible modes of community life." And the choice is a choice in community alliance, for "As in political revolutions, so in paradigm choice-there is no standard higher than the assent of the relevant community."  (p.94) Kuhn, of course, is acutely aware of the controversial nature of his denying standards higher than the assent of the relevant community. Traditional philosophy of science is premised on the claim that the logical structure of scientific theories provides intrinsic reasons for theory change independently of the prevailing social conditions. Kuhn is quick to confirm the impression that he rejects this working assumption. (p.95)

8. Historical Examples Show that Science is Non-Cumulative: Kuhn’s hunch is that scientific change brings about a change in the entities that are taken to be primitive and unexplained. For example, Aristotelians said that a stone fell because of its ‘nature’ drove it toward the center of the universe. Afterwards the normal seventeenth-century tradition of scientific practice insisted that "the entire flux of sensory appearances, including color, taste, and even weight, was to be explained in terms of the size, shape, position, and motions the elementary corpuscles of base matter." (p.104) The attribution of other qualities to the elementary atoms was a resort to the occult and therefore out of bounds for science. Famously, Molière ridiculed the doctor who explained opium’s efficacy as a soporific by attributing to it a dormitive potency. (p.104) Kuhn sees this not as a criticism of postulating mystical entities per se but of postulating an entity not accepted as primitive at the time. In that vein, Kuhn remarks that "During the last half of the seventeenth century many scientists preferred to say that the round shape of the opium particles enabled them to sooth the nerves about which they moved."

9. Incommensurability: The lesson that Kuhn draws from these examples is that:
"... the reception of a new paradigm often necessitates a redefinition of the corresponding science. Some old problems may be relegated to another science or declared entirely "unscientific." [e.g. alchemy] Others that were previously non-existent or trivial may, with a new paradigm, become the very archetypes of significant scientific achievement. [e.g., tidology, the study of the tides] The normal-scientific tradition that emerges from a scientific revolution is not only incompatible but often actually incommensurable with that which has gone before." (p.103)
     As far as the main conclusion is concerned, Kuhn’s argument is convincing. There is no doubt that the world view that emerges from a scientific revolution may be incommensurable in the weak sense described above - that new terms are not straightforwardly translatable, and the new paradigm leads to at least some incompatible predictions.
However, there is a concomitant claim that should not go unchallenged. Are there really no objective standards on the acceptance of primitive entities, as Kuhn insinuates? Surely, it is reasonable to demand that postulated entities can at least be measured. Isn’t it conceivable that Molière’s disdain for the term ‘dormitive potency’ arose from the fact that the label adds nothing but a new way of saying "soporific"?

10. Development of Science’s Problems and Standards is Also Non-Cumulative: Kuhn has been arguing against a straw man thus far, in that very few philosophers thought that theory change was cumulative in any naive sense. But far more did, and still do, hold dearly to the view that the standards by which scientific theories are judged (goodness-of-fit, simplicity, unification, and so on) are constant across theory change. Kuhn, on the other hand, develops a line of argument that is more or less an inference to the best explanation - if we suppose that problems and standards change across paradigm shifts, then we may explain why there is so often a breakdown in communication. In the first place, there will be disagreements about the what counts as a solution to a problem because the problems are different, or at least redescribed. (pp.109-10) Furthermore, they are unlikely to agree on what the important problems are. (p.110) However, Kuhn has no argument to show that models from different paradigms cannot be compared on grounds of fit, simplicity, or unification, at least in principle.

11. Paradigms Transform Scientists’ View of the World: Psychological experiments show that expectations cloud and obscure perception (§3), and a paradigm produces expectations, and so a paradigm may skew observation:
     The man who first saw the exterior of the box from above later sees its interior from below. Transformations like these, though usually more gradual and almost always irreversible are common concomitants of scientific training. Looking at a contour map, the student sees lines on paper, the cartographer a picture of a terrain. Looking at a bubble-chamber photograph, the student sees confused and broken lines, the physicist a record of familiar subnuclear events. (p.111)
     In one psychological experiment, volunteers fitted with inverting lenses report that everything appears up side down at first, but they grow accustomed to the inversion. They learn to accurately predict where their hand should move in order to intercept an object, say, and report the experience of seeing everything the right side up. (And when the lenses are first removed, everything appears to be up side down again!) According to Kuhn, (p.112) "literally as well as metaphorically, the man accustomed to inverting lenses has undergone a revolutionary transformation of vision." Yet Kuhn is quick to acknowledge that these known psychological facts about vision do not tell the affect, if any, of scientific training on vision.  (p.113) And prima facie, it is somewhat implausible to naively extend the idea to, say, the convert to a new paradigm. For example, "Looking at the moon, the convert to Copernicanism does not say, ‘I used to see a planet, but now I see a satellite.’ That locution would imply a sense in which the Ptolemaic system had once been correct. Instead, a convert to the new astronomy says, ‘I once saw the moon to be (or saw the moon as) a planet, but I was mistaken’."  (p.115) At the same time, there are clear analogies to the psychological phenomenon of "seeing what one expects to see" as in the case of seeing a black three of hearts as a three of spades. For instance:
     "Sir William Herschel’s discovery of Uranus provides a first example and one that closely parallels the anomalous card experiment. On a least seventeen different occasions between 1690 and 1781, a number of astronomers, including several of Europe’s most eminent observers, had seen a star in positions that we now suppose must have been occupied at the time by Uranus... Herschel, when he first observed the same object twelve years later, did so with a much improved telescope of own manufacture. As a result, he was able to notice an apparent disk-size that was at least unusual for stars. Something was awry, and he therefore postponed identification pending further scrutiny. That scrutiny disclosed Uranus’ motion among the stars, and Herschel therefore announced that he had seen a new comet! Only several months later, after fruitless attempts to fit the observed motion to a cometary orbit, did Lexell suggest that the orbit was probably planetary." (p.115)  
     The question is whether there is evidence for the more radical conclusion that new paradigms can change the way we see the world, as opposed to changing the way we interpret what we see. Or maybe we should say that new paradigms actually change the world in which scientists work?!

124. Paradigms Change the World in which Scientists Work: Kuhn is testing the waters, wondering how far he should go. Should he really irritate the philosophers by claiming that paradigms change the world itself?
     Using traditional instruments, some as simple as a piece of thread, late sixteenth-century astronomers repeatedly discovered that comets wandered at will through the space previously reserved for the immutable planets and stars. The very ease and rapidly with which astronomers saw new things when looking at old objects with old instruments may make us wish to say that, after Copernicus, astronomers lived in a different world. In any case, their research responded as though that were the case.  (pp.116-7)
     Cautiously, Kuhn pushes the thesis a little further:
"At the very least, as a result of discovering oxygen, Lavoisier saw nature differently. And in the absence of some recourse to that hypothetical fixed nature that he "saw differently," the principle of economy will urge us to say that after discovering oxygen Lavoisier worked in a different world."(p.118)
     So far Kuhn has provided no detailed argumentation to support his analysis. Those details are really quite interesting in their own right, irrespective of whether they actually succeed in supporting the stronger conclusion. By way of elaboration, Kuhn contrasts the Aristotelian and Galilean view of a simple pendulum:
     "To the Aristotelians, who believed that a heavy body is moved by its own nature from a higher position to a state of natural rest at a lower one, the swinging body was simply falling with difficulty. Constrained by the chain, it could achieve rest at its low point only after tortuous motion and a considerable time. Galileo, on the other hand, looking at the swinging body, saw a pendulum, a body that almost succeeded in repeating the same motion over and over again ad infinitum. And having seen that much, Galileo observed other properties of the pendulum as well and constructed many of the most significant and original parts of this new dynamics around them. From the properties of the pendulum, for example, Galileo derived his only full and sound arguments for the independence of weight and rate of fall, as well as for the relationship between vertical height and terminal velocity of motions down inclined planes. [Galileo Galilei, Dialogues concerning Two New Sciences, trans. H. Crew and A. de Salvio (Evanston, Ill., 1946), pp.80-81, 162-66.] All of these natural phenomena he saw differently from the way they had been seen before."  (pp.118-9)
     But the conclusion drawn from this example is that "until that scholastic paradigm was invented, there were no pendulums, but only swinging stones, for the scientists to see. Pendulums were brought into existence by something very like a paradigm-induced gestalt switch." (p.120) This is a radical conclusion.

13. A Confusion between Seeing and Seeing As? At first glance, Kuhn’s extreme position appears to arise from a simple confusion between ‘seeing’ and ‘seeing as’. Isn’t the situation simply this: Both Aristotle and Galileo saw the thing that we now describe as a pendulum, even though Aristotle did not describe it in those terms. This is the same sense in which we say that someone totally ignorant of matters astronomical actually succeeds in seeing the planet Venus when they point to the evening star. On the other hand, Aristotle did not see pendulums as pendulums, in the same sense that our astronomical ignoramus does not see the evening star as the planet Venus. On this alternate view, our ignoramus sees the same thing as the expert astronomer - the difference is that the expert interprets what she sees as the planet Venus. But if this is the only alternative, then Kuhn thinks that we should continue to say the post-paradigm scientist literally lives in a transformed world:
     "What occurs during a scientific revolution is not fully reducible to a reinterpretation of individual and stable data. In the first place, the data are not unequivocally stable. A pendulum is not a falling stone, nor is oxygen dephlogisticated air. Consequently, the data that scientists collect from these diverse objects are, as we shall shortly see, themselves different. More important, the process by which either the individual or the community makes the transition from constrained fall to the pendulum or from dephlogisticated air to oxygen is not one that resembles interpretation. How could it do so in the absence of fixed data for the scientist to interpret? Rather than being an interpreter, the scientist who embraces a new paradigm is like the man wearing inverting lenses." (pp.121-22)

14. Radical Incommensurability: When Kuhn asserts in the last passage that "the data are not unequivocally stable," he is stepping beyond the weak version of his incommensurability thesis. For if the data are incommensurable (because "A pendulum is not a falling stone, nor is oxygen dephlogisticated air"), then it would be false to claim that the two paradigms make incompatible predictions-the one is making predictions about a falling stone, and the other about a pendulum. The strong incommensurability thesis says everything the weak thesis says plus the claim that the empirical data for a given theory cannot be translated in a way that are neutral between competing theories. Of course, it is not hard to resist this stronger thesis (and so we should): The Aristotelian and Galilean predictions are about the same thing (an stone swinging on a rigid rod), even though they are described and even seen as differing in some respects (as a falling stone in the first instance, or as a pendulum in the second). There is no reason to deny that the predictions of the two paradigms are commensurable. Kuhn’s only argument to the contrary conclusion appears to revert back to his "gestalt shift" picture of paradigm change:
     "Scientists then often speak of the "scales falling from the eyes" or the "lightning flash" that "inundates" a previously obscure puzzle, enabling its components to be seen in a new way that for the first time permits its solution. On other occasions the relevant illumination comes in sleep. Nor ordinary sense of the term ‘interpretation’ fits these flashes of intuition through which a new paradigm is born. Though such intuitions depend upon the experience, both anomalous and congruent, gained with the old paradigm, they are not logically or piecemeal linked to particular items of that experience as an interpretation would be. Instead, they gather up large portions of that experience and transform them to the rather different bundle of experience that will thereafter be linked piecemeal to the new paradigm but not to the old." (pp.122-23) 
     The last sentence in the quotation is another reference to Butterfield’s description of theory change (§7, above). There is a sense in which this description strengthens the case against interpretation, since to exploit new relations is not to reinterpret old relations. However, it offers no support for a stronger incommensurability thesis. For the relations amongst data can change without the data themselves changing.

15. Need We Be So Concerned about Immediate Experience? Suppose we grant that "the immediate content of Galileo’s experience with falling stones was not what Aristotle’s had been." (p.125) But is experience at this personal level crucial to a deeper understanding of scientific change? In response to this objection, Kuhn emphatically denies that operations and measurements are paradigm independent, "the measurements to be performed on a pendulum are not the ones relevant to a case of constrained fall." (p.126) While Kuhn is undoubtedly right on this point, it is not enough to dismiss the objection. As philosophers of science, maybe we should discuss the concrete operations and measurements that the scientist performs in her laboratory even though they are paradigm dependent? Their paradigm dependence does not exclude them from being predicted by a competing paradigm. For example, it is perfectly sensible to ask whether Kepler’s laws predict the 3-D positions of Mars as inferred from the Copernican model better than the Copernican model best fitted to past data. There is nothing absurd about this, as Kuhn believes.

16. There Is No Such Thing as a Neutral Observation Language: Kuhn then goes on to argue against a second proposal to bypass consideration of "immediate experience": "[The analysis] might, for example, be conducted in terms of some neutral observation-language, perhaps one designed to conform to the retinal imprints that mediate what the scientists see."  (p.125) Kuhn’s reaction to the neutral observation language proposal is that retinal impressions do not map one-to-one onto perceptual experiences: (pp.126-27) "The duck-rabbit shows that two men with the same retinal impressions can see different things; the inverting lenses show that two men with different retinal impressions can see the same thing." However, it is not immediately clear how this fact undermines the proposal. Maybe Kuhn thinks that the absence of such a mapping means that a description of retinal impressions could not be unambiguously understood by all sides. But I think this is false. A theory-neutral observation language is not needed to compare the fit of models from different paradigms. The merit of the model is measured by how well it allows past data to predict future data, and the success of models from different paradigms can be compared against a common data set even when the data in question is theory-laden.

17. Philosophers have Disguised the Confirmation Situation: In considering traditional philosophical accounts of scientific confirmation, Kuhn first complains that philosophers have misrepresented confirmation as a relation between theory and evidence. Kuhn disagrees on two counts: First the question should be one about scientific communities rather than theories: (p.144) "What causes the group to abandon one tradition of normal research in favor of another?" Second, (p.145) "testing occurs as part of the competition between two rival paradigms for the allegiance of the scientific community." Thus, confirmation or verification is not a relation between a theory and evidence, but a process of selection from amongst rival candidates: (p.146) "Verification is like natural selection: it picks out the most viable among the actual alternatives in a particular historical situation." Kuhn’s point about the comparative nature of confirmation is a good one.
     Other philosophers of science, like Popper, have denied the existence of any verification procedures at all. Theories are never justified (because of problems with induction), but they may be disproven. Popper’s philosophy is that of falsificationism: Theories should be severely criticized in an attempt to falsify them. Survivors of such critical scrutiny are rationally acceptable, and are said to be corroborated.
     "... the role thus attributed to falsification is much like the one this essay assigns to anomalous experiences, i.e. to experiences that, by evoking crisis, prepare the way for a new theory. Nevertheless, anomalous experiences may not be identified with falsifying ones. Indeed, I doubt that the latter exist. As has repeatedly been emphasized before, no theory ever solves all the puzzles with which it is confronted at an given time; nor are the solution already achieved often perfect. On the contrary, it is just the incompleteness and imperfection of the existing data-theory fit that, at any time, define many of the puzzles that characterize normal science. If any and every failure to fit were ground for theory rejection, all theories ought to be rejected at all times. On the other hand, if only severe failure to fit justifies theory rejection, then the Popperians will require some criterion of "improbability" or of "degree of falsification." In developing one they will almost certainly encounter the same network of difficulties that has haunted the advocates of the various probabilistic verification theories." (pp.146-47)

18. Incommensurability is a Cause of Communication Breakdowns: Almost by way of review, Kuhn recalls that (p.148) "We have already seen several reasons why the proponents of competing paradigms must fail to make complete contact with each other’s viewpoints." The three reasons for communication breakdown are all "fundamental aspects of incommensurability." First, not all past successes are recognized as relevant problems in the new paradigm. For example, Cartesian vortex theory had a ready explanation of why all the planets revolve around the sun in the same direction. The problem was seen as extraneous to mechanics by the Newtonians who considered it as a question about the origin of the solar system. This is an example of what some refer to as Kuhn loss (I heard the example and the term from Alan Musgrave):
     "In the first place, the proponents of competing paradigms will often disagree about the list of problems that any candidate for paradigm must resolve. ... Lavoisier’s chemical theory inhibited chemists from asking why the metals were so much alike, a question that phlogistic chemistry had both asked and answered. The transition to Lavoisier’s paradigm had, like the transition to Newton’s meant a loss not only of a permissible question but of an achieved solution." (p.148)
     The second fundamental aspect of incommensurability might be called meaning variance, and is often simply equated with incommensurability:
     "Communication across the revolutionary divide is inevitably partial. Consider, for another example, the men who called Copernicus mad because he proclaimed that the earth moved. They were not either just wrong or quite wrong. Part of what they meant by ‘earth’ was fixed position. Their earth, at least, could not be moved." (p.149)
     The third and most fundamental aspect of the incommensurability of competing paradigms might be referred to as world changes:
     "In a sense that I am unable to explicate further, the proponents of competing paradigms practice their trades in different worlds. One contains constrained bodies that fall slowly, the other pendulums that repeat their motions again and again. ... in some areas they see different things, and they see them in different relations "one to the other." (p.150)
     I will not repeat what has already been said about the merit of these theses.

19. How Communication is Restored by Conversion to the New Paradigm: What I want to highlight here is the further thesis that (p.150) "before they can hope to communicate fully, one group or the other must experience the conversion that we have been calling a paradigm shift." For example, (p.151) "Max Planck ... sadly remarked that ‘a new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.’" Or, (p.152) "Conversions will occur a few at a time until, after the last hold-outs have died, the whole profession will again be practicing under a single, but now different, paradigm." Partly by way of explication, and partly by way of argument, Kuhn tells us how conversion is induced and how resisted. There are three things that commonly sway scientists towards conversion.
     The first is the claim to have solved crisis-provoking problems:
"Probably the single most prevalent claim advanced by the proponents of a new paradigm is that they can solve the problems that have led the old one to a crisis. ... Copernicus thus claimed that he had solved the long-vexing problem of the length of the calendar year, Newton that he had reconciled terrestrial and celestial mechanics, Lavoisier that he had solved the problems of gas-identity and of weight relations, and Einstein that he had make electrodynamics compatible with a revised science of motion." (pp.153-54)
     The second road to conversion follows a claim to novel predictions:
"... particularly persuasive arguments can be developed if the new paradigm permits the prediction of phenomena that had been entirely unsuspected while the old one prevailed.
Copernicus’ theory, for example, suggested that planets should be like the earth, that Venus should show phases, and that the universe must be vastly larger than had previously been supposed. As a result, when sixty years after his death the telescope suddenly displaced mountains on the moon, the phases of Venus, and an immense number of previously unsuspected stars, those observations brought the new theory a great many converts, particularly among non-astronomers. In the case of the wave theory, one main source of professional conversions was even more dramatic. French resistance collapsed suddenly and relatively completely when Fresnel was able to demonstrate the existence of a white spot at the center of the shadow of a circular disk. That was an effect that not even he had anticipated but that Poisson, initially one of his opponents, had shown to be a necessary if absurd consequence of Fresnel’s theory.
" (pp.154-55)
     The third and final category of claim that may cause conversion is the claim to simplicity:
"Fortunately, there is also another sort of consideration that can lead scientists to reject an old paradigm in favor of a new. These are the arguments, rarely made entirely explicit, that appeal to the individual’s sense of the appropriate or the aesthetic-the new theory is said to ‘neater,’ ‘more suitable,’ or ‘simpler’ than the old. Probably such arguments are less effective in the sciences than in mathematics. ... Nevertheless, the importance of aesthetic considerations can sometimes be decisive. ... To see the reason for the importance of these more subjective and aesthetic considerations, remember what a paradigm debate is about. When a new candidate for paradigm is first proposed, it has seldom solved more than a few of the problems that confront it, and most of those solutions are still far from perfect. Until Kepler, the Copernican theory scarcely improved upon the predictions of planetary position made by Ptolemy. ... In short, if a new candidate for paradigm had to be judged from the start by hard-headed people who examined only relative problem-solving ability, the sciences would experience very few major revolutions." (pp.155-57)  
     These three considerations are parallel those put forward by William Whewell circa 1840, and predate that time. Whewell referred to the three tests of hypotheses as being (1) the prediction of phenomena of the same kind as those which the hypothesis was invented to explain, (2) the prediction of facts of a different and novel kind (the consilience of inductions), and (3) the tendency to simplicity and unity.

20. The Evaluation of Paradigms is Future Oriented: Kuhn correctly points out that new paradigms, when proposed, are not veterans in the problem-solving business, and cannot compete on the basis of current problem-solving accomplishments. Rather, they are selected on the promise of future success. Such promises are hard to assess, which is why it is so rare for new paradigms to take hold unless the old one is in crisis (see §5). (pp.157-58)  The further claim that such decisions can only be made on faith is the substance behind charges that Kuhn sees science as an irrational or arational enterprise on a par with religious conversion. However, Kuhn adds that there must be a basis for this faith, though it need not be a rational basis:
     "... crisis alone is not enough. There must also be a basis, though it need be neither rational nor ultimately correct, for faith in the particular candidate chosen. Something must make at least a few scientists feel that the new proposal is on the right track, and sometimes it is only personal and inarticulate aesthetic considerations that can do that.... When first introduced, neither Copernicus’ astronomical theory nor De Broglie’s theory of matter had many other significant grounds of appeal. ... This is not to suggest that new paradigms triumph ultimately through some mystical aesthetic. On the contrary, very few men desert a tradition for these reasons alone. Often those who do turn out to have been misled. But if a paradigm is ever to triumph it must gain some first supporters, men who will develop it to the point where hardheaded arguments can be produced and multiplied." (p.158)

POSTSCRIPT

21. Paradigm as Disciplinary Matrix: In his original essay, Kuhn slowly introduces ‘paradigm’ to displace the more common philosophical usage of ‘theory’. In response to charges of vagueness and even equivocation in his use of the word ‘paradigm’, Kuhn now wishes to substitute the term, ‘disciplinary matrix’:
"As currently used in philosophy of science ... ‘theory’ connotes a structure far more limited in nature and scope than the one required here. Until the term can be freed from its current implications, it will avoid confusion to adopt another. For present purposes I suggest ‘disciplinary matrix’: ‘disciplinary’ because it refers to the common possession of the practitioners of a particular discipline; ‘matrix’ because it is composed of ordered elements of various sorts, each requiring further specification." (p.182) 
     He then proceeds to define four components of a disciplinary matrix: symbolic generalizations, metaphysical presumptions, values, and exemplars.
     Symbolic generalizations. (p. 182) "One important sort of component I shall label ‘symbolic generalizations’" like f = m a, or I = V/R, or "elements combine in constant proportion by weight," or "action equals reaction." This corresponds most closely with what have been traditionally referred to as ‘theories’ or ‘laws’.
     Metaphysical presumptions. (p.184) "Rewriting the book now I would describe such commitments as beliefs in particular models, and I would expand the category models to include also the relatively heuristic variety: the electric circuit may be regarded as a steady-state hydrodynamic system; the molecules of a gas behave like tiny elastic billiard balls in random motion."
     Values:
"Probably the most deeply held values concern predictions: they should be accurate; quantitative predictions are preferable to qualitative ones; whatever the margin of permissible error, it should be consistently satisfied in a given field; and so on. There are also, however, values to be used in judging whole theories: they must, first and foremost, permit puzzle-formulation and solution; where possible they should be simple, self-consistent, and plausible, compatible, that is, with other theories currently deployed. ... Other sorts of values exist as well-for example, should (or need not) be socially useful-but the preceding should indicate what I have in mind." (pp.184-85) 
     Exemplars:
"Turn now to a fourth sort of element in the disciplinary matrix... For it the term ‘paradigm’ would be entirely appropriate, both philological and autobiographically; this is the component of a group’s shared commitments which first led me to the choice of that word. Because the term has assumed a like of its own, however, I shall here substitute ‘exemplar.’ By it I mean, initially, the concrete problem-solutions that student encounter from the start of their scientific education, whether in laboratories, in examinations, or at the ends of chapters in science texts. ... All physicists, for example, begin by learning the same exemplars: problems such as the inclined plane, the conical pendulum, and Keplerian orbits; instruments such as the vernier, the calorimeter, and the Wheatstone bridge." (pp.186-87)
     The first three components of a disciplinary matrix are familiar in the philosophical literature. Kuhn’s fourth idea of an exemplar is the most interesting element, especially in light of denial of the traditional presumption that theories and laws determine the empirical content of science. Kuhn’s claim is startling at first: (p.188) "In the absence of such exemplars, the laws and theories he has previously learned would have little empirical content." How do exemplars determine empirical content? As an example:
     "Galileo found that a ball rolling down an incline acquires just enough velocity to return it to the same vertical height on a second incline of any slope, and learned to see that experimental situation as like the pendulum with a point-mass for a bob. Huygens then solved the problem of the center of oscillation of a physical pendulum by imagining that the extended body of the latter was composed of Galilean point-pendula, the bonds between which could be instantaneously released at any point in the swing. ...Finally, Daniel Bernoulli discovered how to make the flow of water from an orifice resemble Huygens’ pendulum. ... The three problems in the example, all of them exemplars for eighteenth-century mechanicians, deploy only one law of nature. Known as the Principle of vis viva, it was usually stated as: ‘Actual descent equals potential ascent.’" (pp.190-91)

22. Tacit Knowledge Gained by Training with Exemplars: Kuhn elaborates upon the importance of exemplars as contributing to "tacit knowledge acquired through practice, which cannot be articulated explicitly." (See §2, A Paradigm is...?) (p.191): "To borrow once more Michael Polanyi’s useful phrase, what results from this process is ‘tacit knowledge’ which is learned by doing science rather than by acquiring rules for doing it."
     Kuhn denies the charges of subjectivity and irrationality that usually follow close on the heels of such psychological solutions (remember Hume), at least on two counts: (p.191) "First, if I am talking at all about intuitions, they are not individual." The sort of tacit knowledge, manifested in intuition, which Kuhn appeals to is common to all members of a community. There is no variation within a community. "Second they are not in principle unanalyzable." On the second count, Kuhn suggests that neuropsychology might provide such an analysis, in stark contrast with the rule-based analysis that philosophers have traditionally sought (p.192). Kuhn’s suggestion predates the current explosion of research into neural networks, parallel distributed computers, or connectionism.
     A possible objection is that neural states must be subject to laws, and therefore we may properly conceive of it as something we achieve by applying rules and criteria. Kuhn responds, in effect, by drawing the distinction familiar to the philosopher of mind between a merely rule-governed process (which covers everything) and a rule-following procedure, which follows a program or set of rules. Neural processing is rule-governed, but not rule-following:
     "... the fact that the system obeys the same laws ... provides no reason to suppose that our neural apparatus is programmed to operate the same way in interpretation as in perception or in either as in the beating of our hearts. What I have been opposing in this book is therefore the attempt, traditional since Descartes but not before, to analyze perception as an interpretive process, as an unconscious version of what we do after we have perceived.
     What makes the integrity of perception worth emphasizing is, of course, that so much past experience is embodied in the neural apparatus that transforms stimuli to sensations." (p.195)
     Without providing any answers, it is worth raising some questions. (1) Exactly how much past experience is embodied in the neural apparatus that transforms stimuli to sensations? (2) Assuming that scientific judgment (or intuition?) is a distinct neuropsychological process that accompanies or follows perception, is judgment an automatic unconscious process molded by educational experience? Or is judgment a slow, interpretive, rule-following procedure of the classical kind? In what follows, I will assume that Kuhn wishes to conflate perception (sensation) and judgment, and say the same about both - they are automatic, unconscious, non rule-following, procedures highly adapted to past educational experiences. (p.196) Educational training results in a stock of neurally encoded information that is used to make scientific judgments of various kinds - judgments of the relevance of data to theory, judgments of confirmation, judgments of what models to try next, judgments of simplicity, judgments of family resemblance. This does not fit the traditional view of knowledge, but the fact that it is subject to change through the discovery of misfits with the environment makes it worthy of the term:
     "But it is strange usage, for one other characteristic is missing. We have not direct access to what it is we know, no rules or generalizations with which to express this knowledge. Rules which could supply that access would refer to stimuli not sensations, and stimuli we can know only through elaborate theory. In its absence, the knowledge embedded in the stimulus-to-sensation route remains tacit." (p.196) 

23. Communication Breakdown Can Be Overcome by Inter-Group Translation: As we have seen (§12) how Kuhn vacillates between a weak and strong stance on incommensurability. In the Postscript (written 7 years later) Kuhn appears to back off from the strong incommensurability thesis, which implied a on-going and insurmountable failure of translation across paradigms: Now Kuhn suggests that:
     "... what the participants in a communication breakdown can do is recognize each other as members of different language communities and then become translators. Taking the differences between their own intra and inter-group discourse as itself a subject for study, they can first attempt to discover the terms and locutions that, used unproblematically within each community, are nevertheless foci of trouble for inter-group discussions. (Locutions that present no such difficulties may be homophonically translated.) Having isolated such areas of difficulty in scientific communication, they can next resort to their shared everyday vocabularies in an effort further to elucidate their troubles. Each may, that is, try to discover what the other would see and say when presented with a stimulus to which his won verbal response would be different. If they sufficiently refrain from explaining anomalous behavior as the consequence of mere error or madness, they may in time become very good predictors of each other’s behavior. Each will have learned to translate the other’s theory and its consequences into his own language and simultaneously to describe in his language the world to which that theory applies. That is what the historian of science regularly does (or should) when dealing with out-of-date theories."  (p.202)
     Part of what had previously obscured Kuhn’s view of the matter was his conflation of ‘persuasion’ with ‘conversion.’ In his original essay, Kuhn assumed that the only way that someone could understand the opposing viewpoint was to be converted to the new paradigm. Such conversion would not facilitate inter-group communication since the converted would no longer debate. But now Kuhn concedes that someone may see the other point of view, and even be persuaded by it, without being converted to the new paradigm (p.204) 

24. Rationality: Is Kuhn’s Description of Science Also a Prescription? A remarkable feature of Kuhn’s essay is how little Kuhn says or implies, in detail, about what would happen if things in science were different. Mostly, such implications are contained in his discussion of how the scientific community is different from other communities (see §24 Yet Science Does Progress, and §25 Progress by Natural Selection). Kuhn even restates his position in purely descriptive terms:
     "Though scientific development may resemble that in other fields more closely than has often been supposed, it is also strikingly different. ... Consider, for example, the reiterated emphasis, above, on the absence or, as I should now say, on the relative scarcity of competing schools in the developed sciences. ... Or think again about the special nature of scientific education, about puzzle-solving as a goal, and about the value system which the scientific group deploys in periods of crisis and decision." (p.209)  
     The implication is that if the scientific community weren’t like that, then it would not accumulate problem-solutions. Kuhn reports that (p.207) "Some critics claim that I am confusing description with prescription, violating the time-honored philosophical theorem: ‘Is’ cannot imply ‘ought.’" By some kind of inference to the best explanation, he thinks his theory is warranted because scientists "behave as the theory says they should." If we include the suppressed premise that scientists behave rationally (so they behave as they ought to behave), then it might be a good argument. But it seems just all too quick and easy! For one thing, there is no reason why the same argument can’t be used to argue, say, that religious fundamentalists ought to read the bible if their goal is to know the truth about evolution, since their methods have been developed and selected for their success, do in fact behave as the theory says they should.
     I think that Kuhn would be better of if he denied that his claims are prescriptive, or normative, at all. For example, the implication above is not that scientists ought to aim at the accumulation of problem-solutions, only that if that is their goal, then certain structural features of their community will be causally influential in bringing about this end. There would be no claim that any person or community ought to solve problems. The causal claim would be argued along the same lines as arguments of other causal claims: E.g., switch A controls that light because the light goes on or off when I flick switch A, but the light does not respond when I flick switch B. Likewise, problem-solutions are achieved by communities of type A, but such achievements are not replicated by community B. There is no inference of ‘ought’ from ‘is’; just a causal conclusion being inferred from relevant evidence.
     The problem is that Kuhn’s causal inferences are unsupported. As in the light switch example, no clue is given of the causal mechanisms behind those problem-solving achievements. What are the quite special characteristics of the world that allow for such achievements? As Kuhn stated in the closing paragraph of his original essay, such questions like "what must the world be like in order for us to know it?" are left unanswered.