1. The growth of mature sciences is punctuated by periodic revolutions involving an overthrow of an older paradigm in favor of a new one. This "paradigm switch" changes the scientists' world-view in a way which makes phenomena that were unexpected or anomalous on the old paradigm now seem expected when nature is viewed from within the new paradigm.
2. During the periods of normalcy between revolutions, scientists practice their calling by virtue of the common possession of a paradigm. A paradigm comes to mark the maturity of a science and thereby defines its first period of "normalcy" by what is seen as a string of successful achievements in explaining some of the phenomena pertinent to that field -at least as seen through the eyes of those researchers who share the paradigm.
3. At the same time the paradigm leaves open-ended a field of possible further problem solutions modeled on the original successes that got the paradigm going. These archetypical applications of theory are learned as exemplars forming the basis of the scientists training. Becoming a good scientist involves learning how to model new puzzle solutions after the patterns suggested by such exemplars.
4. Possession of such a paradigm generally involves accepting specific applications of the central paradigm theory to the solution of problems raised in explaining phenomena which the paradigm stipulates as being particularly revealing of nature's fundamental patterns of behavior. But a paradigm should not be simply identified with a central theory. Different paradigms could both conceivably exploit the resources of the same theory, and yet involve fundamentally different conceptions of both nature and the basic problems of the field. Furthermore, a theory might change and develop, through normal articulation, so as to be dramatically transformed over the course of a long range of normalcy.
5. Thus the paradigm is in effect manifested in the cognitive behavior and language of a community of practicing researchers. The shared language, standard problems, accepted solutions, techniques for obtaining data, agreed methodologies for conducting both theoretical and experimental work, and ultimately ontological commitments about the sort of entities which may be invoked in explanations enable the scientists sharing a paradigm to easily reach consensus on what it is rational to accept. They mark out a certain "disciplinary matrix" in which those who share the paradigm perform their research.
6. A disciplinary matrix, among other things, provides a technical vocabulary for describing phenomena, the standards for what constitutes an acceptable explanation of a phenomenon, the sorts of acceptable answers to questions, the sorts of entities which may be invoked in acceptable explanations, their interrelations, methods by which observations may be made, and in short, all of those beliefs pertinent to form the scientist's judgment of what constitutes doing "good" science within that discipline.
7. As each paradigm contains within itself the appropriate language for describing observation and the standards for acceptable problem solutions, in the shift from one paradigm to another, the facts which it is assumed to be the job of theory to explain will change. Furthermore, the standards which determine what counts as an acceptable solution will also change. Hence Kuhn argues (as does Hanson) for the theory-dependence of observation, or, in his words, the paradigm-dependence of the "world of the scientists' research engagement." In other words, what one can see in observing the world is at least in part determined by the conceptual scheme of the paradigm and what one's paradigm leads one to look for (and does not lead one to look for).
8. Hence, in a revolution the facts which form the observational evidence on the basis of which the scientist makes decisions to accept or reject beliefs will change. Insofar as "the world" is constituted of such facts, the world he/she will live in will change. Of course scientists may use the same words, i.e. they may "say" the same thing, but inasmuch as meaning is dependent on the whole constellation of one's beliefs, a change of paradigm may involve a change in meaning of the theoretical and/or observational terms. Thus Kuhn asserts that meanings, too, are paradigm dependent, at least partially. Nevertheless, Kuhn holds that his position does not necessarily imply the radical idealist thesis that a change of thought implies a change in reality; i.e. that "reality" is totally determined by human thought. What it does require is more modest: a change in the constellation of the beliefs of the scientist will produce a change in the world "of the scientists' research engagement."
9. Nevertheless, without a paradigm at least theoretical terms can have no meaning in isolation, and since theory determines what we will observe, there can be no neutral observation language in which to express the claims of competing paradigms. For the same reasons, there can be no "theory neutral" set of standards by which one paradigm can be objectively evaluated as "better" than another. Thus, while Kuhn is careful to emphasize that there are good reasons which may be advanced pro and contra for each scientist who makes a paradigm switch, these reasons can never be compelling, nor need they be the same for all scientists. In short there can be no "logic" which determines theory choice, as was held to be the case in the received view's analysis of scientific methodology.
10. Thus the choice to adopt a new paradigm is always, at the time when such a choice is made, something of a "judgment call" which involves more than what "the empirical evidence" will dictate. Consequently, in converting scientists to a new paradigm, methods of persuasion other than those based upon the "facts" must be utilized. This does not entail, as has been charged, that successful theories in science convert by "mob psychology," but it does entail that in the final analysis there is no other standard than "what the community of scientists adopts" as that which determines theory choice in science.
11. Thus Kuhn locates the "authority" for theory choice, i.e. that to which we appeal to justify whether we should accept or reject a theory, in the collective decision of the people who make up the community of scientists working within that paradigm. This move can be seen as abandoning the older view that theory choice depends on an objective scientific method which, if correctly applied will dictate that choice on the basis of allegedly "given" observational evidence. Such choices may now be seen as "subjective" because that two equally competent scientists faced with the same observational "evidence" may equally rationally defend opposing choices. This consequence of Kuhn's analysis has brought forth a debate which goes under the name of "the rationality crisis in philosophy of science."
12. A more radical thesis, however, is entailed by the fact that if paradigms constitute mature sciences as Kuhn maintains they do, and if paradigms determine observation in the way that Kuhn claims they do, there is no getting around the conclusion that scientists operating in different paradigms will experience reality differently. Kuhn may well grant that there is a scientist-independent reality that is what it is quite independently of the scientist and his beliefs. But this does not alter the fact that what is used to provide the empirical evidence for or against adopting a theory is not some "reality" as it exists independently of the scientific observer, but rather reality as experienced by the practicing scientists who, by the very nature of science, cannot practice their science without commitment to some paradigm.
13. Thus one can meaningfully speak of science as learning more about "the world" or increasing (or "accumulating") our knowledge of the world only within the context of a period of normalcy. Over a revolutionary divide knowledge does not necessarily quantitatively increase, but changes qualitatively its mode of explanation. Thus it may happen that fewer phenomena stand as satisfactorily explained after the revolution than before it (but importantly the crisis provoking anomaly now rests with the explained phenomena). This phenomenon is sometimes called "Kuhn loss."
14. Insofar as one accepts a correspondence notion of truth, i.e. a statement is true to the extent that it corresponds to the way things really are, if reality is "reality as experienced by the scientist" and that reality is paradigm dependent, there can be no growth in or closer approach to "the truth" over a revolutionary divide.
15. Thus progress in science does not involve growth in truth. Instead it is to be understood in terms of developing paradigms which are successful in solving puzzles over broader and broader ranges of phenomena, or developing ever more extensive or more precise ways of seeing the world such that it more extensively or more precisely behaves the way our way of seeing the world leads us to expect it to behave.
16. Thus if Kuhn's analysis is correct we have no reason to believe that as science grows, it is telling us more and more about reality. The history of science is not the story of how we have learned more and more about "the one world which really exists" apart from human scientific experience, rather it is the story of how human beings have created more and more sophisticated ways of viewing the world, or "more powerful conceptual and instrumental 'tools'" such that it led them to live in a world which behaved as such human beings expect it ought to behave.
17. If "scientific realism" is defined as the view that science gives us knowledge about the nature of an independent reality, Kuhn's philosophy seems to leave no grounds for such a realist interpretation of science, and instead it gives us every reason to accept the anti-realist view that both truth and the nature of reality as it exists independently from the phenomena we observe are irrelevant to the acceptance and/or rejection of scientific explanations.