Putnam's Realism and Van Fraassen's Constructive Empiricism Realism in Mathematics and Elsewhere By Hilary Putnam I. Defines realism: (317) A. borrowing from Michael Dummet, says that a realist with respect to some theory holds: 1. that the sentences in that theory are true or false 2. that what makes them true or false is something outside the theory B. Skip the rest of pp. 317-9 having to do with realism in mathematics II. Realism with respect to empirical science (319) A. rests on two sorts of arguments: 1. negative, pointing to failures of other philosophies, that either attempt to reduce everything to experience or operationalize theoretical concepts 2. positive, the argument that only scientific realism can explain the success of science. It is the only philosophy that does not make the success of science appear to be a miracle B. Explaining the success of science 1. the scientific realist holds that the only scientific explanation of the success of science rests on the following realist premises: a. that the terms in mature scientific theories typically refer to real things b. that the theories in mature sciences are typically at least approximately true c. that when the same term occurs in different theories it can nevertheless refer to the same thing (320) 2. the scientific realist is not claiming that these three premises must be true, but only that they are part of the only explanation of the success of science, and hence of any adequate account of the relation between science and its objects C. Putnam similarly believes that realism is the only philosophy that does not make the success of mathematics a miracle 1. indeed, he thinks it is not possible to be a realist about physics and a nominalist about mathematics 2. for instance, if one is a realist about the physical world, then one wants to say that the law of universal gravitation makes a statement about not just actual bodies, but about certain quantities associated with them Alternatives to Realism By Bas van Fraassen I. Introduction (322) A. defines scientific realism in terms of the aims of science: 1. science aims to provide a literally true story of what the world is like 2. to accept a theory is to believe that it is true B. anti-realism is then the position that: 1. the aim of science is not to give us a literally true description of the world 2. to accept a theory is not necessarily to believe that it is true, but it may be to propose that it is empirically adequate, or possesses some other virtue C. Anti-realists divide into two kinds in accordance with their positions regarding literal truth: 1. science aims at the truth, but not at the literal truth 2. that the language of science is to be construed literally, but that science does not necessarily aim at the truth. This is van Fraassen's own position. (322) D. will not provide a philosophical analysis of the concept of "literal" 1. that job belongs to the philosophy of language 2. but will assume our common sense understanding of it, for instance when we speak of a literal as opposed to an allegorical, metaphorical, or analogical reading of the Bible E. Van Fraassen's decision to accept a literal interpretation of the language of science distinguishes his views from both positivism and instrumentalism (323) 1. on a literal construal, general statements in science are genuine statements, capable of being true or false -- and not merely rules for deriving one statement from another 2. a literal construal can elaborate but not change logical relationships a. for instance, if statistical mechanics says that temperature of a gas is the average kinetic energy of its molecules, this may be elaborating on what heat is, but it does not remove the logical implication that heat must exist b. on the positivist interpretation of science, on the other hand, theoretical terms take on a meaning only in virtue of their relationships to the observable (323) 1.) hence, for the positivist, two theories that contradict each other but nevertheless have the same observational consequences would have the same meaning 2.) it is possible for two theories that contradict each to be regarded as saying the same thing only if they are not literally construed N.B. Hume in effect makes this argument in denying any difference between theism and atheism II. Constructive Empiricism (323) A. to insist on a literal construal of science is to rule out a metaphorical reading N.B. thus van Fraassen's constructive empiricism should not be confused with social constructionism, which is not committed to a literal reading of scientific theories B. to hold a literal interpretation of the language of science does not make one a realist. 1. that is, it implies nothing about either: a. epistemic questions, such as whether we must believe that scientific theories are true, or that the entities that they postulate are real b. what our aims should be 2. a literal interpretation has to do only with what a theory says (323, q.v.) C. van Fraassen advocates a literal construal of scientific theories, but in place of realism, proposes "constructive empiricism:" Science aims to give us theories that are empirically adequate; and acceptance of a theory involves as belief only that it is empirically adequate. (324) 1. by "empirically adequate," he means that what it says about the observable things in the world is true, that it "saves the phenomena" 2. that is, the theory has at least one model into which we can fit all of the phenomena. By "all," he means not just the ones that have been observed, nor even those observed at some time in the past, present, and future D. theory acceptance 1. acceptance of a theory involves more than just belief a. since theories are never complete, for a scientist to accept a theory is to commit him or herself to work in a particular research program b. even for a non-scientists, it may involve a commitment to use the theory in scientific explanations c. For van Fraassen, acceptance of a theory also entails the confidence that your acceptance of it will be vindicated 2. but these issues don't divide the realist from the anti- realist the way that epistemic issues do a. neither the realist belief that a theory is true, nor the constructive empiricist theory that a theory is empirically adequate, implies, or is implied by, the belief that acceptance of the theory will be vindicated b. for instance, both the realist and the constructive empiricist might think human beings will destroy themselves in a war before science can vindicate their theory choice c. whether either philosophy can be equated with the belief that a theory choice would be vindicated under ideal conditions in the long run is a separate question (324-5) 1.) van Fraassen thinks it has no answer, as he thinks counterfactuals are neither true nor false 2.) however, the question may be irrelevant since even an affirmative answer would not undermine the distinction between realism and constructive empiricism we've already drawn III. Demons and the Ultimate Argument (325) A. Hilary Putnam, over the course of his career, has given several different arguments for realism B. First, he defends the argument that theoretical concepts and mathematical entities are indispensable 1. he defends it against factionalism, the view that these entities are just useful fictions 2. Putnam begins by rejecting bad arguments against Fictionalism, and then provides his own reasons for rejecting it 3. the bad argument against fictionalism comes from the verificationist theory of meaning a. for the logical positivists, who were verificationists, the meaning of a theory is a function of the empirical results that would confirm or disconfirm it b. hence, two different theories with the same empirical content nevertheless have the same meaning. For example, Rutherford's atomic theory and Vaihinger's theory that, although perhaps there are no such things as atoms and electrons, the observable world behaves as if there were (325) 4. this is a bad argument for Putnam, because one theory says there are electrons, and the other allows that there may not be 5. the positivists would reply that Putnam's position leads to skepticism a. that is, that there are things we cannot prove or disprove by experiment, and thus that we cannot really know what the world is like b. even worse, you have no reason to reject the belief in demons 6. Putnam's answer to the charge of skepticism is that when faced with 2 hypotheses with the same testable consequences, reject the one with the lower prior probability, in the Bayesian sense 7. van Fraassen finds this disappointing, and argues that it actually favors Vaihinger's anti-realist stance over Rutherford's realist beliefs 8. yet Putnam has done us the favor of refuting verificationism, by arguing that theories could agree in empirical content and differ in truth-value 9. in the end, for Putnam, the realist has to make a leap of faith that is not dictated by reason or evidence: that is, once you have evidence of the truth of a theory, there is nothing more that you could ask for to make it rational for believe the theory in a realist sort of way (326) C. in a subsequent publication, Putnam adopts what van Fraassen calls the "ultimate argument" for realism 1. Putnam begins by laying Dummet's formulation of realism. A realist is someone who holds: a. that the sentences of a theory are true or false b. that they are true or false in virtue of an independently existing external reality (326-7) c. they are true or false independently of our knowing whether they are true or false (327) 2. an anti-realist, on the other hand, would hold that statements are to be understood only by reference to the sort of thing that would count as evidence for them 3. this differs from the usual understanding of realism, in that it is a philosophical position about statements rather than one about unobservable entities. a. in this sense, as Dummet recognizes, a nominalist could be a realist. That is, one could deny that a certain sort of entity such as mathematical sets exist, yet believe that there are true statements about them (327) b. nor could one say that Dummet has provided just a necessary condition for realism -- one could be a realist about elementary particles, yet say that if "particle x has momentum p at time t" is true, then "particle x has position q" is neither true nor false. 4. van Fraassen says that Dummet may be right that what is really at stake or is important in the realism debates are questions about language, but is uncomfortable with Dummet's use of traditional philosophical terminology to talk about these issues (327-8, q.v.) a. van Fraassen would want to define realism in such a way that it did not necessarily imply that every statement in a theoretical language was either true or false b. and van Fraassen's own constructive empiricism would not be anti-realist in Dummet's sense, since he too holds that statements are true or false independent of human knowledge of them 5. But van Fraassen does not really conceive the dispute as being about language 6. in the same paper, Putnam then seemed to abandon Dummet's weak realism for Boyd's realism. Boyd is the source of what van Fraassen calls the "ultimate argument" a. accordint to which realism is the only philosophy that does not make the success of science to be a miracle b. the argument says the following: 1.) that the terms of mature theories typically refer 2.) that the accepted theories in a mature science are typically approximately true 3.) that the same term can refer to the same thing even when the term is being used in different theories c. these three things are not necessary truths for the realist but are part of any scientific explanation of the success of science D. for van Fraassen, the realist's explanation of the success of science looks more Scholastic than scientific 1. van Fraassen would argue instead that science is a biological phenomenon, an activity by one kind of organism that helps it adapt to its environment. 2. to make the point, compares two explanations of the mouse who runs from the cat a. a traditional account, going back to St. Augustine, in terms of what the mouse perceives as agreeing with what is true in nature (328-9) b. a Darwinian account, that says species that do not escape their enemies go extinct 3. in the same way, van Fraassen would say that the success of current scientific theories is no miracle, because only successful theories survive