Science and Relativism: Some Key Controversies in the Philosophy of Science / Edition 1 available in Paperback
Science and Relativism: Some Key Controversies in the Philosophy of Science / Edition 1
- ISBN-10:
- 0226469492
- ISBN-13:
- 9780226469492
- Pub. Date:
- 08/15/1990
- Publisher:
- University of Chicago Press
- ISBN-10:
- 0226469492
- ISBN-13:
- 9780226469492
- Pub. Date:
- 08/15/1990
- Publisher:
- University of Chicago Press
Science and Relativism: Some Key Controversies in the Philosophy of Science / Edition 1
Paperback
Buy New
$26.00Buy Used
$15.99-
-
SHIP THIS ITEM
Temporarily Out of Stock Online
Please check back later for updated availability.
-
Overview
Product Details
| ISBN-13: | 9780226469492 |
|---|---|
| Publisher: | University of Chicago Press |
| Publication date: | 08/15/1990 |
| Series: | Science and Its Conceptual Foundations series |
| Edition description: | 1 |
| Pages: | 194 |
| Product dimensions: | 5.25(w) x 8.00(h) x 0.70(d) |
About the Author
Read an Excerpt
Science and Relativism
Some Key Controversies in the Philosophy of Science
By Larry Laudan
The University of Chicago Press
Copyright © 1990 The University of ChicagoAll rights reserved.
ISBN: 978-0-226-46949-2
CHAPTER 1
Progress and Cumulativity
DAY 1, MORNING
Pragmatist: Gentlemen, I think we should begin, since we are already a bit behind schedule. Having been named chair of this committee, I should say that I interpret my charge to be that of seeing to it that our discussions remain focused on our central tasks and that we do not chase after too many wild hares. We already know one another, having crossed swords on several previous occasions, so I think that no preliminaries of that sort are called for. But we probably should give some thought to selecting the key topics that will form our agenda.
Relativist: Since our brief is relativism, especially as regards scientific knowledge, and since I am the only card-carrying relativist here, I have some suggestions to make about what the salient issues should be. Above all, I think that we should start with the collapse of positivism and foundationalism and move from there ...
Realist: Forgive me for interrupting, Quincy, but the fact that you are keen on relativism gives you no special claim to set our agenda. All of us here have thought about relativism for a long while; the fact that we have rejected it, and that you have accepted it, is neither here nor there.
Positivist: I wonder if, rather than attempting to legislate our full agenda here and now, we couldn't avoid this procedural wrangling by simply agreeing to start somewhere and then take up the topics as they flow naturally from the exchange?
Pragmatist: I wholeheartedly concur, Rudy. Why don't you suggest a place for us to begin.
Positivist: Well, as we all know, one of the key issues in the epistemology of science has concerned the question of the growth of scientific knowledge; thinkers from Peirce to Popper have insisted on the centrality of that problem to scientific epistemology Nor is it philosophers alone who are preoccupied by it. Scientists and laymen similarly agree that one of the striking features of the diachronic development of science is the progress that it exhibits. The philosophical challenge is to find ways to characterize that "progress" as clearly and as unambiguously as we can. Perhaps therefore this would be an appropriate place for us to begin our explorations. And since our brief is to examine the status of contemporary relativism, maybe we could ask Quincy to kick off the discussion with a characterization of the relativist view of cognitive progress.
Relativist: I have no objections to our starting there, if you like, since I share your view that people have an abiding faith in the progress of science. Equally, however, I want to go on record straightaway as having grave reservations as to whether there is any robust, objective notion of the growth of knowledge. I happily grant you that our theoretical understandings and representations of the natural world change dramatically through time, though whether those changes represent "progress" or simply change is unclear. But I think that it would be reversing the natural logic of this subject for me to start things off. Most of us relativists reject the notion of progress because the two well-known accounts of scientific progress — associated with positivism and realism respectively — have been dismal failures. Accordingly, and I can assure you not out of any shyness on my part, I would urge Rudy or Karl to tell us whether they have a coherent theory of scientific progress to put forward. I will fill out my position in response to what they have to tell us.
Pragmatist: That's certainly agreeable as far as I'm concerned. Perhaps we can call on Rudy to outline the problem of progress or theory change as he sees it, since the positivist account of scientific progress probably remains the best known.
Positivist: Gladly. In a nutshell: science is the attempt to codify and anticipate experience. The raw materials of science are observational data or measurements. We develop theories and laws to correlate, explain, and predict those data. A science progresses just to the extent that later theories in a domain can predict and explain more phenomena than their predecessors did. Since the seventeenth century, the sciences — at least the natural sciences — have done just that.
Relativist: Hang on a minute. When you talk about what a theory "can" predict and explain, are you referring to what it has explained and predicted or are you talking about everything that it might be able to predict and explain?
Positivist: You can take it in either sense since science exhibits impressive credentials of both sorts.
Relativist: Well, if we focus on the first sense, what one might call demonstrated progress, I will grant you that some theories have managed to predict and explain some things not guessed at by their rivals. But I'm not sure that this difference gives us a sound basis for maintaining that one theory is really better or truer than another. After all, the fact that one theory has more proven predictive successes to its credit than a rival might be just an artifact of how long each has been around, how assiduously their applications have been explored, how many scientists have worked on them, etc. You surely don't want to argue that the goodness of a theory is a matter of such accidental circumstances as these?
Positivist: Well, as I said earlier, genuinely progressive theories are those which have the capacity to explain and predict a larger range of facts than their rivals. That is, in part, a prospective matter. I accept your point that one theory's known successes might have more to do with these accidents of history than with anything about the theory itself.
Relativist: But if you're saying that, in deciding whether one theory represents progress over another, we have to compare their prospective explanatory and predictive ranges, then I don't see how we could ever settle that issue since — as you just noted — we can never know all the consequences of any theory, infinite as that set is. You face a dilemma, Rudy: comparing the known achievements of rival theories can be done but is indecisive since those achievements will be in part a function of various accidents of the distribution of labor in the scientific community; yet it is impossible to compare the potential but unknown capacities of rival theories.
Positivist: Quincy is right, up to a point. A theory, any theory, has an infinite range of consequences only some of which will have been examined at any chosen stage of inquiry. But we are nonetheless often in a position to make dependable judgments about the prospective scope of rival theories, even when (as will always be the case) we have actually tested only some of those theories' consequences. Indeed, I can describe for you a procedure which will allow us to make such judgments even if we have never tested any of the rivals. Let us suppose that we have two theories under consideration, call them T1 and T2. Suppose further that we can show that T2entails T1. Under such circumstances, we know that T2 must have all the consequences of T1, as well as some additional consequences besides (provided that T1 does not entail T2). Hence if a later theory entails an earlier one, but is not entailed by it, then we know that the later theory must be more general than the earlier.
Relativist: I'm not sure I see the point ...
Positivist: It's simply this: A few moments ago, you said that the comparison of the prospective successes of rival theories was impossible, suggesting thereby that we positivists have no viable theory of scientific progress. What my latest example shows is that we can often demonstrate, even prior to any testing, that one theory is more general than another.
Pragmatist: But surely generality alone, in the sense of a maximally large class of entailments, is not scientific progress. If it were, then the adumbration of tautologies — which imply all true statements — and contradictions, which entail everything, would represent the ideal end point of science.
Positivist: What's wrong with tautologies and contradictions alike is not their lack of generality but their nonamenability to empirical test and thus their low information content. When I say that one theory represents progress over another as long as the former entails the latter (but not vice versa), I mean to refer only to theories per se, i.e., to sets of universal statements which are genuinely empirical by virtue of their prohibiting certain states of affairs. Tautologies and contradictions prohibit nothing and are thus not in the class of theories.
Relativist: What you appear to be saying is that scientific progress can occur only if (a) one testable theory succeeds another and if (b) the later theory entails the earlier. But Duhem, Quine, and a host of others have shown that scientific theories are not falsifiable and hence not testable. They showed specifically that any theory whatever can be retained in the face of recalcitrant evidence, provided we are prepared to make drastic enough changes elsewhere in our framework of beliefs.
Pragmatist: I think, Quincy, we should tackle one issue at a time. We all know you believe theories are nonfalsifiable in principle, and I for one am prepared to set aside one of our later sessions to deal specifically with that issue. But I wonder if for now we shouldn't allow Rudy to finish setting out his position on scientific change and progress.
Positivist: Thanks for the intervention, Percy. Our relativist friend was overhasty, for I should be the last to claim that greater generality in our theories is a sufficient condition for scientific progress. My claim thus far was simply that greater generality was a necessary condition for making a well-founded claim of progress.
Pragmatist: Well, what more is required?
Relativist: Rudy is doubtless on the verge of answering that question, but I wonder if I could interrupt here. He just told us that greater generality was a necessary condition for scientific progress, and I have no quibbles with that claim as far as it goes. But I think we have been too hasty in accepting Rudy's assimilation of greater generality to some logical relation of entailment.
Positivist: You're unclear about what an entailment is?
Relativist: In fact I am, since it seems to me that what a theory entails depends upon the other theories and assumptions with which it is conjoined; but that is not my worry for the moment. My problem is this: I grant you that if one theory entails another theory and not vice versa, then the former is more general than the latter. But surely we can think of situations in which one theory is more general than another even when the relevant entailment relations fail to obtain.
Positivist: Such as what?
Relativist: Well, I'm inclined to think that the theories of quantum mechanics are more general than theories of ecology, even though I couldn't begin to derive the latter from the former.
Positivist: You're surely right, but I don't think I'm committed to claiming that one theory is more general than another only if an entailment relation obtains between them. I see one-way entailments as a sufficient condition for making judgments of generality, not a necessary condition.
Relativist: So, on your view, one theory might be more general than another, and thus potentially progressive, even though neither theory entailed the other?
Positivist: Of course; why are you bothered?
Relativist: It seemed to me that you were suggesting that deductive logic was enough to settle issues of progress, and I have the impression that it is otherwise.
Pragmatist: I wonder if we couldn't return to where we were before this digression. I think that Rudy had said that progress judgments require more than greater generality. What else is involved?
Positivist: Well, above all, we expect the more progressive theory to be better confirmed than its predecessor. We expect it to have enabled us to explain and predict phenomena which its predecessor either couldn't explain at all or which its predecessor predicted incorrectly.
Realist: But if the earlier theory made an incorrect prediction and the later theory — by virtue of entailing its predecessor — and I believe you called that a necessary condition for progress — exhibits all the consequences of its predecessor, then any false prediction made by an earlier theory is also going to be made by its successor. Hence how could a later theory possibly both entail its predecessor and predict something correctly which "its predecessor predicted incorrectly"?
Pragmatist: Are you suggesting, Karl, that the old Baconian ideal that later theories should "contain" their predecessors is bankrupt because it would require later theories to incorporate all the failures of earlier ones?
Realist: Exactly. Once we realize that most theories in science are given up precisely because we have found them to be false, then it follows that the last thing we want to insist on is that their successors must capture all their empirical consequences! Perhaps I can put my challenge to the positivists most concisely in this form: I can see that a later and more general theory might well make some predictions — including correct predictions — concerning matters about which its predecessor was wholly silent; but how could the later theory manage to avoid the incorrect predictions made by its predecessor if it entails that predecessor?
Positivist: You have a point, Karl. Answering it will require me to be a bit more precise and detailed than I have been thus far. Let us distinguish, within the context of what I have been loosely calling a theory, between two elements: the theory per se and the associated experimental laws. Laws coordinate observations, and theories coordinate laws. Now, when we discover that a theory has broken down, that is has some false consequences, what we are really discovering is that some of the lawlike generalizations coordinated by that theory are false, i.e., they are not laws at all. Of course, as Karl says, we don't want to demand that a later theory must replicate the known failures of its predecessors.
Relativist: So now what is your story about the relation between successive theories in a progressive science?
Positivist: I suggest that what we expect a progressive theory change to do is to produce a successor theory which (a) retains all the nondiscredited, lawlike statements associated with the earlier theory, (b) drops out those pseudo-laws which have already been refuted, and (c) introduces some new lawlike regularities not previously encompassed within the predecessor theory. Things are even clearer if (d) some of the lawlike statements associated with the later theory, and not embraced by its predecessor, correctly predict hitherto unexplained and unpredicted phenomena. When all these conditions obtain, then we have a paradigmatic case of progressive theory change.
Realist: Although my positivist friend and I disagree about many matters, I find myself almost wholly in agreement with the characterization he has just offered. Perhaps consensus on these issues is within our reach.
Pragmatist: Let's not be too hasty. Rudy's definition of theoretical progress in science does beg a few questions. For one thing it requires us to accept a distinction between a theory per se and the lawlike statements associated with it. As I understand it, that distinction is to be drawn chiefly in light of a distinction between observational terms — which is what occurs in the "laws" — and nonobservational terms which occur in the theory per se. Have I got that right?
Positivist: Of course.
Pragmatist: In that case, I cannot accept your distinction, since I doubt that there is a sharp line between theoretical terms and observational terms; and I suspect that Quincy will find it equally objectionable. And so should Karl, despite his initial willingness to accept Rudy's account of progress, since Karl's realism about science hinges crucially — if I understand it — on the repudiation of any sharp observational/theoretical dichotomy. But even if I were to grant you, Rudy, that there is a viable distinction between what is observational and what is not, I would still have problems with your characterization of scientific progress.
Positivist: And what would those be?
Pragmatist: You originally told us that a sine qua non for progress was that later theories must entail their predecessors. When some of us noted that such a policy would involve retaining all the failures as well as the successes of earlier theories, you backed off and conceded that entailment between theories was too strong a condition. Now you are telling us that one theory is better than another if, among other things, the later theory retains all the nondiscredited lawlike statements of its predecessor.
Positivist: Quite.
Pragmatist: But how does one tell what those are? Since no one believes that we are fully aware of all the empirical or observational consequences of any theory, how can we ever be reasonably confident that a later theory has retained all the (correct) observational consequences of its predecessor?
Relativist: That sounds a bit like my earlier worry about how we can possibly compare the unknown, prospective features of different theories.
Pragmatist: Indeed it does. For all we know, a new theory might have ignored or dropped out many of the correct but unknown laws associated with an earlier theory. Thus, giving up the earlier theory, because of certain known failures, and replacing it by a theory which incorporates the earlier theory's known successes and avoids its known failures — if we could manage to do that — offers us no assurance that the new theory will generally work better than the old one might have.
(Continues...)
Excerpted from Science and Relativism by Larry Laudan. Copyright © 1990 The University of Chicago. Excerpted by permission of The University of Chicago Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
PrefaceNote to the Reader
1. Progress and Cumulativity
2. Theory-ladenness and Underdetermination
3. Holism
4. Standards of Success
5. Incommensurability
6. Interests and the Social Determinants of Belief
References
Index