+ All Categories
Home > Documents > On Scientific Justification by Consensus

On Scientific Justification by Consensus

Date post: 18-Nov-2023
Category:
Upload: luc
View: 0 times
Download: 0 times
Share this document with a friend
9
Springer is collaborating with JSTOR to digitize, preserve and extend access to Zeitschrift für allgemeine Wissenschaftstheorie / Journal for General Philosophy of Science. http://www.jstor.org On Scientific Justification by Consensus Author(s): Paul K. Moser Source: Zeitschrift für allgemeine Wissenschaftstheorie / Journal for General Philosophy of Science, Vol. 17, No. 1 (1986), pp. 154-161 Published by: Springer Stable URL: http://www.jstor.org/stable/25170737 Accessed: 31-12-2015 22:48 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTC All use subject to JSTOR Terms and Conditions
Transcript

Springer is collaborating with JSTOR to digitize, preserve and extend access to Zeitschrift für allgemeine Wissenschaftstheorie / Journal for General Philosophy of Science.

http://www.jstor.org

On Scientific Justification by Consensus Author(s): Paul K. Moser Source: Zeitschrift für allgemeine Wissenschaftstheorie / Journal for General Philosophy of

Science, Vol. 17, No. 1 (1986), pp. 154-161Published by: SpringerStable URL: http://www.jstor.org/stable/25170737Accessed: 31-12-2015 22:48 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

On Scientific Justification by Consensus

PAUL K. MOSER

Zusammenfassung

Nach vielen gegenwartigen Wissenschaftstheoretikern ist die Wissenschaftstheorie des Logi schen Empirismus, wie sie in den Schriften von Carnap, Russell, Reichenbach und Hempel vertreten wird, durch die neue Wissenschaftstheorie wesentlich verbessert worden, wie sie von

Hanson, Polanyi, Toulmin und Kuhn entwickelt worden ist. Aber keiner der letzteren Gegner des

Logischen Empirismus hat im Detail die Erkenntnistheorie herausgearbeitet, welche der neuen

Wissenschaftstheorie zugrundeliegt. Kurzlich jedoch hat Harold I. Brown, in Perception, Theory and Commitment The New Philosophy of Science (University of Chicago, 1979), eine klare

Formulierung dieser neuen, consensualen Erkenntnistheorie vorgelegt. In dem vorliegenden Artikel entwickele und bewerte ich die Ansichten von Brown und Kuhn als Reprasentanten der

neuen Erkenntnistheorie. Alles in allem begriinde ich, daft die neue Erketintnistheorie bestenfalls

eine aufterst unvollstandige Alternative zur logisch-empiristischen Erkenntnistheorie liefert.

According to many philosophers of science, logical empricist philosophy of

science, as represented by the writings of Carnap, Russell, Reichenbach and

Hempel, has been much improved upon by the new philosophy of science

developed by Hanson, Polanyi, Toulmin and Kuhn. But none of the latter

opponents of logical empircism has worked out in detail the epistemology

underlying the new philosophy of science. In fact, it is somewhat doubtful that these theorists would all subscribe to a single epistemology. Recently, however, Harold Brown has tried to provide a clear formulation of the

epistemology implicit in the new philosophy of science.1 In many ways the

epistemology developed by Brown extends the epistemological views of

Thomas Kuhn. In fact, one might argue that a unified epistemology arises from

the views of Kuhn and Brown. In any case, both writers advance what I call

'consensual epistemology', i. e., an account of scientific justification by consensus. So, in what follows, I shall look mainly to their writings for the

subject-matter of my assessment of consensual epistemology.

I

A central tenet of consensual epistemology is that we should reject any mathematical model of the justification of scientific theories. That is, we

should oppose the view that the justification of scientific theories arises from

the application of an algorithm. Rather, according to consensual epistemology, we should recognize the pervasive role of fallible human judgment in all

scientific justification. Brown emphasizes this point by using Aristotle's

1 Harold I. Brown, Perception, Theory and Commitment: The New Philosophy of Science

(Chicago: University of Chicago Press, 1979), Part II, especially Chapter 10.

Zeitschrift fiir allgemeine Wissenschaftstheorie XVII/1 (1986) ? Franz Steiner Verlag Wiesbaden GmbH, Sitz Stuttgart

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

On Scientific Justification by Consensus 155

notion of the man of practical wisdom to clarify what scientific rationality is.2

In a scientific context, the man of practical wisdom is the person experienced in

deliberating on and making crucial scientific decisions which cannot be made

by invoking an algorithm. According to consensual epistemology, such a

trained scientist, and not the rules he wields, determines which scientific

theories are justified, and thus what is rational from a scientific point of view.

More accurately, Kuhn and Brown hold that the community of trained

scientists determines what is rationally acceptable. Kuhn is especially clear on

the view that there is no epistemological standard higher than the consensus of

the scientific community. He claims that the best way to decide whether one

scientific theory is more rationally acceptable than another is as follows:

Take a group of the ablest people with the most appropriate motivation; train them in some science and in the specialties relevant to the choice at

hand; imbue them with the value system, the ideology current in their

discipline (and to a great extent in other scientific fields as well); and,

finally, let them make the choice.3

This view, of course, makes scientific justification intrinsically social. It implies that we can explain the justification of scientific theories, not by an appeal to

unambiguous epistemological rules, but only by an examination of the relevant

scientific community, particularly its shared values, likes and dislikes.

The shared values of a scientific community, according to Kuhn, provide the best guidance in the community's decision whether one theory is more

rationally acceptable than another. Among these values we usually find the

following: (1) accuracy, i. e., a theory should agree with the results of

previously accepted experiments and observations; (2) consistency, i. e., a

theory should be internally consistent and consistent with other currently accepted theories; (3) broad scope, i. e., the consequences of a theory should extend beyond the particular observations it was designed to explain; (4) simplicity, i. e., a theory should bring order to otherwise unordered

phenomena; and (5) fruitfulness, i. e., a theory should reveal new phenomena and new relations among phenomena. Each of these values is, of course,

somewhat vague, and consequently scientists often disagree about how they should apply

a particular value to a

particular decision. For example, one

scientist may construe simplicity

as quantitative simplicity, whereas another

scientist may construe it as qualitative simplicity. Furthermore, scientific values often conflict with each other when one applies them together. It might be the case, for instance, that simplicity or fruitfulness guides the scientist to choose one

theory, whereas accuracy guides him to choose another. And

unfortunately we cannot reconcile all value-conflicts simply by positing 2

Brown, Perception, Theory and Commitment, p. 149. Cf. Brown, "On Being Rational," American Philosophical Quarterly 15 (October 1978), 245-46.

3 Thomas Kuhn, "Reflections on My Critics," in Criticism and the Growth of Knowledge, eds.

I. Lakatos and A. Musgrave (Cambridge: Cambridge University Press, 1970), p. 237. Cf. Kuhn, The Structure of Scientific Revolutions, 2d ed. (Chicago: University of Chicago Press, 1970),

pp. 94, 170, 200, and Brown, Perception, Theory and Commitment, p. 150.

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

156 Paul K. Moser

accuracy as the most decisive value. For often one of two competing theories is

very accurate in one area, whereas the other theory is very accurate in another

area. The oxygen theory, for example, could explain weight relations in

chemical reactions better than the phlogiston theory, but the phlogiston theory could better explain the metals' being much more alike than the ores from

which they came. Consequently, Kuhn stresses that the scientific community must decide in which area accuracy is more important.4 And the consensus of

the scientific community is important in the application of the other scientific

values also, since those values provide only a rough guide for the community's choice of the most reasonable theory. Accordingly, Kuhn emphasizes that the consensus of the scientific community provides the highest standard for the

justification of any scientific theory. But one might inquire further how Kuhn's emphasis on the significance of

scientific values to scientific justification fits with his emphasis on the

significance of community consensus to scientific justification. A natural

question in this connection is whether the scientific values in question have any

epistemological significance independently of the decision of the scientific

community to countenance those values. The proponent of a thoroughgoing consensual epistemology will, of course, oppose an affirmative answer here.

For instance, Richard Rorty has said the following about the scientific values

emphasized by Kuhn:

We would do well to abandon the notion of certain values floating free of

the educational and institutional patterns of the day. We can just say that

Galileo was creating the notion of 'scientific values' as he went along, that it was a splendid thing that he did so, and that the question of

whether he was 'rational' in doing so is out of place.5

According to consensual epistemology, then, the epistemological significance of scientific values derives wholly from a decision made within the scientific

community. Thus, on that view, scientists do not countenance certain scientific

values because it is rational to do so; rather, it is rational to countenance

certain scientific values because the community of scientists does so. Accord

ingly, the consensus of the scientific community, rather than the set of

scientific values, is epistemologically fundamental. Of course, once a scientific

community comes to countenance certain scientific values, those values will

influence the subsequent cognitive decisions of that community, but we should

not infer from this that those values have an epistemological significance

independently of the community consensus. For community consensus,

according to consensual epistemology, is basic to all epistemological assess

ment.

4 The present paragraph follows Kuhn, "Objectivity, Value Judgment, and Theory Choice," in

The Essential Tension (Chicago: University of Chicago Press, 1977), p. 323. Cf. Kuhn, The

Structure of Scientific Revolutions, 2d ed., pp. 184-86, 199. 5

Rorty, "From Epistemology to Hermeneutics," in Acta Philosophica Fennica, Vol. 30: The

Logic and Epistemology of Scientific Change, eds. I. Niiniluoto and R. Tuomela (Amsterdam: North-Holland Publishing, Co., 1978), p. 20. Kuhn's writings give no reason to believe that he

would disagree with the substance of this quote.

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

On Scientific Justification by Consensus 157

The following principle captures the basic idea of the consensual epistemol

ogy advanced by Kuhn and Brown:

A person S is justified in believing a scientific proposition P at present if

and only if at present most members of a scientific community C rele vant to S agree that P explains, in accordance with certain scientific

values accepted by C, certain problem-propositions, which according to

the members of C need explaining, better than any available competing

propositions.

According to this principle, the consensus of the scientific community plays a

decisive role at three levels. First, a community gives rise to a set of scientific

values that provides the community with a number of general epistemological

guidelines. As noted above, according to Kuhn our current scientific values

entail that an acceptable theory should be accurate, consistent, comprehensive,

simple and fruitful. Second, a consensus of the scientific community determi nes which scientific problems an acceptable theory needs to solve, and thus

which problem-propositions an acceptable theory needs to explain. Typically, such a consensus arises from scientists' participation, by

means of a common

education, in what Kuhn calls a 'disciplinary matrix'6, i. e., a complex of shared

beliefs accounting for general agreement about scientific problems. Third, a

consensus of the scientific community determines which propositions are

justified by their providing the best explanation, in accordance with the

community's scientific values, of the relevant problem-propositions. Accord

ing to consensual epistemology, then, a

community consensus underlies any

epistemological assessment in science.

Let us turn now to some problems facing this account.

II

One likely objection to consensual epistemology is that it confuses

descriptive and normative matters. Justification, it seems, is a normative

feature, and hence should not be defined in terms of something factual, such as

the consensus of the current scientific community. Obviously, the scientific

community can reach a consensus in many ways, and it is doubtful that all of these ways are

rationally acceptable. Consider, for instance, the notorious

Lysenko affair that marked the scientific community of the Soviet Union in the late 1920s. The politically forced consensus of the Lysenkoites brought about

widespread rejection of modern genetics in the Soviet Union, and this consensus held sway over Soviet biology well into the 1960s. Given the real

possibility of such political demogoguery, it seems absurd to analyze scientific

justification in terms of the consensus of the current scientific community. By way of reply, the proponent of consensual epistemology might stress

that this theory relies on the consensus of the world-wide scientific commun

6 See Kuhn, "Second Thoughts on Paradigms," in The Essential Tension, pp. 293-319. Cf.

Brown, Perception, Theory and Commitment, pp. 97-101.

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

158 PaulK. Moser

ity. Of course the world-wide scientific community did not endorse Lysenko ism, and thus one might conclude that consensual epistemology can easily avoid the objection at hand. But this reply is clearly too facile. For the crucial

objection is that it is a real possibility to have a politically forced consensus of the world-wide scientific community, but that such a consensus would be

rationally unacceptable. One natural counter-reply is that consensual epistemology does not permit

just anyone to participate in forming a rationally acceptable consensus; only trained scientists can form such a consensus.7 But this sort of reply will not go very far. For it is always a real possibility that trained scientists will choose a

theory on epistemologically inappropriate grounds, such as on the basis of financial considerations. In fact, some neo-marxist theorists have argued that this is more than a possibility at present. In any case, it is difficult to see how a

consensus formed primarily out of financial interests, for example, can settle

any epistemological issues about scientific justification.

Consequently, the consensual epistemologist might naturally propose that what is scientifically justified is determined by certain scientific values which

may or may not be exemplified in the cognitive decisions of the current so

called "scientific" community. On this proposal, certain scientific values are

essential ingredients of any genuine scientific community. Each scientific

community can nonetheless rank and apply these values in its own way, but once a

community neglects or renounces these values it ceases to be scientific.

Evidently this kind of proposal will enable the consensual epistemologist to

acknowledge the normative character of scientific justification. For it enables one to deny that just any community consensus determines scientific

justification, and to hold that a community

consensus is rationally acceptable

only if it conforms to a specified set of scientific values.

Apparently consensual epistemologists such as Kuhn and Brown would

have little objection to the foregoing proposal. But Brown complicates that

proposal by arguing that we can derive norms of scientific justification from an

examination of the practices of actual scientific communities.8 Specifically, he

holds that we can look back over the history of science, and discover what has

helped as well as what has hindered the development of science. On the basis

of this discovery, Brown claims, we can make general recommendations about

scientific procedure. We might recommend, for instance, that contemporary scientists make theory choices in accordance with the scientific values

emphasized by Kuhn. In any case, Brown holds that we can derive "ought" from "is"; that we can determine how scientists ought to make epistemological assessements of theories from an examination of how scientists have in fact

made such assessments.

But there are at least three problems facing Brown's effort to derive "ought" from "is". First, we must face the difficult problem of determining which

7 This is Brown's claim in Perception, Theory and Commitment, p. 160. Cf. Kuhn,

"Reflections on My Critics," in Criticism and the Growth of Knowledge, p. 263. 8

Bown, Perception, Theory and Commitment, pp. 157-58.

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

On Scientific Justification by Consensus 159

communities, whether past or present, qualify as genuine scientific com

munities. Evidently, we need to have at hand a list of defining characteristics, such as a list of shared scientific values, before we can profitably search history for the scientific groups providing us with the desired scientific values. But if

this is so, it appears that we must presuppose that certain values are scientific

before we can successfully examine the scientific communities from which we

intend to derive those values. Brown apparently finds this kind of circularity to

be virtuous and not vicious, on the ground that all historical inquiry requires

presuppositions. But it seems clear that the circularity here is less than virtuous. For if we must presuppose that the values we intend to derive are

scientific, and thus epistemologically significant, our purported derivation will

be futile.

A second problem facing Brown's proposal is that the history of science is

highly diversified in such a way that the various so-called "scientific"

communities do not share anything like a uniform system of scientific values.

Some scientific communities show little, if any, recognition of the five

scientific values listed above. The ancient Babylonian astronomers, for

instance, were not concerned, so far as we know, with theoretical simplicity and generality in the way that most

speculative astronomers are. Rather, they

apparently were concerned almost exclusively with practical achievements in

celestial forecasting. And without much difficulty one could multiply examp les of cases in which some group of scientists consistently neglects some of the

scientific values in question.9 Thus, we should oppose any suggestion that all of the scientific values in question are exemplified in every scientific

community, or that a

recognition of each of those values is a necessary

condition of a scientific community. At most, the recognition of only some of

those values is a necessary condition of a scientific community. Of course the consensual epistemologist might propose that due to their prominent role in the development of modern science, those values have become essential

ingredients of a contemporary scientific community. But such a proposal will lead to an additional problem.

A third problem with Brown's proposal arises from the assumption that certain kinds of behavior can help the development of science. This assumption raises two difficult questions: What is science developing toward? And how can we distinguish the helping from the hindering of scientific development?

Obviously the success of Brown's proposal to derive "ought" from "is"

depends on plausible answers to these questions. Although Brown himself has not provided the needed answers, it is clear that he would oppose answers

invoking any absolute, theory-independent kind of truth toward which science

develops. And the same holds for Kuhn also. According to both Kuhn and

9 On the ancient Babylonian astronomers see, for example, Stephen Toulmin and June

Goodfield, The Fabric of the Heavens (New York: Harper & Row, 1961), pp. 23-48. Additional

evidence for the variability of scientific values is available in Toulmin, Human Understanding, (Princeton: Princeton University Press, 1972), I, chapter 4, and Larry Laudan, Progress and Its

Problems (Berkeley: University of California Press, 1977), chapter 4.

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

160 PaulK. Moser

Brown, theory-independent truth is irrelevant to the evaluation of theories, since our theories provide

our only

access to reality.10

In short, we have no

access to theory-independent truth. Moreover, it is not clear that any other

plausible account of scientific development is readily available to the consen

sual epistemologist; at least, neither Kuhn nor Brown has provided one.

Clearly, we cannot now rely on a scientific consensus to tell us what helps the

development of science, since, after all, the basic question at hand concerns the

conditions under which a scientific consensus is rationally acceptable. Consen

sual epistemology, then, faces at least three major problems arising from

Brown's proposal to derive "ought" from "is".

Furthermore, even if the consensual epistemologist can derive certain

scientific values from an examination of the history of science, there is still the

crucial question whether a scientific consensus formed in accordance with

those values can provide epistemic justification, i. e., whether it can provide one with a good reason to believe something to be true, or at least likely to be true. As suggested above, Kuhn and Brown deny that a scientific consensus

about the rational acceptability of a theory enables one to conclude that this

theory is likely to be true in any theory-independent sense of truth. Rather, Kuhn and Brown propose that the rational acceptability of a theory lies mainly in its ability to solve certain problems the prevailing scientific community

deems important. Thus, the consensual epistemology of Kuhn and Brown is

relevant not to what I have called epistemic justification, but at most to some

kind of rational acceptability that does not entail a high probability of truth.

But even if we acknowledge a distinction between epistemic justification and

rational acceptability, the consensual epistemologist must face an additional

problem. A consensus of the scientific community, it seems, can bear on the

rational acceptability of one's beliefs only if one has good reason to believe that

there is such a consensus. That is, if the consensus of the scientific community alone can provide one with good reasons for one's scientific beliefs, and if one

has no good

reason to believe that such a consensus supports one's beliefs, then

one will have no good reasons for one's beliefs. For in that case one's beliefs

will, from one's own epistemological perspective, lack good grounds, and thus,

despite the consensus of the scientific community, one will not be justified in

holding one's beliefs. Consequently, to have justified beliefs, one must be

justified in believing something about the views of the relevant scientific

community. In short, one must be able to determine with good reason what

the relevant scientific community agrees on. But how can consensual

epistemology provide for this requirement? According to the earlier statement

of consensual epistemology, one is justified in holding some scientific belief so

long as that belief is supported by the consensus of the relevant scientific

community. But the present considerations indicate that one must first acquire

justified beliefs about what the relevant scientific community agrees on. And

typically one can acquire such beliefs only by reading the writings of the

10 See Brown, Perception, Theory and Commitment, pp. 153-54, and Kuhn, The Structure of

Scientific Revolutions, 2d ed., pp. 206-7.

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions

On Scientific Justification by Consensus 161

relevant scientists, and by considering whether their views are widely agreed upon. But this means that typically the justification of one's beliefs about a

scientific consensus, and thus of one's scientific beliefs, will ultimately arise

from evidence endependent of that provided by an external scientific consensus. Insofar as consensual epistemology neglects this consideration, it is

in need of revision.

An equally troublesome gap in the consensual epistemology of Kuhn and Brown is the absence of an account of what constitutes a consensus of the

scientific community. Must there be a unanimous decision in favor of the

acceptability of a theory? Or is a simple majority decision sufficient? Further, what if the scientific community has a split decision on the acceptability of a

particular theory? Should we then say that both the theory and its denial are

rationally acceptable? In any case, there is rarely, if ever, complete agreement among the members of a scientific community, and even if there were such

agreement, it would be very difficult to verify. But, as noted above, one's

verifying that there is some sort of scientific consensus is centrally important to the justification of one's scientific beliefs. Consequently, until the consen

sual epistemologist can provide an account of this procedure, and meet the additional objections raised above, we have good reason to conclude that consensual epistemology is inadequate as an alternative to the more traditional

empiricist epistemology underlying logical empiricist philosophy of science. In

sum, then, consensual epistemology provides at best a seriously incomplete alternative to logical empiricism.

Adresse des Autors:

Paul K. Moser, Assistant Professor of Philosophy, Dept. of Philosophy.

Loyola University of Chicago, 6525 North Sheridan Road, Chicago, Illinois 60626, USA

This content downloaded from 147.126.1.145 on Thu, 31 Dec 2015 22:48:05 UTCAll use subject to JSTOR Terms and Conditions


Recommended