The ways in which the sciences are and are not value free

The paper below is a later version.

Presented at the conference, "Value Free Science:

Illusion or Ideal?", Center for Ethics and Values in the

Sciences, University of Alabama at Birmingham,

February 23-25, 2001.

Impartiality, Neutrality and Autonomy:

Three components of the idea that science is value free [1]

Hugh Lacey

Swarthmore College

ABSTRACT:

The idea that science is value free is well captured by the thesis: Impartiality, neutrality and autonomy are, and ought to be, constitutive values of scientific practices and institutions. Impartiality affirms that theories should be accepted in the light of criteria that are completely distinct from commitments with respect to social and moral values; neutrality that applications of established theories do not especially favor any particular value perspectives; and autonomy that social and moral values have no place within scientific methodology, and that science should be conducted free from interference derived from value commitments. These three ideas, and their presuppositions, are analyzed in detail - enabling us to explore fruitfully the ways in which the sciences are and are not value free.

I identify three distinct ideas: impartiality, neutrality and autonomy, that are in play in discussions of science as value free. Provisionally, (1) science is impartial: there is no proper role for moral, social and any other non-cognitive values, alongside the cognitive (or epistemic) values, in the appraisal of the soundness of scientific understanding;[2] (2) well conceived scientific practices produce a body of understanding that is neutral among contending value-outlooks: neither cognitively nor in practical applications do they favor the claims or interests of any particular value-outlooks; and (3) its research practices are autonomous: their methodologies should be unencumbered by political, religious and other non-cognitive interests. I submit that the view of science as value free, that is integral to the self-understanding of the modern scientific tradition, is well captured by the thesis:

Impartiality, neutrality and autonomy are, and ought to be, constitutive values of scientific practices and institutions.

Note that this thesis is consistent with moral and social values playing many roles within science, e.g., in motivating research interests, in setting the direction and determining the legitimation of applied research and applications, even in making judgements about the adequacy of available evidence and assessing the testimony of scientific experts — and in many other ways.[3]Proponent of science as value free have never denied this. They hold only that at its core — where theories are accepted and research directions of basic research are set — moral and social values (unlike cognitive values) have no proper role, and that no viable value-outlook should be especially favored by the advances of scientific understanding. Too often, science is value free has been dismissed by pointing to the legitimate play of values at the periphery rather than at the core of scientific practices.

According to my thesis science as value free represents a value of scientific practices and institutions. A particular account of the nature of values underlies my argument, which unfortunately there is no time to develop in this paper. I believe that the discussion of science as value free has too often been marred by uncritical endorsement of non-cognitive accounts of values. I have proposed an account in which values of various kinds (personal, moral, social, etc) are held together in value-outlooks.[4] These are complex sets of values, rendered coherent, ordered and rationally worthy of being held by certain presuppositions about human nature (and nature) and about what is possible. Since suppositions about human nature and the possibilities of things are open to some measure of empirical investigation, it follows that the outcomes of scientific inquiry may contribute to support or undermine the rational credentials of a value-outlook. I will say that a value-outlook is viable provided that its presuppositions are consistent with soundly accepted scientific knowledge. Viability is a necessary condition for the rational adoption of a value-outlook. Not all value-outlooks that have been entertained are viable, but the advance of science leaves open a range of viable ones. That is why it is coherent to see scientific developments as having played a significant rational role in (e.g.) the demise of the value-outlook of medieval Christendom, while also maintaining that science is impotent to adjudicate the great moral disputes of our age[5] — and it perhaps explains why normally no incoherence is recognized when the modern scientific tradition endorses both science as value free and the primacy of progress.[6]

* * * * * * *

I will now explicate the three component values of science as value free and their respective presuppositions.[7]

Impartiality

Impartiality presupposes that cognitive and other kinds of values can be distinguished.[8]

It designates the value that:

A theory is accepted of a domain of phenomena if and only if it manifests the cognitive values to a suitably high degree (according to the highest standards) in the light of available empirical data; and a theory is rejected if and only if a theory inconsistent with it has been soundly accepted.

Accordingly there is no proper role for moral and social values (and the practical uses to which theories may be put, and by whom) in the judgments involved in choosing theories; and neither is there a role for religious dogma, political ideology or metaphysical commitments.

A theory may be soundly accepted in accord with impartiality and at the same time be an object of social value (e.g., be useful on application for particular social interests). Impartiality forbids a role to social values only in grounding the judgments involved in choosing theories, not in making judgments of their significance for a value-outlook on application, where a theory is significant for a value-outlook if it may be applied so as to further the manifestation of (some) component values of the outlook, without (on balance) undermining the outlook as a whole.[9]

Impartiality is presented as a value of scientific practices and institutions, even though it is well known that in actual fact numerous theories have been accepted in violation of it. Nevertheless, such lapses (of great importance to "anti-science" criticisms) are consistent with impartiality being genuinely held as a value in scientific practices provided that there are plenty of exemplary cases that do manifest it highly, that attempts are made to identify the mechanisms that can cause violations of it and steps are taken to prevent their actual operation,[10] and that the trajectory of scientific practices points to more and a greater variety of theories being accepted in accord with it.

Neutrality

Neutrality presupposes first impartiality, second that scientific theories do not entail that any particular values should be adopted, and third that the body of soundly accepted scientific theories leaves open a range of viable value-outlooks.

Then, neutrality designates the value that:

Each viable value-outlook is such that there are soundly accepted theories that can be significant to some extent for it;

and applications of soundly accepted scientific theories can be made evenhandedly, so that overall there is no viable value-outlook for which the body of theories should have favored significance.

Neutrality expresses that scientific research provides, as it were, a menu of soundly accepted theories, among the items of which (in principle) each value-outlook may have its tastes for application (good or bad) catered to.[11] Note that the significance of a theory for a value-outlook cannot be inferred from its being acceptance in accordance with impartiality.

Like impartiality, neutrality can be maintained as a value of scientific practices despite actual lapses, provided that conditions parallel to those I listed for impartiality are in place. Within a wide range of mainstream scientific practices the trajectory is indeed in the direction of the higher manifestation of impartiality — and it is easy to point to exemplary cases of theories that are soundly accepted of certain domains. But, in the case of those same practices, a similar trajectory towards higher manifestation of neutrality is not discernible with the same ease; and, within them, I do not think this can be reversed. Rather than being applied in an evenhanded way, the body of soundly accepted theories of modern science tends overwhelmingly (and often unabashedly) to be significant for value-outlooks that nourish technological progress and "modernization" [that contain what I call below "the modern valuation of control"], so that the actual practices of application provide little evidence that the theories of modern science can be significant to a comparable extent for many other viable valueoutlooks that contest modernization, especially in its current form of "globalization".[12]

Autonomy

Autonomy presupposes first that there is a reasonably clear distinction between basic and applied scientific research, and second that the practices of basic research aim to bring about higher and more widespread manifestations of impartiality and neutrality.[13]

I express it as the value that:

The characteristics of scientific methodology, the adoption of strategies in research, and the priorities and direction of basic research are set, without "outside" interference, by cognitive interests alone.

I think that it is clear that the manifestation of this value to a reasonable extent in scientific practices rests upon a further presupposition that may best be stated as a complementary value:

Basic research practices are conducted in autonomous communities, supported by autonomous institutions — i.e., communities and institutions whose priorities are set without interference from "outside" interests, powers and values (political, religious, economic, personal); consequently the priorities and directions of basic research are set so that the interest, to heighten the manifestation of the cognitive values in the theories of the domains under investigation and to extend research into domains not currently within their compass, is not subordinated to other interests.[14]

* * * * * * *

Now I will offer my appraisal of the components of the thesis with which I have expressed science as value free.

I begin with autonomy. Can the ideal that basic scientific research effectively be "driven by" the cognitive values alone, play a regulatory role in scientific practice? Autonomy might appear to follow from accepting that the aim of science is "to gain understanding of the world — (or) to generate and consolidate theories manifesting the cognitive values highly, and progressively of more and more domains of phenomena and possibilities".

Appealing to the aim of science stated like this, however, can provide no direction to scientific investigation, since it does not — for any domain — point to the relevant kinds of empirical data to procure and the appropriate descriptive categories to use for making observational reports, and to the kinds of theories to posit so that they can be put them into contact with the data. In order to pursue such an aim the "right" kinds of data and theories must be brought into contact, so much so that (logically) antecedent to engaging in inquiry (what I call) a strategy must be adopted;[15] where the key roles of a strategy are to constrain the kinds of theories that may be entertained — and thus to specify the kinds of possibilities that may be explored in the course of the inquiry — and to select the kinds of empirical data that acceptable theories should fit. Without adopting a strategy we cannot address coherently and systematically: what questions to pose, what puzzles to resolve, what classes of possibilities to attempt to identify, what kinds of explanations to explore, what phenomena to observe, measure and experiment upon, what procedures to use.

My notion of strategy has affinities with Hacking's notion of "form of knowledge." In each case questions can be asked about its adoption as a frame for research (logically) prior to the appraisal of any theories developed within the frame. Hacking says:

By a form of a branch of scientific knowledge I mean a structured set of declarative sentences that stand for possibilities, that is, sentences that can be true or false, together with techniques for finding out which ones are true and which ones false. … A form of knowledge represents what is held to be thinkable, to be possible, at some point in time.[16]

In my language, a strategy identifies the general features of the range of possibilities that may be explored within a research program; and, if its claims become hegemonic in a society, that will set limits to the "thinkable" in that society, so that questions that make sense in one frame may be unintelligible in another.[17] Hacking's and my emphases are different: he emphasizes how a form of knowledge may be altered by radically new inventions),[18] I that commitment to conflicting value-outlooks may lead to the adoption of different strategies, though Hacking is sensitive to the contingency of the adoption of a form of knowledge and the role of social factors in explaining their adoption.

Most of modern science tends to adopt virtually exclusively various forms of (what I call) materialist strategies:[19] theories are constrained to those that represent phenomena and encapsulate possibilities (the abstracted possibilities of things) in terms of their being generated (or generable) from underlying structure, process, interaction and law, abstracting from any place they may have in relation to social arrangements, human lives and experience, from any link with value, and from whatever social, human and ecological possibilities that may also be open to them.[20] Reciprocally, empirical data are selected, not only to meet the condition of intersubjectivity, but also so that their descriptive categories are generally quantitative, applicable in virtue of measurement, instrumental and experimental operations.

What are the rational grounds on which the modern scientific community has adopted materialist strategies almost exclusively? [Why, looking ahead, has it been more interested in exploring the possibilities of agro-biotechnology than of agroecology?] I pose this question because I am assuming that in order to explore important classes of possibilities the use of materialist strategies is indispensable, that there is in general no good reason to think that the possibilities of things are exhausted by their abstracted possibilities, and that there are forms of systematic empirical inquiry in which "non-abstracted" possibilities can be investigated. I will — perhaps contentiously — call any form of systematic empirical inquiry a form of "science."[21]

One common answer to the question draws upon materialist metaphysics:[22] Science aims to understand the world as it is — the material world — independently of its relations with human beings. Materialist strategies (and only them) provide categories appropriate to this aim. A secondanswer can be drawn from Kuhn:[23] Not the nature of "the material world," but the current historically-contingent stage of our research practices, demands the adoption of materialist strategies. Adopted in the first place (rationally) because they helped to solve puzzles that had remained anomalous under old strategies, they have continued to predominate because of their fruitfulness: under them the range of theories that have become soundly accepted is large and variegated and it continues to become more so. That is sufficient for the current privilege of materialist strategies for, according to Kuhn, the historical practice of science proceeds best when the scientific community pursues a strategy single-mindedly until its potential is exhausted.

Having criticized these two answers elsewhere.[24] I will proceed directly to a third, the one I take to be most compelling.[25] This answer draws (rationally and not just causally) from a mutually reinforcing interaction between research conducted under materialist strategies and commitment to a set of widely held and embodied social values — modern technological and economic values, the values of "modernization"; I have in mind particularly (what I call) the modern valuation of control, the distinctive role played by control of natural objects and the way in which it is valued in modernity: its scope, centrality in daily life, relative insubordination to other social values, and the deep sense that control is the characteristic human stance towards natural objects, so that the expansion of technologies into more and more spheres of life and into becoming the means for the solution of more and more problems is highly valued.

[Passage taken from Lacey (forthcoming).]It is often assumed that any value-outlook rationally held today must include certain values connected with the control of natural objects and with technological advance, values that together constitute (what I call) the modern valuation of control. Thus, e.g., those who cite the value of organic farming over that of the agricultural practices spawned by recent innovations of genetic engineering tend to be dismissed on the ground that they run counter to the trajectory being shaped by the modern valuation of control ("globalization"). But that dismissal does not follow from currently soundly accepted scientific knowledge. That a value-outlook rests uneasily with the current centers of power implies neither that adopting it violates canons of rationality, nor that it is rendered non-viable.