BEHAVIORAL SCIENTIST QUA SCIENTIST MAKES VALUE JUDGMENTS, THE

Behavior and Philosophy, 2003 by Lacey, Hugh

ABSTRACT: I distinguish three matters about which decisions have to be made in scientific activities: (1) adoption of strategy; (2) acceptance of data, hypotheses, and theories; and (3) application of scientific knowledge. I argue that, contrary to the common view that only concerning (3) do values have a legitimate role, value judgments often play indispensable roles in connection with decisions concerning (1)-that certain values may not only be furthered by applications of the scientific knowledge gained under a strategy, but they may also provide a primary reason for conducting research under the strategy. However, this is compatible with making decisions concerning (2) that in no way draw upon values. While, in my opinion, this account applies to all the sciences, it has special salience in the behavioral and cognitive sciences. The behavioral scientist, qua scientist, makes value judgments when making decisions about which strategy to adopt, but not when deciding which theories to accept as providing knowledge and understanding of specified domains of phenomena.

Key words: value judgments, impartiality, behavioral science, materialist strategies, agroecology

John Staddon, like many authoritative spokespersons of the tradition of modern science, is drawn to endorse that "Value-laden statements should. . .be excluded from [scientific] research entirely. . ." (Staddon, 2001). Why? Apparently because he holds the objective of science to be the discovery of facts and regularities and he believes that, unless "what is fact" is differentiated from "what is value" and value judgments are kept out of scientific research, this objective may not be met and science may not "rise above the level of 'politics by other means'" (Staddon, 2001, p. iii).

There is, however, a straightforward sense in which value judgments cannot be kept out of scientific research. Theories (hypotheses) must be appraised for their epistemic value-the kind of value they have in virtue of their being bearers of knowledge and understanding-in the light of how well they satisfy the appropriate criteria (Scriven, 1974): Is a theory well supported by the evidence? Has it been submitted to appropriately rigorous experimental tests involving sufficiently robust empirical data? Does it display adequate predictive power, a sufficient degree of empirical adequacy, and appropriate explanatory scope? Answering these questions involves making epistemic (sometimes referred to as "cognitive") value judgments (Lacey, 1999a, Ch. 3, 2002d), and there is no science without them. Clearly it is not epistemic value judgments that Staddon wants to keep out of scientific research, but moral, social, political (and other nonepistemic) value judgments and statements laden with these kinds of values. (Below when I use "value" without qualification I will be referring to nonepistemic value.)

In this article, in response to Staddon's statement quoted above, I will sketch an argument that value judgments may play an indispensable role in scientific research. When they do so, however, it is only at specific moments, which does not include the moment at which factual judgments are settled upon. I have developed the argument in detail in several recent works in which 1 address primarily the natural sciences, most comprehensively in Is Science Value Free? (Lacey, 1999a; see also 2002a, 2002d); the reader is referred to them for details.

Where Value Judgments May Play a Role in Research

Ignoring some nuances, Staddon and I agree that:

1. There is a distinction between fact and value and between epistemic value judgments and other kinds (moral, social, political, etc.) of value judgments (Lacey, 1999b).

2. The principal kinds of factual judgments made in the sciences-(a) that a datum is a consolidated experimental result, (b) that a hypothesis is well supported empirically, and (c) that a theory identifies the possibilities that may be realized within a specified domain of phenomena (or, more narrowly, that it is successful in generating predictions)-are not, or ought not be, based in any way on value judgments.

But, in apparent disagreement with him, I hold:

3. There are moments in research practices in which values may play a legitimate and sometimes an indispensable role.

And so I reject:

4. Value judgments and value-laden statements should be excluded entirely from scientific research.

I emphasize that the legitimate role of values is confined to the right moments. I endorse neither the indiscriminate play of values in science nor currently fashionable views that scientific judgments are fully explicable in social terms and that sound science cannot be separated from ideology, although sometimes distinguished scientists put forward as sound science what is in fact ideology (Lewontin, 1992; cf. Dawkins, 1999).

My disagreement with Staddon might not be very great. he reluctantly concedes that there might be fields from which values cannot be excluded; then, qualifying the passage quoted above, he says "Value-laden statements should be . . .(lest we abolish much of social psychology) stated in a responsible way" (Staddon, 2001, p. iii). Perhaps item 3 can be interpreted as filling in what "a responsible way" might be. Nevertheless, where Staddon permits responsibly stated value-laden statements as a concession to social psychology, I hold that their role is pervasive and fundamental in all the sciences.