CEPA eprint 3924

Some principles of postpositivist philosophy of science

Garrison J. W. (1986) Some principles of postpositivist philosophy of science. Educational Researcher 15(9): 12–18. Available at http://cepa.info/3924
Table of Contents
The Problem of Confirmation
The Underdetermination of Theory by Logic
The Underdetermination of Theory by Experience
The Quine-Duhem Thesis
The Theory-Ladenness of Experience
Science, Deliberation, and Practical Wisdom
In recent years there has been a great deal of methodological debate among educational researchers, theoreticians, and practitioners concerning issues such as relativism raised by the so-called “new,” “Kuhnian” or “postpositivistic” philosophy of science. The intensity of this debate notwithstanding, the fundamental principles and their relations that comprise the postpositivistic view have not always been carefully spelled out. Some of the principles discussed will include (a) the problem of confirmation, (b) the underdetermination of theory by logic, (c) the underdetermination of theory by experience, (d) the Quine-Duhem thesis, (e) the theoryladenness of experience, and (f) the incommensurability of theories. No attempt will be made to evaluate these principles. However, those who are prepared to accept all of these will be hard pressed to avoid the dangers of relativism. I will argue that these dangers, if they exist, may be lessened if not eliminated by practicing the pragmatic virtues of epistemological conservatism and good sense.
It is widely held that positivism is dead and that we now live in a post- positivistic era of educational research, although exactly what it means to pronounce this obituary is far from clear. D. C. Phillips has helped clarify the situation by distinguishing four kinds of positivism and arguing that not all of them need to be entirely rejected. The first kind of positivism, that of Mill, Comte, and the 19th century, is dead and has been for some time. The second is logical positivism. A third form, at least in the popular mind, is behaviorism. As Phillips correctly notes, behaviorism does have much in common with logical positivism, not the least of which is a mutual reliance on reduction sentences, that is, operational defini‑tions. The fourth kind is positivism as a synonym for empiricism. With the possible exception of Paul Feyerabend (1969), it is unlikely that anyone would want to suggest empiricism is dead, or even mortal. Phillips (1983) writes:
Some of the four positions that have been identified as receiving the label “positivism” have died, but others are still alive in one form or another. And because each of the four are complex positions, even the ones that have died have not departed “holus bolus” – bits and pieces of them have managed to escape the Grim Reaper. The true picture (always stranger than fiction) is quite complex. (p. 6)
In this paper I will restrict my attention to logical positivism.
In the demise of positivism, the role of “Grim Reaper” has been played by the so-called “new,” “Kuhnian,” or simply “postpositivistic” philosophy of science. This new philosophy of science, like the positivism it is supposed to have slain, is far from monolithic. The swath cut by the scythe of postposi tivism may be as narrow as the rejection of strict verificationism or as broad as the acceptance of relativism, depending on how many of a number of overlapping, but somewhat distinct principles of this new philosophy of science one is prepared to accept. These principles will be discussed in order.
No attempt will be made to evaluate the principles of postpositivistic philosophy of science; that task will be left to the reader. Those, however, who are prepared to accept all of these principles will find it difficult to avoid the dangers of relativism. I want to suggest that these dangers, if they exist at all, may be lessened if not eliminated by practicing the pragmatic virtues of epistemological conservatism and good sense. These pragmatic virtues are characteristic of those possessed of practical wisdom. I will recommend in the conclusion that in the pursuit of wisdom, the opinions of educational researchers, as well as those of educational practitioners, should be considered.
The Problem of Confirmation
The first principle to severely challenge the logical positivist program arises out of a very simple but intractable problem in the process of confirming scientific theories. Confirmation of anything from grand unified theories to the humblest hypothesis consists in drawing a logical implication from some hypothesis (H) or theory (T) to some empirical-experimental conclusion (E) that says something like “If T (or H) is true then E will be observed.” If E is, as a matter of empirical fact, observed, the researchers would like to claim T is confirmed or verified. Schematically, this pattern of argument may be represented in propositional logic this way:
\begin{matrix} P_{1} & T \to E \\ P_{2} & E \\ \hline C & T \end{matrix}
In research the logical sign “→” is usually meant to convey some necessary connection, for instance, causality. The problem is that this schema is logically invalid, representing what is commonly called “the fallacy of affirming the consequent.” Premises one and two (P1 and P2) may be true and the conclusion (C) nonetheless false. This means that even a well-confirmed theory may still prove to be fallacious. In the sciences, the fallacy of affirming the consequent gives rise to the problem of confirmation or verification that is merely the modern logical form of the 17th century philosopher David Hume’s psychological problem of induction.[Note 1]
To circumvent this problem, positivists have developed a number of powerful statistical methods and techniques. Nonetheless, as we all know from our first course in statistics, correlation does not imply causation. There are not, nor will there ever be, absolutely certain and mechanical inductive techniques. This inevitable uncertainty cuts deep into the flesh of daily existence. As W. V. O. Quine (1969) has observed, “the Humean predicament [of induction] is the human predicament” (p. 72). There are no technical substitutes for wisdom, the ability to draw adequate conclusions from inadequate evidence.
Sir Karl Popper was the first to fully appreciate the intractability of the problem of induction and reject verificationism altogether. Popper contends that “There is neither a psychological nor a logical induction. Only the falsity of the theory can be inferred from empirical evidence, and this inference is a purely deductive one” (Popper, 1968, pp. 54-55). Formal logic certainly bears out the latter half of this claim. Consider the following logical schema for refutation:
\begin{matrix} P_{1} & T \to E \\ P_{2} & \text{ not } E \\ \hline C & \text{ not } T \end{matrix}
Unlike the schema for confirmation, this form of deductive reasoning, known as modus tollens, is logically valid. Summing up, Popper declares in effect that all reasoning is either deductive or defective, and that the growth of science consists in a continuing series of conjectures and refutations. Popper (1968) writes: “ … there is no more rational procedure than the method of trial and error – of conjecture and refutation: of boldly proposing theories; of trying our best to show that these are erroneous; and of accepting them tentatively if our critical efforts are unsuccessful” (p. 51). Thus it appears that the method of conjectures and refutations secures the growth of science upon the firmest of logical foundations.
The Underdetermination of Theory by Logic
It is likely that Phillips, for his part, would want to stop the rejection of positivism about here, with the rejection of strict verificationism. Perhaps he is right to do so; certainly such a move would pull us up far short of the perilous cliffs of epistemological relativism, but there are at least plausible reasons to go further. We begin with an aside by noting that it is impossible to refute any claim that something exists, say unicorns, electrons, or cognitive structures.[Note 2] Those who assert the existence of particular entities can always say that researchers have simply failed to detect the object for one reason or another. Theories need individual entities as much as they need universal laws and generalizations. Al- though it is not exactly the problem in the present case, it is important to note that modus tollens arguments only bear directly on the universal statements of a theory.
Theories are internally complex; if we like, they may be viewed as a conjunction of statements. A theory comprised of n number of statements could be written as T = (S1 and S2 and … and Sn). Now let us introduce a simple logical equivalence known as De Morgan’s law. This law says that the logical expression [not (A and B)] is logically equivalent (written ≣) to (not A or not B). Let us look again at the schema of refutation:
\begin{matrix} P_{1} & (S_{1} \text{ and } S_{2} \text{ and }\dotsc S_{n}) \to E \\ P_{2} & \text{ not } E \\ \hline C & \text{ not } (S_{1} \text{ and }S_{2} \text{ and }\dotsc S_{n}) \end{matrix}
Using De Morgan’s law this conclusion becomes (not S1 or not S2 or … or not Sn).
The point of all this logic is that no single experiment or set of experiments, crucial or otherwise, refutes an entire theory, but only one or a few interrelated statements, quite possibly leaving the core of the theory intact.[Note 3] There are many ways of using this basic schema to deflect a refutation away from statements especially cherished by the research tradition. Nor, as it might first appear, is this a bad idea. Epistemological conservatism is good policy, whatever one’s political persuasion. As a rule, it is not wise to quickly abandon those principles that have worked well in the past just because they don’t happen to work well just this once. It is better to adjust the theory elsewhere and in such a way as to least disturb our most basic beliefs. The conservative response is not logical in the narrow sense; it is epistemological. Narrow logic is what generated the predicament in the first place.
Epistemological conservatism is pragmatic; it follows the common sense expedient that declares that it is reasonable to preserve what has worked well in the past. Should refutations begin to multiply, however, the good council of conservatism begins to lose its force and more liberal alternatives may begin, by right, to demand our consideration. One particularly interesting variant of this conservative response is the use of ad hoc hypotheses.
One of the common criticisms brought against functional explanations in the ethnographic literature is that they are frequently ad hoc in nature. Yet, as the following historical example indicates, ad hoc explanations can provide interesting hypotheses for testing and possible verification. Astronomers in the 19th century noted that the perturbations of the planet Uranus seemed to refute Newtonian celestial mechanics. The French and English astronomers Leverrier and Adams independently proposed, ad hoc, the existence of a planet beyond the orbit of Uranus to account for the observed perturbations. This was a conservative response in accord with well- established theories. With the discovery of Neptune, an apparent refutation was turned into a glorious theoretical confirmation. Later, when the problem of Mercury’s perihelion arose, Leverrier postulated the existence of another planet within the orbit of Mercury that he named Vulcan. When observations failed to detect this “new” planet, Leverrier declared it must always be on the other side of the sun from the Earth and therefore unobservable. Vulcan’s existence has never been confirmed.
Solving the problem of the perihelion of Mercury became one of the first confirmations of the new relativistic mechanics. The value of an ad hoc hypothesis in celestial mechanics or educational ethnography cannot be adjudicated in advance of experience. Only those ad hoc hypotheses that postulate unobservable consequences deserve to be rejected out of hand. Which of those ad hoc hypotheses that have empirical consequences we may wish to pursue is, like conservatism, a pragmatic issue; that is, given the physical constraints of theory and the fiscal constraints of funding, which hypotheses seem most practically useful to pursue? We should, however, be conservative in advancing ad hoc explanations and hypotheses.
The Underdetermination of Theory by Experience
In addition to the underdetermination of theory by logic, there is a serious underdetermination of theory by experience, or experiment, as well. One way of expressing this underdetermination is to say that given any finite body of data, an infinite number of theories may be tailored to fit the body. Consider the following oversimplified example. Suppose we are given some imaginary collection of data relating some independent variable, x and some dependent variable, y. Let the “theory” sought to account for this data be comprised of a single statement, some mathematical function that fits all the data points (observations) in Figure 1. The reader is probably thinking the choice must be a linear equation (y = mx + b), yet an infinite number of equations could be constructed to fit the same points. In the present case the choice of equation would probably be compelled by aesthetic considerations, a preference for elegance and simplicity; but what if the data have been plotted in polar coordinates instead? Should theory choice be relative to the background framework in this way? Is it? These complex and difficult issues cannot be taken up here, much less decided.
One way of finessing problems such as this is to make further observations that provide additional data points that may then be interpolated between previously existing data points. Any lines unable to account for these new points would be eliminated until, presumably, we were left with the correct theory. While extremely useful in research practice, it nonetheless remains the case that at the end of this process an infinite number of one-statement theories could be constructed to fit the data base so long as the base remained finite. The choice of the “best” theory, or explanation, remains underdetermined.
The response to empirical under- determination is as pragmatic as the response to logical underdetermination. Simplicity, for instance, is neither an empirical nor a logical constraint; it can only be justified pragmatically. There are also conservative constraints. Usually when we interpret data, we do so against the background of a preexisting and unquestioned theory or, in the case of theory construction, some background paradigm, research tradition, or set of guiding assumptions. Indeed, as a rule the problem is finding something to fit the data given the background constraints. Oftentimes it is only when these determining pragmatic constraints are lifted that underdetermination becomes a problem at all.
Figure 1: An imaginary plot of data.
The Quine-Duhem Thesis
Such difficulties as the underdetermination of theory by logic and experience combine to yield the next principle of postpositivistic philosophy of science, the so-called “Quine-Duhem” thesis. Briefly stated, this thesis says that no scientific proposition, including explanatory hypotheses, can be tested in isolation. On this view, not only the entire theory, but scientific methodology, mathematics, logic, in fine the whole of science, or “The Web of Belief,” as W. V. O. Quine sometimes calls it, must face the “tribunal of sensory experience” together. Quine (1961) expresses the confrontation between science and experience this way:
… total science is like a field of force whose boundary conditions are experience. A conflict with experience at the periphery occasions readjustments in the interior of the field. Truth values have to be redistributed over some of our statements. Reevaluation of some statements entails reevaluation of others, because of their logical interconnections …. Having reevaluated one statement we must reevaluate some others, which may be statements logically connected with the first…. But the total field is so under-determined by its boundary conditions, experience, that there is much latitude of choice as to what statements to reevaluate in the light of any single contrary experience. No particular experiences are linked with any particular statements in the interior of the field, except indirectly through considerations of equilibrium affecting the field as a whole. (pp. 42-43)
This holistic view of science, if true, seriously dilutes the impact of the falsificationist program. Furthermore, with the choice of which statements to reevaluate left underdetermined by logic and experience, the way seems open for many adventitious, nonscientific, and possibly irrational elements to influence scientific decisions.
In general, those who are impressed with the Quine-Duhem thesis resort to “inference to best explanation” by means of such pragmatic values as scope, coherence, testability, fruitfulness, and yes, simplicity and conservatism, to restore rationality to the process of interpreting data and constructing theories.[Note 4] Such pragmatic considerations as these blend the practical judgment of the research practitioner with theoretical judgment – the context of discovery with the context of justification.[Note 5] Needless to say, such a pragmatic “solution” to the problem of underdetermination and holism is highly controversial. I will return to this controversy in the conclusion.
The Theory-Ladenness of Experience
The apparent failure of logic and experience to dictate theory justification and choice threatens the entire logical positivist program. The principle that we are about to introduce goes even further by calling into question the very “objectivity” of experience. N. R. Hanson (1958) coined the term “theory-ladenness” to express the dependency of observation on theory for its meaningful interpretation. Since it is the conceptual structure of theories that weaves the raw fabric of observations into meaningful tapestries, allow us to use the term concept-ladenness interchangeably with Hanson’s theory-ladenness. Concept-ladenness contends that there are no concept (or theory) independent facts, observations, or 0-reports, as the empirical base of science is sometimes called. Unlike the underdetermination of theory by data, which at least preserves the sanctity of the facts, theory-ladenness challenges the very meaningfulness of facts independent of theory.
Concepts organize and structure otherwise disconnected features of perceptual experience into meaningful perception. The distinction between perception and meaningful perception was drawn by the philosopher Ludwig Wittgenstein, among others. Wittgenstein (1969), referring to the famous gestalt figure of the duck-rabbit, distinguished between merely “seeing,” perception, and “seeing as,” where “ ‘Seeing as…’ is not part of perception. And for that reason it is like seeing and again unlike” (p. 197). What exactly is the difference? Wittgenstein answers hesitantly: “Is it a case of both seeing and thinking? or an amalgam of the two, as I should almost like to say” (p. 197).
Hanson (1958), following up on this idea, contends that “seeing as,” or “seeing that,” as he sometimes calls it, “threads knowledge into our seeing; it saves us from re-identifying everything that meets our eyes” (p. 20). Referring to the same gestalt figure as Wittgenstein, Hanson writes: “But we can see the illustration [of the duck-rabbit] now as one thing, now as another. So we interpret it, and see it as we interpret it” (p. 14). In science, theories are the means of interpretation, of “seeing as.” Many of these same insights are well-known to cognitive psychologists, although they have not always appreciated their significance for epistemology.
For the practicing scientist the empirical world is the world of experiments. These experiments are constructed in accordance with, and interpreted by, the concepts of the theory presupposed by the researcher. Even if the theory is not explicitly stated, there are concepts present in research that make atheoretical (or aconceptual) inquiry impossible according to this view. Most philosophers who accept the concept-ladenness of experimental research acknowledge that experience is not totally structured by theory, but rather is at least partially pre-structured independently of the researcher’s conceptual framework. The point is that experience, while not totally passive, is not sufficient to determine any single theory uniquely.
The dependency of experimental facts on the theory that structures them and gives them interpretive meaning is dramatically illustrated by what happened to the carefully conducted research of that portion of 19th century chemistry that proceeded outside of the Proutian atomistic hypothesis. The following comment by the Noble Laureate Frederick Soddy, himself a student of Rutherford, is telling:
There is something surely akin to if not transcending tragedy in the fate that has overtaken the life work of that distinguished galaxy of nineteenth-century chemists, rightly revered by their contemporaries as representing the crown and perfection of accurate scientific measurement. Their hard won results, for the moment at least, appear as of little interest and significance as the determination of the average weight of a collection of bottles, some of them full and some of them more or less empty. (cited in Garrison & Macmillan, 1984, p. 271; see also Soddy, 1932)[Note 6]
A similar fate befell the experimental results of the early structural psychologists such as Wilhelm Wundt who, rather ironically, proceeded inside an atomistic psychological hypothesis. While Wundt succeeded in his attempt to establish “a new domain of science,” his introspective experimental analyses of immediate consciousness into its individual elements appears to be of little more than historical interest.
When the theory that structures and gives meaning to experiments collapses, an entire scientific world, it seems, collapses with it. In the present case the eventual success of the atomic hypothesis in the physical sciences and the success of gestalt and behavioral psychology rendered the efforts of the opposing camps apparently meaningless. Imre Lakatos (1970) sums things up this way: “When an ‘observational’ or ‘interpretive’ theory finally gets eliminated, the ‘precise’ measurements carried out within the discarded framework may look – with hindsight – rather foolish” (p. 140).
One of the major tenets of the logical positivist program was the claim that there is a sharp distinction between facts and the theories we construct on them. The holistic view expressed by the Quine-Duhem thesis weakens the fact- theory distinction, and theoryladenness, if true, annihilates it. If the notions of theory and conceptladenness are correct then all sciences are, to some degree, interpretive and hermeneutical, all observation participant observation.
The most disturbing result of theory-ladenness is that it raises the possibility of finding no neutral empirical ground of shared facts on which to judge between competing theories. This leads, in part, to the last principle of postpositivistic philosophy of science that I wish to consider, the incommensurability of theories, or paradigms.
The incommensurability of theories, as the phrase itself suggests, means that no common ground or proportion can be found between two (or more) competing theories. Half of this incommensurability is due, as we have just indicated, to the lack of a theory neutral commensurate body of facts that may be used to construct crucial experiments. The other half of incommensurability is linguistic. This occurs when proponents of competing theories or paradigms do not share the same concepts, or use similar sounding concepts with significantly different meanings, with the result that the different camps fail to make logical contact and end up “talking past” one another.
For example, the concept “mass” does not mean the same thing in classical mechanics as it does in relativistic mechanics. The difference is significant. For classical mechanics there was no proportion between mass and energy, thus the expression E = mc2 would be a grammatical but nonsensical statement. Similarly, for a Ptolemaic astronomer, the sun and moon along with Mercury, Venus, Jupiter, and so on, were all instances of the concept “planet,” whereas the earth was not. For the Copernican, the sun and moon were not planets at all and the earth was. Research traditions may as easily talk past each other using methodological concepts as theoretical concepts. Elliot Eisner (1981), for instance, has shown that scientific and artistic qualitative researchers may agree on the “phenomena of study,” but not on “the mode in which the study occurs” (p. 5). In this case the shared term “qualitative,” according to Eisner, “not only hides the important distinctions which need to be made, but it is, itself, misleading” (p. 5). Between very different theories and/or methodologies it is often hard to tell what is what, empirically or conceptually.
This incommensurability, along with much of the rest of postpositivistic philosophy of science, suggests that neither logic nor experience, even in combination, is enough to guarantee the totally objective evaluation of theories and hypotheses. Incommensurability has opened the door to relativism, irrationalism, and various kinds of scientific nihilism. It is perhaps wise to keep in mind that although radical incommensurability might be a possibility, it rarely if ever actually occurs. If two theories were that incommensurable it would be hard to even imagine them as competing, although the current debate between the theories of creationism and evolution come close. Further, we should ask ourselves if the traditional methods of justification really have failed us. Harvey Siegel (1980a, 1980b, 1982, 1986), Newton- Smith (1981) and D. C. Phillips (1983) among others, think not, and offer valuable suggestions as to how excess may be avoided. Nonetheless, Phillips is not quite correct to conclude that “the theses of relativism, incommensurability, holism of meaning … have been shown to be dubious at best, and it would seem that the weight of scholarship now is flowing against Kuhn” (p. 11). The target here is Thomas Kuhn’s (1962) The Structure of Scientific Revolutions, the book that might be fairly said to have inaugurated the “new” or postpositivistic philosophy of science.[Note 7]
Science, Deliberation, and Practical Wisdom
I would like to conclude by suggesting two closely related responses to the threat of relativism that seem so far to have been overlooked. Both responses depend on introducing a normative component into the debate at two distinct levels. I would like to preface my suggestions by remarking that we should not lose sight of the fact that there remain rational things to do even when strict logic fails us.
Earlier we pointed out that the Quine-Duhem thesis, along with the theory-ladenness of data, if correct, combine to annihilate the fact- theory distinction. The rejection of this distinction clears the way for the rejection of an even more cherished distinction. It was Richard Rudner (1953), a prominent philosopher of the social sciences, who first realized that Quine’s holistic arguments could be made to do double duty. Not only does the fact-theory distinction collapse if we accept Quine’s arguments, but so too does the fact-value distinction. Rudner’s argument, briefly stated, is that the acceptance or rejection of a theory or hypothesis is a value decision that must be made by the scientist since the choice is left underdetermined by either experience or logic. Rudner concludes that “the statement that scientists qua scientists make value judgments, is also a consequence of Quine’s position” (p. 6).
More recently, Kenneth R. Howe (1985), in a paper whose very title, “Two Dogmas of Educational Research,” indicates its Quinean origin, has argued for the rejection of “the fact-value dogma” in education. Indeed, Howe contends, “social research is doubly value- laden … [since] the very concepts social researchers employ are evaluative of human behavior” (p. 12). What this means is that if we accept all of what post-positivistic philosophy of science would have us believe about science, then normative judgments must become part and parcel of science. What has not generally been recognized is that if science not only rejects the fact–theory distinction, but the fact- value distinction as well, then it actually becomes easier to stem the tide of relativism, provided that scientists are prepared to accept the responsibility that accompanies their theory choices. If educational scientists qua scientists must make value decisions, then we have a right to demand that their decisions be constrained by the same practical virtues that constrain choice and action in everyday living.
As noted earlier, researchers, including educational researchers, often require pragmatic principles such as simplicity and conservatism in drawing their inferences to best explanation. Clearly, it is inappropriate and irresponsible to have recourse to practical judgment and practical virtues unless the principles of inductive and deductive logic have first been exhausted. What has not been made clear is that there is a logic of practical judgment; it is the means-ends reasoning of practical deliberation. In deliberation, what is called for are researchers that resemble Aristotle’s “man of practical wisdom” (Nicomachean Ethics, Bk. VI, Ch. 5). Strictly speaking, Aristotle did not think that we could deliberate about scientific knowledge since it admitted demonstration; but if the postpositivistic critique of underdetermination, the QuineDuhem thesis, theory-ladenness, and all the rest is correct, then Aristotle is simply mistaken, and normative considerations can and must be introduced into scientific theorizing. For Aristotle, those possessed of skill in deliberation, that is, those possessed of the virtues of practical wisdom, are distinguished by the acuity of their judgments. The pragmatic response to holism seems to manifest the virtues of practical wisdom in deliberation.
One consequence of resorting to deliberation and means-ends reasoning is that attention must be paid to the ends or purposes of the research paradigm (Kuhn, 1962), program (Lakatos, 1970) or tradition (Laudan, 1977). Among philosophers of science, Toulmin (1972, 1983) seems unique in recognizing the importance of purpose in giving reason and rationality to scientific inquiry, although even he fails to appreciate the role of means as well as ends in the logic of inquiry. Recently, Robert Donmoyer (1985) introduced the views of Toulmin into the debate concerning relativism in educational research.
I would like to expand the set of pragmatic responses to relativism beyond deliberation to include rhetoric and dialectic. More exactly, I would like to expand deliberation to include rhetoric and dialectic. I take my lead from the late Chaim Perelman (1979), who developed what he called “The New Rhetoric.” For Perelman, this rhetoric was not just an “ornamental art,” merely a matter of style, but rather a means of resolving conflicts where neither logic nor experience are sufficient to decide the issue. The resolution is achieved by appealing to the “universal audience” of those competent to judge – in the present case, educational theorists and researchers. Perelman’s work was, as he himself saw it, an extension of Aristotle’s thoughts on rhetoric and dialectic applied to the sciences. Since to err is human, it is understandable that humankind prefers not to trust its own judgment; yet if the pragmatic view of science is correct, we are thrown back on ourselves and a degree of normative judgment is necessary even in the descriptive sciences. Somehow this seems less disturbing when, as in education, it is the science of human nature with which we are concerned. What Pierre Duhem himself wrote over 50 years ago regarding the physical sciences applies many times over to the “doubly value-laden” social and psychological sciences:
Since logic does not determine with strict precision the time when an inadequate hypothesis should give way to a more fruitful assumption and since recognizing this moment belongs to good sense, physicists may hasten this judgment and increase the rapidity of scientific progress by trying consciously to make good sense within themselves more lucid and more vigilant. (Duhem 1974, p. 218)
Duhem concludes by referring to the following “expression” by Claude Bernard that very much resembles the thinking of Perelman:
The sound experimental criticism of a hypothesis is subordinated to certain moral considerations; in order to estimate correctly the agreement of a physical theory with the facts, it is not enough to be a good mathematician and skillful experimenter; one must also be an impartial and faithful judge. (Duhem, 1974, p. 218, italics added)
There is no substitute for wisdom.
Pragmatic and purposeful deliberation, rhetoric and dialectic may carry us safely over the perils of logical and empirical underdetermination and the Quine-Duhem thesis generally, but it seems, at first glance, inadequate for dealing with the problem of theory-ladenness and strong incommensurability. This would be a serious failure since it is the latter two principles, especially incommensurability, that are seen as the strongholds of relativism. Fortunately the scope of practical deliberation can be readily extended by expanding the “universal audience” of those competent to judge to include not only educational theorists and researchers, but educational practitioners as well, that is, teachers, principals, supervisors, and others.
Practitioners are an important source, perhaps the most important source, of practical wisdom in education. More to the point, the language and shared experience (facts) of the practitioner provide a common, although not necessarily fully neutral, nonscientific ground on which to judge between competing scientific theories. If the relativist is right and, as Protagoras proclaimed, “man is the measure of all things,” then at the very least let us take “man” here in the broadest generic sense as the entire linguistic community of those theoreticians, researchers, and practitioners competent to judge. Finally, let us not fail to include anyone else whose judgments may have practical virtue and who are prepared to accept responsibility for their choices.
Aristotle (1941) Nicomachean Ethics. In: The Basic Works of Aristotle, Richard McKeon. New York: Random House.
Donmoyer R. (1985) The rescue from relativism: Two failed attempts and an alternative strategy. Educational Researcher, 14 (10): 13–20.
Duhem P. M. (1974) The aim and structure of theory (P. P. Wiener, Trans.) New York: Atheneum. (Original work published 1906)
Eisner E. W. (1981) On the differences between scientific and artistic approaches to qualitative research. Educational Researcher, 10 (4): 5–9.
Feyerabend P. (1969) Science without experience. The Journal of Philosophy, LXVI: 791–794.
Garrison J. W. & Macmillan C. J. B. (1984) A philosophical critique of process-product research on teaching. Educational Theory, 34 (3): 255–274.
Hanson N. R. (1958) Patterns of discovery. Cambridge, England: Cambridge University Press.
Howe K. R. (1985) Two dogmas of educational research. Educational Researcher, 14 (8): 10–18.
Kuhn T. (1962) The structure of scientific revolutions. Chicago: University of Chicago Press.
Lakatos I. (1970) Falsification and the methodology of scientific research programmes. In: I. Lakatos & A. Musgrave (eds.) Criticism and the growth of knowledge (pp. 91–196) Cambridge, England: Cambridge University Press.
Laudan L. (1977) Progress and its problems. Berkeley: University of California Press.
Macmillan C. J. B. & Garrison J. W. (1984) Using the “new philosophy of science” in criticizing current research traditions in education. Educational Researcher, 13 (10): 15–21.
Newton-Smith W. (1981) The rationality of science. Boston: Routledge & Kegan Paul.
Perelman C. (1979) The new rhetoric and the humanities. Holland/Boston: D. Reidel, Dordrecht.
Phillips D. C. (1983) After the wake: Postpositivistic educational thought. Educational Researcher, 12 (5): 4–12.
Popper K. R. (1968) Conjectures and refutations in the growth of scientific knowledge. New York: Harper Torchbooks.
Quine W. V. O. (1961) Two dogmas of empiricism. In: From A Logical Point Of View (2nd ed.: 20–48) Cambridge MA: Harvard University Press.
Quine W. V. O. (1969) Epistemology naturalized. In: Ontological relativity and other essays (pp. 69–90) New York: Columbia University Press.
Quine W. V. O. (1974) The roots of reference. La Salle IL: Open Court.
Quine W. V. O. & Ullian J. S. (1970) The web of belief. New York: Random House.
Rudner R. (1953) The scientist qua scientist makes value judgments. Philosophy of Science, 20 (1): 1–6.
Siegel H. (1980a) Objectivity, rationality, incommensurability and more. British Journal of Philosophy of Science 31: 359–384.
Siegel H. (1980b) Justification, discovery and the naturalizing of epistemology. Philosophy of Science 47: 297–321.
Siegel H. (1982) Relativism refuted. Educational Philosophy and Theory, 14 (2): 47–50.
Siegel H. (1986) Relativism, truth, and incoherence. Synthese, 68 (2): 223–254.
Soddy F. (1932) The interpretation of the atom. London: Murray.
Toulmin S. (1972) Human understanding. Princeton NJ: Princeton University Press.
Toulmin S. (1983) The construal of reality: Criticism in modern and post modern science. In: W. J. T. Mitchell (ed.) The politics of interpretation (pp. 99–117) Chicago: University of Chicago Press.
Tuthill D. & Ashton P. (1983) Improving educational research through the development of educational paradigms. Educational Researcher, 12 (10): 6–14.
Wittgenstein L. (1969) Philosophical investigations (3rd ed) New York: Macmillan.
The situation is even more serious than I have indicated. Carl Hempel and Nelson Goodman have each developed different logical paradoxes of confirmation (Hempel) and induction (Goodman) that, like the original psychological problems posed by Hume, have never been satisfactorily resolved.
A colleague uses a 7-foot Pooka that he claims to see in his social foundations of education course to make this point quite effectively. I do not believe in this Pooka or in unicorns either, for that matter. I do believe in cognitive schemas and electrons, although neither I nor anyone else has actually ever seen them. The existence of schemas and electrons must be inferred from evidence that only indirectly confirms them. The difficulties of the cognitive psychologist are not greater, or less, than that of the atomic physicist in such matters.
The full significance of this possibility was first realized by Lakatos (1970).
Although in The Web of Belief Quine and Ullian (1970) consider five “virtues of hypotheses,” in The Roots of Reference Quine (1974) seems to settle on simplicity and conservatism as the highest virtues.
For a rebuttal of those who would blur the context of discovery with context of justification, see Siegel (1980b).
These non-Proutian experiments remain replicable to this day; it is not their truth but their meaningfulness that has vanished. We should not forget that meaning underpins and is even more basic than truth.
Doug Tuthill and Patricia Ashton (1983) do a good job of drawing out some of the implications of Kuhn’s work for the epistemology of educational theory and research. It was Kuhn who introduced the notion of incommensurability into the vocabulary of the philosophy of science. Kuhn himself was concerned with the incommensurability of research paradigms. Qualitative versus quantitative research paradigms would be candidates for incommensurability in the eyes of many.
Found a mistake? Contact corrections/at/cepa.infoDownloaded from http://cepa.info/3924 on 2016-12-30 · Publication curated by Alexander Riegler