Much of the tension that currently exists between applied and academic orientations in sociology can be explained in terms of political, professional, and even interpersonal issues and concerns. However, underlying these factors is a genuine epistemological difference that may prove to be the most recalcitrant of all to resolve. Applied (or practicing) sociologists tend to embrace an essentially pragmatic philosophy of science in which the ultimate test of theory is the extent to which it can produce knowledge that "works." Academic sociologists employ the nomological model that focuses on the development of theory for its own sake: that is, the value of theory is determined by the extent to which its law-like generalizations (nomos) are supported in hypothesis-testing situations. The discrepancies between these two approaches are illustrated by a case study that is among the best-known and most consequential sociological applications of all time: the work of Gunnar Myrdal, Arnold Rose, and others on racial segregation in the United States.
The first chapter of Jonathan H. Turner’s Classical Sociological Theory: A Positivist’s Perspective (Turner, 1993) is entitled “Comte Would Turn Over in His Grave.” In this enlightening essay, Turner demonstrates that Comte’s “positive science” clearly and explicitly included a central role for theory in social research. Similarly, Comte dismissed as unscientific the kind of empirical research that is conducted in the absence of theory. That is, he stressed that sociological method is in principle opposed to practices referred to today by many sociologists as positivism! 1 In Comte’s words, as quoted by Turner, true social science seeks to avoid:
. . . empiricism which is introduced into it by those who, in the name of impartiality, would interdict the use of any theory whatever . . .. [N]o dogma could be more irreconcilable with the spirit of positive philosophy .. no real observation of any phenomenon is possible, except insofar as it is first directed, and finally interpreted, by theory (Comte, 1896 : 242-43; Turner, 1993: 2). 2
Turner follows these observations with a speculative discussion concerning how positivism came “to denote quantitative research divorced from theory.” Here he makes reference to the influence on Western philosophy—and on many sociologists (e.g., Paul Lazarsfeld)—of the Vienna Circle, the logical positivists, who were active during the first half of the twentieth century. In another context, Jonathan and his co-author Steven Turner connect these developments to internal politics and the competition for large research grants at critical junctures in the history of the profession (Turner and Turner, 1990). And, although neither Jonathan nor Steven makes this point explicitly, it seems clear from their comments that the practice of equating positivism with “mindless empiricism” or—better— “abstracted empiricism” 3 is explained at least in part by the failure of many sociologists to have ever read anything about positivism written by Comte.
The discussion that follows in this paper is based on the understanding that the same forces that have conspired, ironically, to characterize Comte’s positivism as anti-theoretical have also distorted and otherwise obscured another feature of his work. For, contrary to prevailing views among academic sociologists today, not only did Comte believe that theory was a vital component of social research, he also believed that the theories developed by researchers should be given their ultimate test in real-world applications. That is, positive science is indeed a theory-directed type of intellectual activity; but it is not an activity whose sole end is self-perpetuation—knowledge for knowledge’s sake. Rather, the ultimate purpose of sociological knowledge is to change the world in determinate ways: to create a Positive Polity, 4 a society based on science rather than on superstition or metaphysics.
Inherent in the (actual) doctrine of positivism is the understanding that until and unless our theories are applied, we will never know if they are valid. Under the strong influence of Newton, Comte was aware that Absolute Truth would always evade the scientist, but that sound theory could, at least, initiate a revolution—industrial in Newton’s case and social in Comte’s. Thus, not only is positivism opposed to the kind of abstracted empiricism with which it is wrongly equated, as Turner has helped us to understand; it is opposed to the apolitical pure-science approach that also has often and incorrectly been labeled “positivism.”
Any discussion of sociological theory such as this must stipulate, if only briefly, how key terms are to be used. In this context, I would like to draw attention to two important distinctions that are inherent but rarely explicated in this field of study. The first has to do with the relationship between scientific theory, in general and sociological theory, in particular; the second relates to the difference between sociological theory and theoretical sociology.
Jonathan Turner is one of a small group of observers who have commented critically on the tendency among sociologists to use the term theory, Humpty-Dumpty fashion, in any way they choose. One of his often-cited examples is the prevailing belief among multivariate analysts that the reduction of statistical variance is tantamount to theory building. This is related to the similarly mistaken notion that theory “emerges” when a sufficient number of facts are gathered about a subject. Other instances of inappropriate uses and/or sheer misunderstanding of the act of theorizing include the following. (1) The common practice of labeling as theory reviews of the lives and works of classical sociologists; (2) the program to develop “middle-range” theory; (3) the continuing interest in new causal theory 5 among proponents of the now-classic Duncan-Blau model of social stratification; and, most recently, (4) the enthusiastic following that has been generated by rational choice “theory.”
I recently examined some popular textbooks on sociological theory while preparing for my graduate seminar on the subject. There I discovered what I view as a kind of ultimate instance of this lack of interest in using the term properly; or, what amounts to the same thing, a lack of interest in using the proper labels for non-theoretic types of scientific activity and their products. The author of this statement is George Ritzer, one of the most prolific and respected contemporary writers on sociological theory. According to Ritzer—whose credentials include service as the Chair of the Theory Section of the American Sociological Association—sociological theories are “things” that “have a wide range of applications, deal with centrally important social issues, and have stood the test of time” (Ritzer, 1996: 4). Unfortunately, Ritzer does not specify what kinds of “things” theories are, how one determines if a range of applications is wide or narrow, or what it means to “deal with” social issues. Thus, one could credibly include under his definition such diverse items as Tylenol, The New York Times, rent control, and The Encyclopedia Britannica-Online.
Eschewing such loose stipulations, scientists and philosophers of science (e.g., Rudner, 1965; Suppe, 1974) generally concur that a theory is, first and foremost, a linguistic formulation. It consists of words and sentences (statements). It is not part of the “object world” of things and actions; rather it is something that can be uttered and/or written in ordinary language, mathematical symbols, etc. In addition, the sentences of which a theory consists are systematically related—as opposed to being a mere collection of discrete statements. Moreover, the system whereby its sentences are connected is deductive logic; which, as you will recall, has five and only five basic connectors: “not,” “and,” “or,” implication (“if . . . then”), and identity (“same as”).
Among the sentences of a theory, some specifically have empirical referents: their subjects refer to objects in the empirical world and their predicates refer to qualities or actions of these objects. Such sentences have the characteristic that Karl Popper and others have referred to as “falsifiability.” This means that their apparent truth can be challenged by sense data that are contrary to what the sentences assert to be the case. Not all the sentences of a theory need to be empirical in this sense; for most theories also include definitions, statements of equivalence, implications, and the like. The truth or falsity of these depends on whether or not they are properly constructed according to the rules of logic. Thus, they cannot be shown false with empirical data. In any case, the empirical sentences of a viable theory must be supported by fact and never (yet) falsified. If and when data are definitively shown to contradict a theory, then the theory is no longer viable and it must be revised or discarded.
The final feature of scientific theory is perhaps its most essential and its most controversial. For, in addition to supported and not-yet-falsified empirical sentences, a theory also includes some such empirical sentences that refer to general classes of phenomena. These are generalizations of the form “all x are y,” or “every a is also b”: (1)”All large, industrial cities have high (compared to an explicit standard) rates of violent crimes,” (2) “every successful charismatic party becomes routinized,” and so on. Like the other empirical sentences in a theory, these are believed to be true—based on observation. And because of the scope of their truths, they are reminiscent of what were once referred to as “laws.” However, they must also be falsifiable, so that their truth cannot be deemed “eternal” or “unassailable.” For this reason they are properly called “law-like generalizations.”
These generalizations play a pivotal role in the way theory is used and developed. When combined with empirical observations, they are productive of hypotheses. In turn, when such hypotheses are tested in relation to further empirical observations, they reflect on the viability—or lack of viability—of the theory from which the generalization is drawn. For example, suppose that we have developed a theory that contains sentences about the nature of life in large industrial cities (in comparison to smaller settlements), the nature of certain types of crime, and related matters. And suppose that one of the sentences in the theory is the generalization stated above as number (1). Finally, suppose that we have determined that a specific city, “D,” is large and industrial. We can then set up the syllogism:
All large industrial cities have high rates of violent crime
“D” is a large industrial city
Therefore “D” has a high rate of violent crime.
In this formulation, the third sentence, which is the valid conclusion of a logical argument, is, in fact, a research hypothesis. Like all hypotheses, it is an “educated guess.” The “educated” part comes from the fact that the theory indicates that it should be true. The “guess” part indicates that we do not know that it is true, we only believe that it should be true. As in all cases of hypothesis testing, its truth depends upon what the empirical evidence indicates. Thus, if we count up the number of violent crimes and divide by the relevant population size, and if the outcome is a high number in relation to pre-established standards, then we accept the hypothesis. In such cases, the theory is also upheld (for the time being) because it produced a hypothesis supported by the facts. If the crime rate of “D” turns out to be low—by accurate and reliable measurements, then the hypothesis must be rejected. In cases like these, the theory has been shown to be weak and in need of revision or abandonment.
In my view, the controversy surrounding this property of theory arises in part from the failure of critics of “positivism” to grant that, by contemporary standards, scientific laws—and thus scientific truths—are understood to be highly provisional. These critics fail to acknowledge that scientists and philosophers of science are well aware that the viability of a theory is of necessity a fleeting thing, because law-like generalizations and other empirical sentences depend upon always-imperfect observation. Instead, the critics characterize their opponents as seekers of absolute certainty. Then, based on such characterizations, they argue that because such certainty is impossible to attain, the program of “positivism” should be abandoned. Here we have a perfect example of the straw man argument (as well illustrated in Seidman, 1991, and in the introductory essay and several specific chapters of Ritzer, 1997).
It is not always easy to determine why these critics continue to insist that scientific theory somehow promises to produce unassailable wisdom. But it appears that the strict stipulations for viable theory are so daunting that the critics do not believe one can actually abide by the rules. Thus, it is claimed, scientific theory (or at least sociological theory) cannot be produced: theorizing in the sense outlined here is impossible. There is something to be said for this argument. For it is highly unlikely that in any scientific field all of the so-called theories are in “perfect” condition: that all logical relations are valid, that all empirical sentences have been substantiated, that no law-like generalizations have ever been contradicted by fact. Rather, the rules governing the creation of viable theory constitute an ideal type (like a perfect vacuum or Weber’s bureaucracy) that serves as a point of reference for the actual activities of imperfect, real-world scientists. Thus, it is not necessary to interpret these rules as if they were absolute directives saying “thou must relate all sentences in a logical system,” and so on. Instead, they can be understood as important objectives to “strive to the utmost to relate all sentences in a logical system.” Implicit in this less stringent set of criteria is the view that it is better to attempt to be as correct as possible than—as the critics seem to be arguing—to abandon the attempt altogether because the task is too difficult.
With this noted, we are left with a characterization of scientific theory and theorizing that is at once positivistic (in the true sense) and yet within the grasp of ordinary human beings. It is what Israel Scheffler (1971) once termed “modest empiricism;” and it comports with the actual views and activities of contemporary science much more accurately than do the stilted and over-blown caricatures created and promulgated by the anti-science critics. What, then, is scientific theory? Based on the preceding discussion, I would say that it is no more and no less than previously established knowledge that is organized to facilitate hypothesizing. Is it possible to attain? Of course. Is it absolute, eternal truth? Of course not. Does it or can it exist in sociology? Let us take up this question in the following section.
The question of how sociological theory is related to scientific theory can be addressed in at least two ways. On one hand, and most directly, it can be argued that the former is simply one type of the latter. In this sense, physicists use physical theory, biologists use biological theory, sociologists use sociological theory, and so on. The category scientific theory is thus an abstraction based upon the common properties of the types of theories used in each of the scientific fields. Sociological theory can exist because sociologists—like any other scientists—can establish a base of knowledge, organize it logically, and use it (correctly) to generate hypotheses. Moreover, sociological theory does exist because one can identify in the works of its practitioners linguistic systems that satisfy the above-stated criteria. In fact, the purpose of Turner’s (1993) book and many (but by no means all) well-known works on the subject is to state for the reader the main generalizations and other parts of leading theories.
An alternative, and more probing, approach to establishing if and how sociological theory is related to other kinds of scientific theories begins with another question: Is sociology a coherent field whose terminology corresponds to real-world objects and events? That is, are there properly sociological phenomena whose behavior can be comprehended with authentically sociological concepts embodied in strictly sociological sentences? If the answer to this is “yes,” then one can conclude that sociological theories are like all other scientific theories in structure or form but that they differ because they refer to distinctive units of observation/analysis and variables. Physical theories thus refer to physical objects such as light and sub-atomic particles, biological theories refer to biological objects such as organisms and genes, and sociological theories refer to human interactions, groups, organizations, and the like. But if the answer is “no, there are no strictly sociological phenomena,” then the status of sociological theory is in doubt. In this case, it is possible that sociologists—certified or self-proclaimed as the case may be—have developed and effectively used theories. But it would be a gloss to refer to them as “sociological.”
This less-comforting scenario has been debated in one form or another many times over the past two hundred years or so. For example, it is closely related to the issue of emergence versus reductionism that concerned, among others, Herbert Spencer and Emile Durkheim. If all so-called sociological phenomena can be reduced to psychological ones, then sociological theory is at best superfluous and at worst misleading in understanding how people behave, etc. But if sociological phenomena have a property—which Spencer first termed the “superorganic”—that defies reduction, then sociological theory is not only possible, it is required in accounting for human actions.
At the other extreme, it is reasonable to assert that phenomena labeled sociological cannot in reality be separated from properties that are economic, legal, philosophical, political, cultural, and even geographical. From this perspective, it would be more appropriate to conceive not of sociological units, variables, and so on, but of social scientific theory. By extension, any viable understanding and/or explanation of social phenomena must be what we now refer to as interdisciplinary. This argument is all the more credible when we consider the backgrounds of the major, classical sociological theorists. Ibn Khaldun was a philosopher and historian. Adam Smith and his colleagues constituting the Scottish Moralist school were moral philosophers. Malthus was a theologian and political economist. Herbert Spencer was a philosopher and economist. Marx was a philosopher and legal historian—and a professional revolutionary. Weber was an economic historian, and Durkheim—although commonly identified as a sociologist—was trained in philosophy and law. How, one might wonder, can such a cast of characters produce theory that is somehow distinctively sociological?
My purpose in raising these perennial issues is hardly to resolve them. For one thing, I am not sure that they are subject to resolution. It may be that they ultimately involve value judgments and thus will always serve to divide sociologists and others into relatively distinct schools of thought. But I do believe that when we consider such apparent dilemmas—and suggest the many related ones not explicitly noted here—we can better appreciate how complex is the reality to which sociological theory must be addressed. It is true that in this complexity, and in the superorganic level at which social phenomena operate, some unique solutions may lie. For where other sciences depend solely on theories to generate credible hypotheses, sociologists have mechanisms such as Verstehen to anticipate why people might behave as they do. 6 Nevertheless, it is a kind of complexity that I believe makes the task of creating and properly employing sociological theory more difficult, by degree if not in kind, than that faced by other scientists.
In summary, sociological theory must be formulated and used under uniquely challenging epistemological and ontological conditions. More than in other scientific fields, the correspondence between sociological terms and the phenomena to which they are meant to apply is in doubt. Moreover, when sociologists seek to establish generalizations, they typically find themselves trapped between the Scylla of reductionism and the Charybdis of interdisciplinarity. For this and related reasons, it is apparent that sociological theory (1) is indeed a kind of scientific theory but (2) it is also a kind that is especially difficult to “establish.”
In the face of such challenges, sociologists—past and present—have created a type of linguistic formulation, which, although lacking some relevant features, is nevertheless labeled theory. This type of account is based upon one or more simplifying assumptions about the objects of sociological analysis. Once stipulated as true, the assumptions play the role of law-like generalizations, which, in combination with empirical observations, are productive of apparently credible hypotheses. When tested, such hypotheses are rarely if ever entirely supported by the evidence. Nevertheless, they are accepted as approximately true in view of the abstract character of the “laws” from which they are derived. Under these circumstances, the formulations not only appear to be viable they seem irrefutable.
No doubt, these formulations have the appearance of sociological theory. However, they are more accurately understood to be theoretical sociology—with an emphasis on “theoretical” in the ordinary language sense of “unreal.” That is, these are sociological accounts derived from unsubstantiated—or clearly false—premises. In his theory seminars, Alvin Gouldner often compared these to a system of gymnastics based on the assumption that people had no skeletons. Under the rules of such a system, performers could tie themselves in knots, roll themselves up into compact little balls, and perform other wonderful feats. Unfortunately, this system would not be very useful in helping observers understand actual gymnastic performances. Similarly, regardless of how ingenious its assumptions may be, theoretical sociology is of little use in accounting for the actual properties of social relations.
Two well-known examples come to mind in this regard, both of which have a distinguished heritage—and at least one of which is currently experiencing a highly celebrated revival. These are (1) evolutionary “theory” and (2) rational choice “theory.” The former extends back to the very origins of sociology, to a time when the ideas of Darwin and Spencer had captured the imaginations of social thinkers. The main simplifying assumption in this approach is that human systems—groups, societies, and cultures—change according to the same principles that apply to organisms and species: natural selection, competition, environmental adaptation, extinction, speciation, and the like. Although it is true that, for nearly as long as social scientists have held this view, able critics have challenged it. 8 Nevertheless, explicitly and implicitly, it continues to pervade much work in macrosociology. Thus, we read about the shifts from mechanical to organic solidarity, from Gemeinschaft to Gesellschaft and from feudalism to capitalism. We then assume that they are essentially the same processes whereby the dinosaurs came to rule the earth, deciduous trees came to dominate in temperate climates, or the necks of giraffes became elongated.
As critics of this approach, including Robert Nisbet (1969), have pointed out, the simplifying assumption—that sociocultural change is the “same” as organic evolution—does not agree with the facts, as we understand them. What is true is that evolution is a powerful metaphor for much that occurs as human systems are transformed over time. But to build statements into a theory-like formulation that literally equate the two processes is virtually to ensure misunderstanding of empirical phenomena. Without delving too deeply into the critique of this type of theoretical sociology, I will just remind you (if such a reminder is necessary) that intention, planning, and willful acts always enter into change in human systems in ways that cannot be attributed to their organic counterparts. Although unintended outcomes are perhaps as common as those that are intended, the fact that intent is a characteristic component of sociocultural transformations sets them apart from the kinds of changes studied in other sciences.
The second example of theoretical sociology, rational choice “theory,” seems to have attracted many contemporary sociologists—for reasons that I frankly do not quite understand. This approach has been an integral part of microeconomic theory for generations. In that context, it has been developed extensively, tested widely, and it has become highly formalized through the adaptation of its basic principles to differential and integral calculus. In brief, it is among academe’s best-established and most sophisticated attempts to understand human behavior. Thus, it is a puzzle as to why it should now be heralded as a novel type of “theory” and why educated people should take the crude efforts—so crude that they invite ridicule among economists—of its sociological proponents seriously.
Notwithstanding these peripheral concerns, rational choice “theory”—in its economic and sociological versions—is seriously flawed. For, like evolutionary “theory,” it is based upon an attractive but false set of simplifying assumptions. The central principle (usually disguised as a law-like generalization) is that human individuals always act in such a way that they (1) first consider the behavioral choices before them and (2) then select the option that they believe will maximize the benefits forthcoming to them. Assuming that this is true, one can account for individual and aggregate behavior by ranking the potential courses of action available to relevant actors and identifying the ones that are most likely to pay off.
Of course, as any sociologist should know, people do not always behave this way. They do not unerringly know all of the possible courses of action they might take. They do not always act as individuals (for they are sometimes representatives of others, members of groups, faces in the crowd, etc.). They may not care to or be able to discern which options will reward them the most. Several different and possibly contradictory rewards can result from a single act. And sometimes people act for the benefit of others, not themselves. Moreover, we do not really have an adequate way of deciding which set of conditions is conducive to behavior that follows the rational choice principles and which is conducive to the other possible outcomes. In brief, if people were rational actors, their acts could be explained effectively. But they are not, and the “theory” cannot account for actualities.
I do not want to suggest that this is a new or an original critique. It is not. Moreover, microeconomists and their sociological counterparts have over the years developed a well-rehearsed set of responses to such concerns. One of the most familiar is the supposed refutation of the last-mentioned issue, the possibility that people may at times behave altruistically. To this objection, the rational choice theorists argue in the following manner: Some people, sometimes do behave in ways that appear to benefit others with little regard for the self—perhaps even by exposing themselves to danger. For example, a citizen bystander may rush out into heavy traffic to rescue a person who is in distress and likely to be struck by a car. However, such acts are undertaken because the actor-rescuer will receive considerable psychic benefits, perhaps even a reward. In such instances, the potential dangers of running out into traffic are weighed against the benefits that accrue in our culture to those who perform a heroic act, and the latter course is selected.
Such a response seems to satisfy the proponents of rational choice theory. But some of the more astute critics remain unimpressed. Kristin Monroe (1996; Monroe, Barton, and Klingermann, 1990), whose research on altruism is among the most effective contemporary work on the subject, has pointed to the major flaw in this counter-argument. According to her, if what appear to be the most selfish and the most selfless acts can both be explained by the same formulation, then there is no longer a need for empirical research. Any behavior has a ready-made explanation. Questions: Why does John eat an apple? Why does Joan eat a pear? Why does John give his apple to Joan? Why does Tom steal Joan’s pear? Answers: rational choice, rational choice, rational choice, and rational choice. Neat, isn’t it? Neat, but not very enlightening. Although I admit that this is something of a caricature, it does illustrate the pitfalls involved in conflating sociological theory and theoretical sociology.
If the issues associated with the meaning and uses of theory are as serious as I am suggesting, and if the abuses are as blatant as I claim, then why, one might wonder, have they not been addressed more effectively? Put less politely, how can so many sociologists get away for so long with such apparent lack of concern about such an important subject? The answer, I believe, can be found in the insular character of academic sociology. For as long as sociologists are communicating only with each other, and as long as the test of sociological “theory” lies exclusively in the hands of sociological peer reviewers, there is neither incentive nor means to rectify logical and ontological problems. If a so-called theory can pass the scrutiny of disciplinary elites (for example, journal editors and tenure and promotion committees), it isn’t necessary for it to produce credible—and testable—hypotheses; it isn’t even necessary for it to be a theory.
This diagnosis suggests that the problems discussed above will never be meaningfully addressed until and unless sociologists take their theoretical work beyond purely academic forums. That is, the time may have come to return to the most harsh but most relevant proving ground for sociological theory, the real world. To repeat a well-worn (but perhaps inadequately understood) quote: “Philosophers have interpreted the world in many ways. The point, however, is to change it.” In this spirit, the following sections discusses two contrasting kinds of knowledge (or, better, knowledge quests) based on the work of Nico Stehr (1992): the nomological model and the pragmatic model.
Amidst the perennial and often excessively introspective debates about whether or not sociology is/can be an “authentic” science, one or another commentator occasionally suggests that sociologists have a rather strange and distorted idea about what other scientists do. 9 Because, it is argued, sociologists misunderstand the nature of scientific activity both in general and in the fields such as physics and biology, which they most ardently seek to emulate, they have set before themselves an impossible task (see Turner and Turner, 1990). This observation certainly appears to apply to the anti-scientific critique that accuses “positivists” of seeking absolute truth. For, it seems, one or another (or both) of the camps divided over the question of the very possibility of sociological theory appears to have embraced the seriously mistaken view that other sciences (e.g., physics) can attain certainty. Based on this assumption, those who nevertheless continue to attempt to develop “laws” of social relations—thus strictly following the nomological model—are bound to be frustrated. Whereas those who altogether reject the project of developing a scientific sociology are perpetually reinforced in their misdirected nihilism.
In contrast, those who are capable of stepping outside of their narrow, disciplinary roles and of learning something about how the non-social sciences actually operate are well aware that the search for certain knowledge was abandoned decades, if not centuries, ago. Scientists do not overtly and directly seek to establish irrefutable inductive truths. They know, as sociologists should also know, that “irrefutable inductive” is a contradiction in terms. Rather, they seek exceptions to widely accepted beliefs; contradictions to what ought to be observed; and unfulfilled expectations in experimental outcomes. In this respect, all scientific knowledge is provisional; otherwise, there could be no such thing as scientific progress. As noted earlier, the key generalizations of scientific theories are referred to as law-like for a good reason. Namely, they are bound to be revised or abandoned eventually as they generate more and better hypotheses.
It is a safe guess that every sociologist in the United States, at least, has learned in an introductory course in statistics the difference between a research hypothesis and a null hypothesis. It is equally certain that in the process the point was made very clearly that the appropriate scientific form is to test and to strive to uphold the null. Only when the laws of probability dictate otherwise do we then conclude that the research hypothesis is probably true (within stated limits of confidence that may approach but never reach 100 percent). This lesson is not gratuitous. Nor is it intended only to establish a ritualistic formula for writing the third section (following “Review of Literature” and “Theory”) of a journal article. Instead, it meant to help the aspiring sociologist think like a scientist. And this means to avoid at all costs any suggestion that the truths that are discovered with the aid of the scientific method are certain or absolute. To the contrary, those who have learned these lessons well are especially sensitive to the provisional character of inductive proof.
Lest I be misunderstood here, I am not saying that scientific theory routinely contains obvious falsehoods. For, as mentioned in our introduction, a theoretic formulation that includes one or more sentences known to be false is, strictly speaking, not a theory. Each statement, and especially the law-like generalizations, must have been supported by empirical evidence on some occasions and never (yet) refuted. But if such statements are truly empirical, then it remains possible for them to be refuted at some future date—and, in the long run, it is likely that they will be refuted. To expect more, as some sociologists seem to do, is tantamount to guaranteeing that theirs will forever be an “impossible science.”
If we abandon the search for absolute theoretical truth, are we not then left without a criterion for judging the viability of theory? How, after all, can one determine if theory A is superior to theory B, if both are equally likely to produce false hypotheses at one point or another? These appear to be the very kinds of questions that keep alive the debate about whether or not sociology is/can be a scientific discipline. They are also based on the prevalent misunderstanding of science outlined in the last section. For in posing them, one conjures up an image of a battle between, say, two physicists, each having developed a theory of the behavior of a newly discovered sub-atomic particle. The first physicist generates and tests a series of ten hypotheses, all of which are upheld by observation. The second physicist is successful in testing nine hypotheses, but falters on the tenth. At that point, an impartial judge declares that the first theory is the one that shall prevail, because it is “truer” or “more valid.”
This neat and somewhat suggestive depiction of how “real” science works supports the pessimistic appraisal of sociology’s prospects as a science. And perhaps for that reason it is held to be more or less accurate—at least implicitly so. However, we have known for some time that this image does not portray the mechanisms whereby some scientific theories survive and others are discarded. Rather than judgments being made on the basis of veridicality—the capacity to generate truth, they are far more typically made on the basis of practicality—the capacity to be effective in application to practical problems.
In the title of his recent book on William James, Harvey Cromier (2000) put it succinctly: “The truth is what works.” As Stehr (1992) noted somewhat earlier, knowledge that has practical consequences, consequences that can be observed and which are in some way valued, is knowledge that survives and becomes certified. Knowledge that cannot be shown to make a difference in the world is likely to be forgotten. Moreover, these outcomes are largely unrelated to the ultimate truth-value of the theories behind the knowledge. As long as it “works,” a scientific theory is good enough.
Before illustrating and elaborating on this point, let me immediately indicate what this means for sociological theory. Following Stehr, I am suggesting that the true test of theory is the extent to which it makes a difference in practice. If this is so, then the value of specific theories and even the question of whether or nor there can be such a thing as sociological theory can never be resolved solely within academic contexts. For in such contexts judgments must be made only in relation to the (inappropriate) nomological model. We must, in brief, take our theories into the real world if we have any hope of achieving our cherished goal of becoming real scientists.
This account of practical knowledge has its immediate roots in the pragmatist epistemology of the late nineteenth and early twentieth centuries, first outlined by Josiah Royce (1969), developed by C.S. Peirce (1998; Hookway, 2000), and applied by William James (Cromier, 2000)—all of them American philosophers. Thus, I have termed it the pragmatic model. As students of the pragmatist (or as James preferred “pragmaticist”) movement are aware, this theory of knowledge was developed in conscious opposition to the idealist underpinnings of the nomological model.
Peirce (1888; 1985) provided one of the most succinct statements in his essay “How to make our Ideas Clear,” first published in Popular Science. Here Peirce characterizes the mind as a hierarchy of beliefs. At the top of the hierarchy, nearest the surface, are our fleeting opinions that are supported only weakly by actual experiences. For instance, we may come to believe that our new neighbors are friendly based upon an initial meeting and short conversation. On such a basis, we would be cautious about what we might expect from them and not terribly surprised if we eventually learn that they are not as nice as we first believed. At the deepest level are the beliefs that we refer to as “knowledge.” These have been verified time and again as we put them to the test in everyday situations: e.g., the belief that when we return from work our house will be where we left it in the morning. We are inclined to put considerable stock in such beliefs and to act on them with confidence (in this example, we would routinely proceed home in the usual direction).
The important thing to note, Peirce argues, is that the highest and lowest levels, and all levels in-between, differ from one another only in degree not in kind. The beliefs that we designate as knowledge are those that “work.” And every act we perform is, in one way or another, a test of our beliefs. Thus, we cannot say that even our best verified beliefs represent eternal truths. During the recent earthquakes in Turkey, thousands of people returned from work in good faith, “knowing” the location of their homes, only to be dismayed that this knowledge no longer worked. In such cases, one is compelled to have a “change of mind” through a reconsideration of what one really knows.
So it is with scientific theories. Every time we act upon them, we are putting them to the test. But the entire range of argumentation and formal hypothesis testing one can marshal will never prove that our law-like generalizations constitute knowledge. If we can act effectively on the basis of such beliefs, then they are tantamount to knowledge—until they fail to work. At that point, our theories need to be corrected. Otherwise, like the startled, homeless earthquake victim, we will not be able to negotiate effectively through this world.
For more than two centuries, scientists and laypersons alike held Newtonian physics to be the ideal case of eternal scientific truth. They—unlike Newton himself—were convinced that his law-like generalizations were in fact laws. Moreover, it was commonly assumed that the reason for Newton’s success lay in the unerring ability of his system to produce verifiable hypotheses. Of course we now understand that Newton’s laws were, at best, partial truths. Recent developments in physics such as relativity theory, quantum mechanics, and the widespread acceptance of the principle of indeterminacy have helped us to understand the limitations of the classical theories. For, considering what is now known about physical properties at the sub-molecular and super planetary levels, it is clear that Newtonian mechanics apply only to a limited range of observable phenomena (the middle range). And, even for the kinds of phenomena for which the classical principles still appear to be valid, they are only approximately so under normal conditions.
These differences between the physics of the early eighteenth century and the physics of today have caused scientists to rethink the nature of the physical world. In addition, they have prompted philosophers and sociologists of science to examine critically the nature of empirical truth. 10 For example, it has often been remarked that newer discoveries such as quantum mechanics have “proved” that Newton’s laws were “false.” But this kind of observation is, at best, a metaphor. It would be more accurate to note that Newton’s laws were never “true” in the first place; they were only not (yet) falsified. More important, perhaps, we now understand that the enduring character of the Newtonian system is not the result of its unerring and unexceptionable generalizations. Instead, his system worked; it proved to be good enough. Good enough for what, we might ask? As Stehr (1992) has argued, it was good enough to form the basis of the Industrial Revolution. Although it was by no means absolute knowledge, it was extremely effective as practical knowledge.
In the context of discussing Newtonian physics, Stehr points to another instance of a scientific system whose veridicality is now in serious doubt, but which has nevertheless proved to be effective in practice. This is psychoanalysis.
It is perhaps unnecessary to review in detail the scientific critique of psychoanalytic “theory,” because virtually the entire program of contemporary behavioral psychology is built upon such a critique. However, it might be helpful to recall that psychoanalysis as developed by Freud and his students is held to be highly dependent upon phenomena that are unobservable: the libido, introjection, the unconscious, and the like. Because they cannot be observed, such entities cannot be the subjects of empirical sentences and, therefore, no authentic theories of them can be established or tested. In addition psychoanalytic “theory” tends to resolve to logically meaningless propositions. That is, a single formulation can be used to explain an observed phenomenon and its opposite. For example, the fact that a person was socialized by an overbearing father might be the cause of the person’s own aggressive behavior, but it might also be the cause of the person’s passive behavior. As critics would argue, one cannot have it both ways.
Although defenders of psychoanalysis dismiss such arguments as unfair and as based upon a misunderstanding of the approach, there is little doubt that psychology has changed dramatically over the past several decades—in part because of the research inspired by Freud’s observations. However, as in the case of Newton, these changes do not render Freud “wrong.” For it is not the absolute truth of his generalizations concerning libido and the unconscious that afforded him such a central role in the development of science. Rather, the importance of psychoanalysis stems from the fact that it works. Many thousands of people have been helped, perhaps in some senses even cured, by psychoanalysts—despite the potential weaknesses of the science on which it is based. Today, many of the psychological disorders once treated with psychotherapy are now approached as problems in psychopharmacology. But Freud and the Freudians first identified most of these disorders. We can now surely say that on nomological grounds what they developed was not a flawless science. However, it was practical knowledge that was good enough for several generations—and, in some respects, is still good enough today.
A final example of practical knowledge takes us into quite a different realm. This is the field of race science (Rassenwissenschaft) which was developed during the late nineteenth and early twentieth centuries principally in the United States, Germany, and England (see Weinstein and Stehr, 1999). Inspired by the Darwinian Revolution and strongly influenced by a centuries-long tradition of racial theorizing, the race scientists established and elaborated a program in which it was presumed that the behavioral, mental, and moral characteristics of individuals were essentially formed by heredity. Throughout this work, the key independent variable is race itself, defined as a discrete, sub-specific group of humanity. Among the multitude of race’s alleged consequences, two received special attention: intelligence (as measured by standard IQ tests) and criminality (as measured by frequency of arrests and/or convictions). Perhaps it goes without saying that Northern Europeans, “Nordics,” always proved to be the most intelligent and most law-abiding group. In contrast, Africans were always shown to be the least intelligent and most criminally deviant. Whereas Jews—classified by practitioners as a “race” with little dispute—displayed an especially sinister combination of high intelligence and low morality.
Of course, pockets of support for race science continue to exist in academic and intellectual circles today. 11 However, this doctrine is no longer considered to be in the mainstream, as it was some eighty years ago. In fact, to those willing to attend to the evidence, the program of race science has been repudiated. This includes evidence on the actual causes of intelligence and criminality and the questionable ties once believed to exist between “race” and other factors of interest to race scientists—such as musical ability, business acumen, capacity for hard work, and sense of humor. It also involves the inadequacy of indicators such as IQ scores and even the definition and measurement of race. Yet, as with Newton’s physics and Freud’s psychoanalysis, the success and popularity of Rassenwissenschaft never did depend upon its ability to lead one reliably and without exception to truths. Instead, it was practical knowledge; good enough to bring results in appropriate contexts.
Unfortunately, the context that turned out to be the most appropriate for the practical application of race science was governmental eugenics policy, especially but not exclusively that embraced by Germany’s Nazi government. If, race science proponents argued, certain races embody desirable inherited traits whereas other races are bound to carry in their gene pools obvious undesirable characteristics, is it not unjust to allow the latter groups to reproduce? For the sake of the population as whole, inferior races should be prohibited from contributing their flawed intellects and consciences to the general reproductive pool. And, as any demographer will attest, there are only two ways to achieve this: (1) force members of the inferior races to emigrate or (2) sterilize, prohibit from having intercourse, and/or kill members of those races before they can reproduce. Of course, in the name of racial hygiene and for the sake of the Nordic race, the Nazis used Rassenwissenschaft to justify the pursuit of exactly these policies—with considerable success.
Once more, it should be stressed, to those who would object that race science is not “authentic” science, the people who counted—within and outside of the academic community—believed that it was as authentic as any other field. And despite the fact that most rational people today give little credence to racial explanations, it proved to be good enough to underwrite a very effective program of population control we now call the Holocaust. Which is to say that the results of practical knowledge are by no means guaranteed to be beneficial—as they were in the case of Newtonian physics and Freudian psychoanalysis. If turned to evil ends, such knowledge can be highly effective in bringing about evil outcomes. Most disturbing of all, from the perspective of most Nazis and many “ordinary Germans,” as well, applied race science was meant to be beneficial (for them).
The contrast between the nomological and the pragmatic models, and the issues associated with the uses of knowledge raised by our discussion of race science, suggest a question posed some six decades ago by Robert Lynd: Knowledge for What? That is, what is our purpose in developing and applying scientific theory? As noted above, from the nomological perspective, which is essentially the position to which most contemporary academic sociologists subscribe, the answer is “for knowledge.” Yet, sociology aside, many pragmatists also accept this position. While granting that the ultimate test of our beliefs is in their application to real-world situations, one might still hold that the purpose of such testing is to establish which beliefs deserve to be maintained and which need to be revised. That is, the end of the pragmatic knowledge-creation process need not be to change the world but rather simply to create better knowledge—where “better” is defined in relation to effectiveness.
It is at this juncture that Lynd and others have argued on behalf of social scientific exceptionalism. For the aspects of the world that would be altered by a pragmatist sociologist in the search for knowledge differ from those of interest to a pragmatist non-social scientist. In particular, whereas a physicist might wish to establish how a physical theory applies to the operation of a machine, a sociologist would need to affect change in an organization, social relationship, or other part of the moral order. So, the argument goes, the social scientist has a special responsibility to operate in good faith by testing knowledge not for manipulative ends but rather to bring about an authentic improvement in the setting(s) employed for such knowledge tests. Although it is obvious that, on an abstract plain, “improvement” is a highly value laden concept, Lynd and the preponderance of sociologists who have preceded and succeeded him have in mind the movement toward more inclusive forms of policy formulation and decision making. That is, the brief but well-grounded answer to the question of “knowledge for what?” is “for democracy.” 12
Today it is perhaps easier to understand that non-social scientists, too, might have moral responsibilities. Following the events of World War II and the important educational work of the Bulletin of Atomic Scientists, most physicists are now overtly dedicated to promoting peaceful uses of atomic energy. Similarly, based on the discoveries of Rachel Carson, Barry Commoner, and other environmentalists, most contemporary biologists are openly concerned with the preservation of species and their environments. Thus, sociologists no longer are (if they ever were) the only scientists who routinely face ethically-charged theoretical problems.
Nevertheless, it seems to me that Lynd’s observations still apply to social science in a special sense. Looking back on the century that has just passed—and its clear lessons about what happens when social scientists allow others to monopolize social “engineering,” it appears at best pointless and at worst destructive to plead the case for “pure” sociology. Those who hold that their only interest in developing sociological theories is to generate more research to improve theories, etc., are either enormously self-involved, oblivious to the world around them, or intellectually dishonest. Even more important, the argument against a (democratically) committed sociology, and in support of the development of apolitical theory, is seriously discontinuous with the foundations as well as the authentic intellectual breakthroughs of the discipline.
Thus, when C.Wright Mills posed his now oft-cited opposition between abstracted empiricism and grand theory, on one hand, and intellectual craftsmanship, on the other, he was simultaneously addressing epistemological and ethical concerns. Strongly influenced by the work of Veblen and the Chicago School tradition (note the Veblenesque subtitle of his dissertation: “the Higher Learning in America”), Mills explicitly connected sociological approaches to theory with the sociologist’s moral commitments (Mills, 1969). In his view, the discipline was experiencing a period of moral bankruptcy in which practitioners routinely avoided engagement with such critical issues as the erosion of democracy in America. This stance was achieved and supported by a characteristic distortion of scientific activity, whereby empirical research and theory were pursued in isolation from each other. Empiricists did not theorize, but rather chose to let the facts “speak for themselves,” a la Bacon’s ant. Theorists did not ground their generalizations in empirical research, preferring instead, like the spider, to develop the all-encompassing—but empty—schema of theoretical sociology.
Mills’s alternative, intellectual craftsmanship (parallel to Bacon’s bee and, significantly, to Comte’s positivism), combines theory and research in such a way that effective law-like generalizations can be created, tested, sustained, and revised. This approach, Mills believed, is productive of knowledge in an authentic scientific manner, without inflated claims of absolute truth or nihilistic assumptions to the effect that knowledge is impossible to attain. Moreover, he felt that developing sociological theory in this manner would help to bring sociologists into intimate touch with the important social issues of our times.
It is at this juncture that another of Mills’s well known formulations emerge: the idea that it is the special task of sociologists to seek to connect biography and history. By placing the lives of individuals in historical context in a manner that only sound theorizing can do we are, Mills believed, better equipped to build a meaningful social science. Moreover, in the process we are also open to understanding and to contributing to the solution of the problems that stand in the way of achieving more fulfilling biographies and greater historical progress. On the other hand, as long as the development of real sociological theory is avoided, eschewed, or held to be a futile undertaking, it will be very difficult for sociologists to realize its potential as a source of enlightenment and a means of improving the human condition.
In this section we consider how the development of pragmatic (and pro-democratic) theory actually occurs. Obviously, it is not possible here to reflect on every last detail entailed in the production and application of practical knowledge. However, the illustrative case I have selected can help to emphasize those aspects of the process that clearly set it apart from the kind of theorizing that characterizes contemporary academic sociology. 13
The example is possibly the best-known applied sociological project ever undertaken in the United States and perhaps the world: the study of the socioeconomic status of African Americans (then called “Negroes”) in the several decades following Reconstruction. The main published work emerging from this study, An American Dilemma by Gunnar Myrdal (1944) and others, first appeared as a full-length—two volume—research report in 1944. 14 The most significant and consequential social transformation effected on the basis of this work was no less than the desegregation of the public schools in the United States. This action followed the decision of the U.S. Supreme Court in “Brown et al. V. Board of Education of Topeka et al.,” decided May 17, 1954.
Gunnar Myrdal concluded his comprehensive study of America’s “Negro problem” with the following observation.
The rationalism and moralism which is the driving force behind social study, whether we admit it or not, is the faith that institutions can be improved and strengthened and that people are good enough to live a happier life. With all we know today, there should be the possibility to build a nation and a world where people’s great propensities for sympathy and cooperation would not be thwarted. To find the practical formulas for this never-ending reconstruction of society is the supreme task of social science. The world catastrophe places tremendous difficulties in our way and may shake our confidence to the depths. Yet we have today in social science a greater trust in the improvability of man and society than we have ever had since the Enlightenment (1944: 1024).
I cannot think of a more appropriate four-sentence summary of the main argument of this paper. Yet, Myrdal did not by any means intend it as a statement about the role of theory in applied sociology. Rather, it is a personal reflection on several years of exhaustive social research by dozens of individuals and sponsoring organizations on one of the most perplexing questions of our time. That is, why and in what ways are Black Americans assigned an inferior status in their society and what can be done to rectify the situation? 15
Myrdal and his colleagues entered into this study armed with a substantial set of findings drawn from research conducted in several disciplines over the previous several decades. During the course of the fieldwork, additional quantities of data were produced via interviews, secondary data analysis, and documentary sources. But the study could not have had such significant impact if this information had not been organized into a coherent theoretical framework. And, as Myrdal acknowledged, a key component of this framework proved to be “the theory of the vicious cycle,” also known as the principle of cumulative and cyclical causation (Myrdal, 1944:75-78, 101, 109, 172, 207-209, 791, 794, 1011, 1065-70, etc.).
In the context of the study, this theory proposes that the maintenance of segregated facilities for whites and blacks is part of a self-sustaining process in which the dominant belief in the social inferiority of blacks is reinforced. Separate facilities have always meant unequal facilities, so that blacks are of necessity more poorly housed, educated, cared for medically, etc. than whites. The results of such treatment serve to support the belief in the inferiority of blacks (“they don’t care for their homes,” “they are ignorant,” “they are unhealthy,” etc.) and thus to justify further segregation.
In identifying the role that this theory was to play in the study, Myrdal (1944:75) stipulated that “Throughout this inquiry we shall assume [emphasis added] a general interdependence between all the factors in the Negro problem. White prejudice and discrimination keep the Negro low in standards of living, health, education, manners and morals. This, in turn, gives support to white prejudice. White prejudice and Negro standards thus mutually ‘cause’ each other.” Note that this process is at first “assumed,” with the understanding that only the facts can determine whether or not things actually occur in that manner. In a similar tentative vein, Myrdal (1944:78) observes that, to give some of the key concepts in this theory “clearer meaning—which implies expressing also the underlying social value premises—and to measure them in quantitative terms represents from a practical viewpoint a main task of social science. Fulfilling that task in a truly comprehensive way is a stage of dynamic social theory still to be reached but definitely within vision.”
The researchers thus began their study of race relations with the expectation that they would find that a certain dynamic, specified by the theory of the vicious cycle, underlay observed social inequalities in all of their manifestations. They also believed that, in the course of their work, they would be able to refine the theory, to give “clearer meaning” to their concepts, and to “measure them in quantitative terms.” As we now know, they were uniquely successful on both counts. The principle of cyclical and cumulative causation did indeed fit the facts; not, we have since learned, perfectly or completely, but certainly in a way that was “good enough.” This much is quite evident from reading An American Dilemma. And the theory was strengthened in several ways, through addition of previously unanticipated factors and the elimination of aspects of the process that proved to be irrelevant. Evidence of the how the theory was refined in these ways can be found, for example, in Myrdal’s later work on international stratification (Myrdal, 1968; 1970).
But Myrdal and his colleagues articulated one additional expectation that makes their work an especially effective model of practical knowledge. For they intended to use what they would learn about the Negro problem and, especially, about its cumulative and cyclical character, to improve the situation; in brief, to use their theory “for democracy” (e.g., Myrdal, 1944: 78-80).
If this theoretical approach is bound to do away in the practical sphere with all panaceas, it is, on the other hand, equally bound to encourage the reformer. The principle of cumulation—in so far as it holds true—promises final effects of greater magnitude than the efforts and costs of the reforms themselves.
That is, if indeed a vicious cycle is the main process sustaining the social inferiority of black Americans, then the undoing of this cycle will have tremendous payoffs in the realm of race relations and far beyond. If, in fact, segregated facilities produce unnecessary socioeconomic waste—because they are a cause of inequality and not a solution to it, then integrated facilities—especially in the realm of education—will unleash a substantial stock of unutilized productivity.
This interest in practical application was expressed in full knowledge of the legal framework of segregation, the Fourteenth Amendment to the Constitution, and relevant rulings by the United States Supreme Court (although the researchers include no explicit discussion of Plessy v. Ferguson 16 ; Myrdal, 1944: 480-81, 555, 601, 628-630, 633, 1110, 1231-32, etc.). Thus the authors were well aware of the revolutionary implications of suggesting and, later, demonstrating, that segregated schools and the like were having an effect the very opposite of that assumed and sanctioned by law. With the benefit of hindsight, then, it is no surprise that when the Supreme Court headed by Earl Warren essentially overturned this legal framework, the definitive argument was based on Myrdal’s findings. 17 In effect, the Court accepted the principle of cumulative and cyclical causation and, on that basis, ruled that separate schools violated the Constitutional rights, and specifically the Fourteenth Amendment rights, of African Americans.
We conclude that in the field of public education the doctrine of “separate but equal” has no place. Separate educational facilities are inherently unequal. Therefore, we hold that the plaintiffs and others similarly situated for whom the actions have been brought are, by reason of the segregation complained of, deprived of the equal protection of the laws guaranteed by the fourteenth amendment (Brown, 1954: 495).
It is now common knowledge that the Brown decision ushered in one of the most critical eras in the recent history of the United States, affecting relations between blacks and whites and much more. Some of the outcomes were anticipated and intended, and many were not. No sociologist would seriously argue that true equality between majority and minority groups has been achieved as a result of that ruling. But it is also unlikely that a serious student of social relations in this country would deny that the relative statuses of whites and blacks have been altered since 1954 and, to some extent at least, for the better.
In fact, even now, after nearly fifty years have passed, information routinely comes to light that reflects on the validity of Myrdal’s research and of his vicious cycle theory, in particular. We now know, for instance, that outlawing segregated public schools does not automatically bring about integration. As Myrdal anticipated, segregated housing is likely to continue to sustain de facto separate educational facilities. Programs of bussing to achieve integrated schools have met with mixed success. And, just as the Brown decision can be credited with increasing the proportion of integrated classrooms in some parts of the country, it is also responsible for contributing to the “white flight” that has left many of our central cities predominately black and our suburbs predominately white.
These, apparently, are the kinds of lessons that are learned when powerful social theory is put into practice. Authentic changes do take place, and theorists are challenged to account for all of them—not merely the desired results. Perhaps this is one thing Myrdal meant by “final effects of greater magnitude than the efforts and costs of the reforms themselves.” But I do not take such complexities as grounds for discouraging the application of theory. For what, after all, are our options? Who among us would prefer that An American Dilemma had not been published, or that the Brown decision had not thereby been rendered? If the soundest sociological theories, based on the most conscientious research possible do not inform public life and public policy, then less sound, less carefully construed formulas will simply take their place. As John Maynard Keynes noted, “The ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed, the world is ruled by little else. Practical men, who believe themselves to be quite exempt from any intellectual influences, are usually the slaves of some defunct economist” (Keynes, 1936: 383-84; also quoted in Weinstein and Stehr, 1999: 26).
It should be obvious that there is a subtext to this presentation. That is, like Robert Lynd, C.Wright Mills, and other—less famous-—sociologists, I believe that our current set of priorities that gives greater importance and a higher status to academic, nomological theorizing in comparison to the pragmatic model should be reversed. In my view, we are currently involved in a classic case of the tail wagging the dog. In part because of the prevalent misunderstanding among sociologists of positivism and their lack of sympathy for the founding conception of sociology as a kind of technology, a disdain of intellectual craftsmanship characterizes the dominant (but not necessarily larger) segment of our profession.
For those who agree with this assessment and who would like to realign our theoretical priorities, several strategies are available—some of which are, to an extent, already being pursued. One such strategy is, of course, to practice what we preach—and to preach what we practice. The more successful we are in demonstrating that knowledge that works “works,” the more likely we are to win support for applied approaches. Another fairly common strategy is to criticize our academic colleagues, pointing out the historical, ethical, and epistemological weaknesses in their search for Absolute Truth and/or their consequent abandonment of the search for any kind of truth. I for one believe that this approach is of limited effectiveness, and that it has a strong potential for alienating those who should be our allies. Such backlash effects are especially likely at this point in the history of the field. For many voices can now be heard from many ideological and methodological quarters claiming that sociology is in a state of crisis (with little agreement about what is meant by “sociology” and “crisis”).
Another strategy, which I think has considerable promise, is to seek converts. If my reading of the situation is in any way accurate, then many academics are “closet” applied sociologists. They work within the nomological framework—if they address theoretical concerns at all—because they believe they need to do so for the sake of academic success. As long as a record of articles published in a few elite journals remains the primary or even exclusive criterion whereby tenure and promotion are granted, most academics will be afraid to stray from abstracted empiricism and pseudo theory. This is the case despite the fact that most of them became sociologists in order to improve society (not merely to improve our curriculum vitae). Of course, early in the process of professionalization, one learns to suppress such “do-gooder” thoughts. But my suspicion is that the urge to develop and apply practical knowledge, to be a kind of Gunnar Myrdal in one’s own back yard, never really dies.
Moreover, I do not think that making a commitment to applied approaches, today, is tantamount to professional suicide. To state my reasons for believing this would take us far beyond the discussion at hand. But I do think that there is a key to “coming out,” to becoming the kind of sociologists we set out to be when we first declared our major. And this key lies with our students, undergraduates and graduates alike. For in their naivete concerning what constitutes an acceptable professional attitude, they entirely understand and sympathize with the search for applied theory and practical knowledge. Those students who are aware, or can be made aware, of the power they have (where would academic sociologists be without them?) have the capacity to exert pressure on their professors to admit to their own original motivations. For, in their naivete, they know that, in Myrdal’s words, “the rationalism and moralism which is the driving force behind social study, whether we admit it or not, is the faith that institutions can be improved and strengthened and that people are good enough to live a happier life.”
Ashley, David and David Michael Orenstein. 1998. Sociological Theory: Classical Statements. Boston: Allyn and Bacon.
Brown. 1954. “Opinion of the Court in the Case of Brown, et al. v. Board of Education, et al.” Syllabi of the Supreme Court of the United States, October Term, 1953. Washington DC (347 U.S. 483).
Comte, Auguste. 1896 . The Positive Philosophy of Auguste Comte, 3 vols. Trans. H. Martineau. London George Bell.
Comte, Auguste. 1912. Systéme de politique positive, 4 vols. Paris: Crès
Cromier, Harvey. 2000. The Truth is What Works: William James, Pragmatism, and the Seed of Death. Lanham, MD: Rowman &Littlefield Publishers.
Gilman, Sander. 1996. Smart Jews: The Construction of the Image of Jewish Superior Intelligence. Lincoln, NE: The University of Nebraska Press.
Glass, John. 1999. “Verstehen as Intervention.” Presented at the 17th Annual Meeting of the Society for Applied Sociology, Dallas, October 30.
Hernnstein, Robert J. and Charles Murray. 1994. The Bell Curve: Intelligence and Class Structure in American Life. New York: The Free Press.
Hookway, Christopher. 2000. Truth, Rationality, and Pragmatism: Themes from Peirce. New York: Oxford University Press.
Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. London: Macmillan.
Kuhn, Thomas S. 1970. The Structure of Scientific Revolution. Chicago: University of Chicago Press.
Lakatos, Imre and A.L. Musgrave, eds. 1970. Criticism and the Growth of Knowledge. Cambridge: Cambridge University Press.
Lynd, Robert. 1939. Knowledge for What? Princeton, NJ: Princeton University Press.
Mills, C. Wright. 1959. The Sociological Imagination. New York: Oxford University Press.
Mills, C. Wright. 1969 Sociology and Pragmatism: the Higher Learning in America. London: Oxford University Press.
Monroe, Kristen Renwick. 1996. The Heart of Altruism: Perceptions of a Common Humanity. Princeton, NJ: Princeton University Press.
Monroe, Kristen Renwick, M.C. Barton, and U. Klingermann, eds. 1990. “Altruism and the Theory of Rational Action: Rescuers of Jews in Nazi Europe.” Ethics (October): 103-22.
Myrdal, Gunnar (with the Assistance of Richard Sterner and Arnold Rose). 1944. An American Dilemma. New York: Harper and Brothers, 2 volumes.
Myrdal, Gunnar. 1968. Asian Drama: An Inquiry into the Poverty of nations. Harmondsworth, UK: Penguin.
Myrdal, Gunnar. 1970. The Challenge of World Poverty: World Anti-Poverty Program in Outline. Harmondsworth, UK: Penguin.
Nisbet, Robert. 1969. Social Change and History. New York: Oxford University Press.
Peirce, Charles Sanders. 1985. How to make our Ideas Clear/ Über die Klarheit unserer Gedanken.. Frankfurt am Main: V. Klostermann.
Peirce, Charles Sanders. 1998. The Essential Writings, ed. by Edward C. Moore. Amherst, NY: Prometheus Books.
Ritzer, George. 1996. Modern Sociological Theory. New York: McGraw Hill.
Ritzer, George. 1997. Postmodern Social Theory. New York: McGraw Hill.
Royce, Josiah. 1969. The Basic Writings, ed. by John J. McDermott. Chicago: University of Chicago Press.
Rudner, Richard. 1965. The Philosophy of Social Science. Englewood Cliffs, NJ: Prentice Hall.
Rushton, J. Philippe. 1995. Race, Evolution, and Behavior: A Life History Perspective. New Brunswick, NJ: Transaction Books.
Sanderson, Stephen K. 1995. Social Transformations. London: Blackwell
Scheffler, Israel. 1971. The Anatomy of Inquiry: Philosophical Studies in the Theory of Science. Indianapolis: Bobbs-Merrill.
Seidman, Steven. 1991. “The End of Sociological Theory: The Postmodern Hope.” Sociological Theory 11: 131-46.
Smith, Mark C. 1994. Social Science in the Crucible: The American Debate over Objectivity and Purpose. Durham, NC: Duke University Press.
Sorokin, Pitirim A. 1956. Fads and Foibles in Modern Sociology and Related Sciences. Chicago: H. Regerny, Co.
Stehr, Nico. 1992. Practical Knowledge: Applying the Social Sciences. Newbury Park, CA: Sage
Suppe, Frederick, ed. 1974. The Structure of Scientific Theories. Urbana: University of Illinois Press.
Turner, Jonathan H. 1993. Classical Sociological Theory: A Positivist’s Perspective. Chicago: Nelson Hall
Turner, Stephen P. and Jonathan H. Turner. 1990. The Impossible Science: An Institutional Analysis of American Sociology. Newbury Park, CA: Sage
Weinstein, Jay. 1996. “The Place of Democracy in Applied Sociology.” Journal of Applied Sociology 13, 1.
Weinstein, Jay. 1997a. Social and Cultural Change: Social Science for a Dynamic World. Boston: Allyn and Bacon.
Weinstein, Jay. 1997b “Leadership Training and Organization Building:” Social Insight 1, 1(April).
Weinstein, Jay and Nico Stehr. 1999. “The Power of Knowledge: Race Science, Race Policy, and the Holocaust.” Social Epistemology 13, 1: 3-36.
1. One of the most thorough, recent critiques of atheoretical (and apolitical) sociological thought on the grounds that it is "positivist" can be found in Ashley and Orenstein (1998). Here the authors establish and sustain throughout their long volume a clear opposition between good, correct sociology, on one hand, and "positivism," on the other. It would be difficult for students not well versed in the history of sociology (that is, the audience to whom the argument is directed) to avoid hating "positivism" upon completion of the book.
2. Here Comte is reaffirming Francis Bacon's admonition to modern scientists to avoid behaving like either the ant, who collects sticks to create a pile of sticks, or the spider, who spins its web entirely out of its own substanceInstead, scientists should emulate the bee, who collects pollen and combines it with her own substance to create honey. This is a quite common, and very sensible, view of the role of theory in scientific research. It is a puzzle as to how so many have failed to associate it with Comte's foundation works.
3. C. Wright Mills is probably one of the last sociologists whom contemporary critics of "positivism" would consider to be a positivist. Yet, in his critique of "abstracted empiricism" and "grand theory," he is merely echoing the warning of Bacon and Comte to avoid the ways of the ant and the spider, respectively. His preferred alternative, "intellectual craftsmanship," is precisely what Comte had in mind in The Positive Science. In fact, the extent to which Mills was successful in balancing theory and data collection is a good measure of the success of his important studies such as White Collar and The Power Elite.
4. This book (Comte, 1912), and the connection between it and The Positive Science may be even less familiar to sociologists than the Positive Science, itself. Yet, Comte was quite explicit in drawing on what he took to be sociological truths in arguing for the creation of a more rational and just social order. In brief, he was not a "pure" or academic scientist, clinically uninterested in the impact of knowledge on society; nor were the Scottish Moralists, Saint-Simon, Marx, or the other founders of the field.
5. Some years ago, I was inspired by a critique by Norval Glenn to observe that new causal theory is neither new, nor is it causal, nor - as pertains to this discussion - is it theory. Moreover, the Duncan-Blau "model" is not a model. Here, I think, is an instance - and a very well known instance in the field -- in which sociologists have misappropriated scientific terminology in an attempt to ward off physics envy.
6. John Glass (1999) of The Family Place has prepared some interesting observations on the use of Verstehen in applying sociology in clinical contexts.
7. I understand that this is a rather superficial critique of existing approaches. Of course, one can find extensive discussions about the problems and benefits associated with many of these "theories." But the purpose of this paper is not to delve deeply into these exchanges. However, as noted, Jonathan Turner has provided effective critiques in several forums. In the case of rational choice "theory," a considerable body of critical commentary has developed, including the work of Kristin Monroe, as discussed in the text. For an excellent critique of evolutionary "theory" along the lines suggested here, see Sanderson (1995).
8. See our discussion of this debate, including the contributions of Robert Nisbet, T. Dobzhansky, and Stephen Sanderson, in Weinstein (1997a: Chapter 2).
9. Mark C. Smith (1994) addresses this and related issues in his history of one of the most formative moments in the debate over the scientific status of the discipline.
10. An obvious and singularly acclaimed landmark in this recent inquiry into the character of scientific truth is Thomas Kuhn's argument concerning paradigm shifts. See Kuhn (1970) and his earlier critics Lakatos and Musgrave (1970), and Suppe (1974).
11. Recent "landmarks" in the revival - or continuation - of Rassenwissenschaft include Herrnstein and Murray (1994) and Rushton (1995). Sander Gilman (1996), among others, has taken explicit note of the strong similarities between the earlier and more recent attempts to connect race with behavior and mental capacities. Also see Weinstein and Stehr (1999: 26-27).
12. Some time ago I published an article in the Journal for Applied Sociology (Weinstein, 1996) that argues in favor of a close connection between sociology and democracy. After reading an early version of the article, Peter Rossi e-mailed me a note with the comment, "I basically agree with you, but the devil is in the details." That observation has served as an important reminder as I have continued to explore the sociology/democracy interface over the subsequent few years. It is increasingly obvious that the claim (which is restated in this paper) to the effect that applied sociology does have a pro-democratic orientation - even a "mission" -- is prone to all manner of foibles. These range from empty abstraction to meaningless minutiae. And it is apparent that the claim is capable of offending just about everyone across a wide ideological spectrum. Despite all of this, I continue to support the view. For a more extended discussion of this claim, see Weinstein (1996; 1997a: Preface).
13. A second case which I had considered including is much more recent and far more localized. Known as "the Taylor Community Action Study," this project sought to discover the sources and contribute to the mitigation of serious social disorganization
in a distressed neighborhood in the Detroit suburb of Taylor, Michigan. Several reports emerged from this study, including a summary article in the journal Social Insight (Weinstein, 1997b), published by The Society for Applied Sociology.
The main social transformations effected on the basis of these reports were the creation of a community services center in the neighborhood and the purchase and renovation of the residential property by the City and a selected developer.
Although these two cases differ considerably in scope, impact, and several other ways, they do share at least one set of features that relate closely to the theme of this paper. That is: 1. both were undertaken with an explicit interest in testing theories considered to be relevant to the problems at hand, 2. both applied the theories in a social action framework, and 3. both entailed substantial planned social changes that shed light - both supportive and critical - on the theories thus employed.
14. Some contemporary scholars have been inclined to dismiss Myrdal's study as not authentic sociology because the principal investigator was widely known for his contributions to "economic science" (the field in which he was awarded the Nobel Prize). Others have been reluctant to grant the work its due status because of its overt sympathies with the plight of its subjects. These critics might be surprised to know the names of some of the other contributors, a partial listing of whom includes: John Dollard, W.E.B. DuBois, E. Franklin Frazier, George Lundberg, Frank Notestein, Howard W. Odum, Robert E. Park, Edward Shils, W.I. Thomas, Jacob Viner, and Louis Wirth.
15. Upon reading this enormous book, written more than fifty years ago, it becomes increasingly clear that it anticipated a very large share of the "discoveries" in the field of race relations now considered to be current. If ever there was an instance of what Sorokin (1956) referred to as "new Columbuses," proliferating beyond control, it is this.
16. As Chief Justice Warren pointed out in his opinion, "the doctrine of separate but equal did not make its appearance in this Court until 1896 in the case of Plessy v. Ferguson, supra, involving not education but transportation "(Brown, 1954: 483). Thus, the Court was expressing a common view that Plessy was irrelevant and presented by the defense in a desperate attempt to generalize on its limited and always questionable applicability. This view is reinforced in the precis, as follows: "The 'separate but equal' doctrine adopted in Plessy v. Ferguson, 63 U.S. 537, has no place in the field of public education."
17. The Court's references to the social science literature appear verbatim as footnote 11 of the opinion, as follows (Brown, 1954: 494-95).
18. K.B. Clark, Effect of Prejudice and Discrimination on Personality Development (Midcentury White House Conference on Children and Youth, 1950); Witmer and Kotinsky, Personality in the Making (1952), c. VI; Deutscher and Chein, The Psychological Effects of Enforced Segregation: A Survey of Social Science Opinion, 26 j. Psychol 259 (1948); Chein, What are the Psychological Effects of Segregation Under Conditions of Equal Facilities?, 3 Int J. Opinion and Attitude Res. 229 (1949); Brameld, Educational Costs, in Discrimination and National Welfare (MacIver, ed., 1949), 44-48; Frazier, The Negro in the United States (1949) 674-681. And see generally Myrdal, An American Dilemma, 1944 (our emphasis).
Theory & Science