Empiricists versus social constructionists: More fuel on the flames--Comment/reply

Thyer, Bruce A

EMPIRICISTS VERSUS SOCIAL CONSTRUCTIONISTS: MORE FUEL ON THE FLAMES

Bruce A. Thyer

In the December 1993 issue of Families in Society, Charles Atherton's article, "Empiricists versus Social Constructionists: Time for a Cease-Fire" (pp. 617-624), reiterates the argument that social work professionals need to respect diverse methodologies in the conduct of practice and research as well as the philosophical perspectives that undergird them. He deplores the sometimes strident voices touting the virtues of one approach over another (e.g., logical empiricism over social constructivism or vice versa). He further contends that the value of subjectivist versus objectivist perspectives cannot ultimately be resolved philosophically but that "empiricists have the weaker position" (p. 622). Because Atherton misrepresented my writings on these matters in various places throughout his article, I have chosen to respond here in order to shed further light on these important issues. First, I will discuss Atherton's misrepresentations of my work and that of like-minded social workers, after which I shall provide some additional commentary on these issues.

MISREPRESENTING THE EMPIRICIST'S POSITION

Atherton states, "Thyer evidently believes that there is a scientific basis for the claim that a measuring instrument has face validity" (p. 620), then cites several authors who support the view that the determination of an instrument's face validity is to a certain extent a matter of subjective appraisal. Atherton referred to an article of mine (Thyer, 1989) as being "patronizing" (p. 620) because of the simplicity of its content. Apparently it was not simple enough for Atherton to understand. I addressed the issue of face validity in only one place in this article:

Validity refers to the extent to which an assessment method actually provides a "true" measure of what it is supposed to measure. Validity of measurement is expressed in a variety of ways (face, content, criterion, construct, discriminant, factorial, predictive, etc.), some of which are qualitative (for example, face, content) whereas others are expressed in a validity coefficient (Thyer, 1989, p. 315, italics added).

Clearly, the above quote shows that the appraisal of a measuring instrument's face validity is not a rigorously scientific process (I even use the word "qualitative" in a nonpejorative manner!). I do not understand why Atherton misrepresents my writings on this point.

Atherton also contends that

Thyer erred in his claim that because random samples are difficult to come by, a realistic compromise is to take random samples from larger convenience samples. However, such an approach merely allows one to generalize to the original convenience samples, which is not an improvement" (p. 620, italics added).

Here is what I actually said about this point:

Truly random selection of clients is often difficult in field studies but various compromises may be made (for example, random selection of clients from a larger convenience sample), in order to approximate the goal of obtaining a representative sample of clients (Thyer, 1989, p. 317).

The fact is that NO ONE, to my knowledge, has ever employed a truly randomly selected sample of clients (or families) chosen from a larger population of interest, and conducted an outcome study using such clients. The actual practice is almost always to employ a convenience sample, selected on the basis of availability (such as clients seen at your agency). One may not generalize the findings from such studies to other clients with similar problems, no matter how large the sample size (Thyer, 1992, p. 145, italics in original).

If one wishes to obtain valid information about clients at a given agency that is generalizable to all the clients in that agency, a random selection of the agency's clients is more effective than a nonrepresentative (i.e., nonrandomly selected) sample. Admittedly, this is a relatively simple principle for guiding practice research; I don't understand why a scholar of Professor Atherton's attainments should find fault with this principle.

Atherton refers to another article of mine (Harrison, Hudson, & Thyer, 1992) as a "hatchet job" (p. 620). The article to which he refers was a response to Witkin's (1991) contention that "virtually any intervention can be justified on the grounds that it has as much support as alternative methods" (p. 158). We responded as follows:

We believe that a major component of professional social work training should consist of educating students to seek out...and apply as a first treatment option those methods with greater degrees of empirical support than alternative interventions. To have them do less constitutes, in our opinion, an abrogation of our professional responsibility to the field and borders on a violation of professional ethics (Harrison et al., 1992, p. 462).

To contend that the social work profession has no standards by which to judge the efficacy of its services and that, therefore, "anything goes," is a destructive point of view that undercuts the progress that has been made documenting effective therapies and welfare programs. In this article, we pointed out Witkin's earlier writings, praising the virtues of single-system and group-research designs and questioning Witkin's failure to provide direct quotes to document the misrepresentations he made regarding the tenets of empirical clinical practice. Readers of Families in Society may examine the articles in question and determine whether our response to Witkin was unfair or whether it rightly highlighted flaws in his article.

Atherton goes on to chastise Harrison et al. (1992) for ignoring two "facts": "(1) social constructionists have not claimed that qualitative methods are superior, just different, and (2) it is difficult to publish qualitative studies, because they are often refereed by empiricists" (p. 620). With regard to the first point, Atherton cites Pieper's (1985) views, which are largely devoted to proclaiming the advantages of qualitative research methodologies over quantitative ones, on the "outmoded, discredited assumptions" of the "pseudoscientific approach" (p. 619). In fact, many qualitative articles and books that advocate so-called alternative methodologies begin by disparaging quantitative approaches (see Guba Lincoln, 1981, 1985). Regarding Atherton's second point, I'm not sure what journals Professor Atherton attempts to publish in, but Glisson and Fischer's (1987) survey of the major social work journals found that less than 45% of articles presented the findings of empirical research. Unlike the fields of sociology, psychology, psychiatry, urban studies, education, and other human services, our practice literature is dominated by qualitative and other nonempirical reports. I suspect that the well-designed empirical study has a more difficult time getting published in social work journals than does its nonempirical counterpart, hence the decision of many social work practitioner-researchers to publish their work in the journals of other professional disciplines.

Atherton praises (p. 621) Hartman's (1990) editorial in which she states, "there are many truths and there are many ways of knowing" (p. 3). However, Atherton fails to acknowledge the most important way of knowing, distinguishing fact from falsehood and truth from fantasy. Whereas empirically oriented research methods provide some guidance in this regard, qualitative methods do not, inasmuch as they lack clear standards for determining "truth." Indeed, questions such as whether a "pro-arrest" policy deters domestic violence, an increase in the cigarette tax decreases tobacco consumption and improves public health, or chronically mentally ill clients being treated at a community mental health center have higher social functioning and less hospital recidivism than do clients who do not receive such services, all of which serve as grist for the empiricist research mill, are not amenable to inquiry via qualitative methodologies. Yet, our profession is charged with answering these very questions. Even if qualitative methods might help us understand what it "means" to be homeless or to "make sense of the alcoholic's experience," is that what our profession is all about? Social workers are charged with designing social programs to help the homeless find safe, affordable, and long-term housing, not with understanding them as they walk the streets. Federal and state legislatures do not allocate millions of dollars to help social workers make sense of drug abuse, but offer funding to prevent and deter substance abuse. The evaluation of programs and services that work toward such goals clearly falls in the domain of quantitative methodologies. Nevertheless, as I stated in an earlier article, "Empirically oriented researchers will be delighted if and when the methodologies of qualitative research demonstrate their utility in providing our profession with clearly effective solutions to problems of social importance" (Thyer, 1989, p. 321). Such demonstrations have not been forthcoming, however, and Atherton's reiteration of stale arguments does not advance discussion.

In his article, Atherton misrepresents the positions of logical empiricists, yet does not provide a single quotation to support his views. Here are some of these false statements, followed by more accurate portrayals of the empiricist's stance.

MISREPRESENTATION ONE

Thyer and his supporters, in upholding logical positivism as the exclusive way to the truth, are defending a false god (p. 622).

In actuality, I have written the following:

For the determination of social work effectiveness there is no substitute for controlled experimental research guided by the philosophy of science known as logical positivism and the tenets of the hypothetico-deductive process....I urge the advocates of qualitative methods to set about the serious business of conducting practice research on problems of social importance (Thyer, 1989, p. 320-321, italics added).

My impression is that the major role of qualitative methods lies in their ability to help the practitioner-researcher generate testable hypotheses about family dynamics. Quantitative methods afford very little help in this regard--the construction of hypotheses (Thyer, 1992, p. 143, italics in original).

Empirical research guided by the principles of logical positivist inquiry is the most reliable means of determining the efficacy of a given social service program or method of intervention....Logical positivism does not denigrate the value of clinical impressions or of intuition in the initial stages of any line of scientific inquiry....Often such sources provide brilliant new insights and may even lead to important advances (Thyer, 199, p. 17, italics added).

Does our profession have room for other philosophies, theories, and practice and research methods? Of course! I have even published a number of qualitative case studies drawn from my own clinical work. Is it appropriate to contend that the methodologies of empirical science have been very useful for our field? Of course! Is it appropriate to contend that alternative approaches have yet to prove their worth with regard to evaluating the outcomes of social work services and that until they demonstrate their value we should expand and further develop training in and adoption of empirical methods? Surely. I challenge Atherton to provide a quote from a single social worker who believes that empirical research is the sole avenue to developing our knowledge base. Are empirical methods the best approach available now? Yes! But neither I nor other empirically oriented social workers contend that this will always be the case. "Best" does not mean "only." We welcome all advances in the field, whether they be quantitatively or qualitatively derived. Empiricists and qualitatively oriented social workers have always agreed upon this point. For Atherton to contend that empiricists claim to have a monopoly in the search for truth is both inaccurate and destructive to informed debate on the subject.

MISREPRESENTATION TWO

The social world...is seen by empiricists as an objective entity...[that] is not affected by the personal values of the person doing the measurement: that is, measurement is "value-free" or objective (p. 618).

I and others have written the following:

Logical positivists recognize that the process of scientific inquiry is not completely neutral or value free...although critics of this approach often accuse them of this....However, the methods and procedures associated with logical positivism are intended to reduce as much as possible the potential for such personal biases to result in enduring erroneous findings. Such efforts are by no means free from all flaws, but like the imperfect nature of democratic government, the alternatives are seen as less desirable or useful for theory testing and practice evaluation in social work (Thyer, 1993, p. 22, italics added).

Also see Gorenstein's (1986) letter, in which he correctly notes that

it makes no sense to reject the potential scientific import of a construct simply because social values may have played some role in its formulation. The question of whether a construct has any scientific import is an empirical one. It has to do with whether the construct exhibits lawful properties (p. 589).

The purpose of reliable and valid measures is to control and reduce to the extent possible subjective bias when measuring psychosocial phenomena and the outcomes of practice. Can such bias be eliminated entirely? In most cases, probably not. Nevertheless, findings obtained from more reliable and valid measures are more likely to approximate "nature's truth" than are results obtained from less valid and reliable measures and procedures. I challenge Atherton or others with similar views to provide a direct quote from any social worker with an empirical orientation who has claimed that clients or their life circumstances can be measured perfectly, completely objectively, or in a value-free manner.

MISREPRESENTATION THREE

Social work researchers and practitioners accept a linear cause-and-effect approach to research and practice (p. 621).

Atherton's claim that the search for linear cause-and-effect relationships is the primary characteristic of empirically oriented social work research reflects a lack of familiarity with elementary social work research textbooks, most of which describe various approaches to quantitative inquiry. Much empirical research is descriptive in nature, as in observational or actuarial studies using valid and reliable measures. For example, the national census is an empirical database used by social work researchers to generate practical descriptive knowledge. The census methodology has some problems; it is not perfect or even value free. Nevertheless, it yields good, useful information about people and the circumstances in which they live. Some empirical researchers use correlation research methods to test theory and multiple-regression and other advanced statistical methods to obtain useful knowledge about nonlinear relationships, some of which may be causal. However, the search for causality is only part of the empirical research enterprise. Sometimes, cause-and-effect relationships are linear, but most often relationships are more complex. Currently, quantitative research methods are the best method for examining such relationships, even those involving multiple determinants.

MISREPRESENTATION FOUR

They have made n pseudoreligion out of science (p. 617).

To deride a particular philosophical position and orientation to social work research as pseudoreligious constitutes an ad hominem argument that attacks the person rather than his or her expressed views. Such attacks are a personal insult to empirically oriented social work professionals who maintain devout religious beliefs. As Atherton correctly notes, "Truth is always subject to further review, as are methods" (p. 622, italics in original). As I have noted,

Published errors in research design, assessment methods or statistical methods will be promptly (and rightly) seized upon by one's colleagues, requiring further explanation, modification or retraction. In this manner, practice research becomes a self-correcting process....This self-corrective feature is not characteristic of qualitative or impressionistic analyses, where the consumers of practice research have no way of judging the adequacy of a researcher's conclusions (Thyer, 1989, p. 320).

In the past decade, single-system research designs and meta-analyses have been useful additions to our methodological armamentarium. Scientific truths change, as new empirically supported data are generated; research methodologies are constantly evolving. Such adaptive features are not characteristic of religion.

CONCLUSION

While decrying the acrimonious wrangling of others, Atherton's article merely adds damp fuel to the fire, producing a lot of smoke but feeble light. The issues he raises are capable of resolution. The solution is simple and is found in the last sentence of the abstract of the article he extensively misrepresented: "The advocates of qualitative research are urged to provide the profession with...positive examples of research on the outcomes of social work practice, and to develop explicit guidelines for the conduct of qualitative studies" (Thyer, 1989, p. 309). Advocates of quantitative research have been providing both for decades. Their methodologies have been useful and the knowledge gained invaluable. "Critical diatribes which ignore the valuable progress which is being made in practice research and denigrate logical positivist methods of scientific inquiry in the absence of reasonable alternative approaches merely exert a divisive influence among social workers and fail to provide the profession with the positive examples of practice all of us can learn from" (Thyer, 1989, p. 321). Atherton's article exemplifies a critical diatribe that fails to advance discussion. It is time to develop the qualitative methodologies advocated by Atherton and others, conduct practice research using such approaches, and publish them in our professional journals where they can instruct and be scrutinized by everyone.

Atherton, C. R. (1993). Empiricists versus social constructionists: Time for a cease-fire. Families in Society, 74, 617-624.

Glisson, C., Fischer, J. (1987). Statistical training for social workers. Journal of Social Work Education, 23, 50-58.

Gorenstein, E. E. (1986). On the distinction between science and valuation in the mental health field (letter). American Psychologist, 41,4880-490.

Guba, E. G., & Lincoln, Y. S. (1981). Effective evaluations. San Francisco: Jossey-Bass.

Guba, E. G., & Lincoln, Y. S. (1985). Naturalistic inquiry. Beverly Hills. CA: Sage Publications.

Harrison, D. F., Hudson, W. W., & Thyer, B. A. (1992). On a critical analysis of empirical clinical practice: A response to Witkin's revised views. Social Work, 37, 461-464.

Hartman, A. (1990). Editorial: Many ways of knowing. Social Work, 35, 34.

Pieper, M. H. (1985). The future of social work research. Social Work Research ad Abstracts, 21(4), 3-11.

Thyer, B. A. (1989). First principles of practice research. British Journal of Social Work, 19, 309-323.

Thyer, B. A. (1992). Promoting evaluation research in the field of family preservation. In E. S. Morton & R. K. Grigsby (Eds.), Advancing family preservation practice (pp. 131-149). Newbury Park, CA: Sage Publications.

Thyer, B. A. (1993). Social work theory and practice research: The approach of logical positivism. Social Work and Social Sciences Review, 4, 5-26.

Witkin, S. L. (1991). Empirical clinical practice: A critical analysis. Social Work, 36, 158-163.

A GREAT DEBATE, BUT LET'S CALL IT QUITS

Howard Goldstein

I'm not sure that the social work factions Atherton calls empiricists and social constructionists are involved in "battle" and that therefore a "cease-fire" is required, although the scholarly commotion stirred up a few years ago by such worthies as Walter Hudson and Bruce Thyer on one side and Martha Heineman-Pieper and Stanley Witkin on the other was certainly rousing. Their polemics usefully served to polarize and therefore crystallize ideas and predispositions that had been loosely floating around in the curricula and literature of social work. Moreover, the sometimes overheated give-and-take discredited the aphorism that the reason academics fight so fiercely is that nothing serious is at stake. The partisans proved that something serious, indeed, the development of professional education and practice, was at stake. Atherton observes that both sides have retreated to their respective comers, from which one hears only echoes of their earlier manifestos. For reviving this important controversy in such an astute and penetrating fashion, Atherton deserves appreciation and commendation.

Atherton succeeds in making transparent and accessible various contentious and, he adds, often arcane positions that have consumed countless pages of articles and chapters in the professional literature. He clarifies important terms that have been abused by the various disciples of one or the other point of view (including a very helpful clarification of logical positivism), concluding that "the discussion has degenerated into what can be best understood as a theological controversy" (p. 622). This is the key point that I wish to underscore: the dispute is after all not a "battle" but a canonical struggle focused on the faith, the rules, and the practice of research--indeed, the practice of social work itself. As such, I contend that the struggle does not lend itself to compromise or resolution, nor should it.

Atherton hopes this breach can be healed and properly suggests that we not lose sight of our professional obligation to help our clients. He notes the few attempts at reconciliation and, for some reason, includes the question I posed (Goldstein, 1992) about social work knowledge; that is, how might the form and substance of social work knowledge appear had the profession identified itself more with the arts and humanities in its formative years than with the sciences? Atherton treats my musings too literally when he concludes that because the profession didn't choose the former path, it is pointless to think about what might have been, thus implying that when you choose one, you give up the other forever.

My question was meant to be metaphorical and certainly skeptical about the conceptual and educational routes the profession has taken. Moreover, I wished to suggest that such decisions were not inevitable or irrevocable, that we are indeed free to reconsider, reject, or revise our professional disposition and direction.

Social work is not a science such as, say, genetics or subatomic physics, whose theoretical foundations are unequivocally grounded in proofs and facts and, in general, progress in accord with systematic and incremental conventions. Social works mission, approach to people, and concern with problems, values, and meanings mark it as a humanistic enterprise. Its locus is not a laboratory or some experimental station but rather an inconstant and unstable social landscape. Social work can and should reflect on its progress and on how it responds to social changes and their dynamics. The physical sciences can't return to the concept of, say, ether as an explanation for how light is transmitted through space, but social work can reconsider and reflect upon, for example, its neglected ties with social reform and the principles of the settlement house movement as instructive elements of its history. Specht and Courtney (1994) argue persuasively that social work must rethink its courtship with psychoanalysis and the medical model--with its keen eye for psychopathology and certainty that behavior can be quantified--as well as its effort to achieve "respectability" by wedding with science.

How do we rethink our professional role and purpose? Atherton leans toward but does not have high hopes for reconciliation between the empiricists and humanists, even as he commends Hartman's (1990) statement that there are "many ways of knowing." Even if reconciliation is possible, is it desirable? The last momentous episode of reconciliation in social work occurred in the 1950s (Kasius, 1950), when the Diagnostic School, based on Freudian determinism, and the Functionalist School, guided by Otto Rank's beliefs in personal creativity and will, drew together and created a hybrid, more amorphous, and certainly less distinguished body of ideas than either of its original parents.

Serious ideas, beliefs, and doctrines that bear on our work with and our studies of human circumstances and behaviors need not be compromised or blended into a less precious alloy. At the same time, ideologies and schools of belief should not strive to dominate or diminish the convictions of other schools in the kind of struggle depicted so clearly by Atherton. Such theories and ideas are not merely functional tools that serve a practical purpose. Rather, they are often strongly held convictions, personal creeds, that, as patterns of the real world and representations of how we construct reality, have a certain doctrinal warrant.

The theories and systems of social work, whether in research or practice, are based on sound, rational, and objective foundations that can be communicated and taught and that shape social workers' identity and knowledge base. For example, whether we refer to the random or stratified sampling associated with quantitative-statistical research or the more conditional theoretical sampling of qualitative modes, both methods need to be understood in terms of their respective applications and shortcomings.

Few such theories and ideas, however, remain value-free and impersonal. Perhaps this is how it should be. It seems fait to say that a number of the constructs and methods that we feel we "own" and are part and parcel of our professional expertise are tied to our inner beliefs and ideologies regarding our world. In fact, Atherton's article shows us how readily intellectual differences can turn into crusades. We inherit and are subject to the "many ways of knowing." We may call ourselves "idealists," "realists," or "pragmatists"; we might feel compelled to search for basic immutable truths or feel at ease with relativities; rigorous analysis may suit some, intuition others.

What we learn, inherit, and acquire from the outside is inseparable from our innate ways of knowing and our particular cognitive styles. One's notion of what is "out there" and "in here" is affected by one's frame of reference and modes of perception and thought (and, of course, shared cultural conventions and language) (Goldstein, 1981, 1984). Discomfort, if not panic, surges when someone opposes our beliefs or values or, even worse, proves them wrong. Such strife is inevitable given the differences in human cognitive styles; the convergent thinker will come away from an event with one set of impressions, the divergent thinker with another. All things being equal, this might explain why some researchers are comfortable reducing a complex event to a confluence of definable variables and numerical quotients, whereas others seek to open the same event to more diverse, qualitative, and perhaps provisional meanings.

Gardner (1983) proposed that our cognitive orientation to the world is also shaped by the nature and type of our intelligence, showing that various forms of intelligence characterize individual talents, interests, and modes of thought and understanding. These include linguistic, musical, logical-mathematical, spatial, bodily-kinesthetic, and personal forms of intelligence. In a recent book, Gardner (1993) provides a fascinating study and analysis of the biographies of seven creative persons, each of whom expressed one of these forms of intelligence. Underscoring the importance of individual styles and realities, the following excerpt compares the creative minds of two men, Sigmund Freud and Albert Einstein, of the same era and similar backgrounds. Both men devoted their life to developing a unique system or school of thought.

Although Einstein might have come up with Freud's discoveries, or Freud's with Einstein's, it is not likely...these men had different kinds of minds and operated with different kinds of symbol systems. Freud was strong in the linguistic and personal intelligences: A shrewd observer of human nature from early life, he thought primarily in terms of language and built a system that was primarily linguistic-conceptual, with hardly any spatial or logical content....In sharp contrast, Einstein's linguistic skills were modest and his interest in the personal sphere strictly limited...his thought was rich with visual-spatial images and possible experiments. He could readily relate these experiments to mathematical formalisms and to concepts that existed within a tight logical-mathematical structure (Gardner, 1993, p. 133).

Freud and Einstein also had the temperament and strength of personal convictions to create systems of thought that contradicted conventional knowledge systems and that transcended problems already posed by others. Obviously, few people possess such genius; nevertheless, one wonders if more modestly creative minds fail to radiate their insights because they do not conform to prescribed educational or collegial codes.

Atherton confines the debate to two camps--empiricists and social constructionists. I propose that the roots of this controversy are both deeper and more diverse. Proponents of each position represent a unique world view, a particular construction of reality. To oversimplify for purposes of comparison, one seeks truths and facts through objective analysis; another finds meaning in subjective experiences. Each approach has its own logic, ideology or philosophy, and cognitions and intelligence. To say which constitutes the "truer" reality or which better fits social work's program is as pointless as trying to determine whether Picasso's Guernica or Hemingway's For Whom the Bell Tolls better represents the Spanish Civil War.

It was a great debate but it is time that we take what is useful from the feud and call it quits. Coherent versions of human experience are not necessarily subject to comparative evaluation or reconciliation. In fact, if we insist on enforced "harmony," we may end up with mediocrity. Both the empiricists and social constructionists make contributions to social work knowledge and practice. The principled, well-rounded, and competent social worker needs to draw upon both the scientific and the humanistic-artistic dimensions of knowledge and practice.

This alternative suggests certain propositions. First, it calls for an opening of curricula to other ways of knowing, thinking, and teaching--a shift from a prescribed and didactic pedagogy to a more collegial, inductive, and experiential mode. Educational programs must avoid being patronizing (e.g., "students are trained to be competent in the use of empirical methods, but we do offer a qualitative research course" or vice versa).

Second, to enlarge Atherton's proposal, courtesy and respect must replace the clamorous polemics of both camps. The differences between empirical and qualitative approaches need to be made clear in order to acknowledge confounding eccentricities in social work thought and education and to avoid the casual, sometimes glib, ways scientific concepts are interchanged with metaphysical or value-based precepts (e.g., the notion that linear causality or determinism and self-determination have some equivalence).

Finally, students deserve a collegial educational climate that respects the world views, visions, and values that animated their choice of social work as a career. At the same time, professional education should challenge their assumptions in an effort to open their minds and create wiser and more perceptive commentators, critics, and enablers of social work's progress.

Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic Books.

Gardner, H. (1993) Creating minds. New York: Basic Books.

Goldstein, H. (1981). Social learning and change. Columbia, SC: University of South Carolina Press.

Goldstein, H. (Ed.). (1984). Creative change. New York: Tavistock.

Goldstein, H. (1992). If social work hasn't made progress as science, might it be an art? Families in Society, 73, 48-55.

Hartman, A. (1990). Editorial: Many ways of knowing. Social Work, 35, 34.

Kasius, C. (Ed.). (1950). A comparison of diagnostic and functional casework concepts. New York: Family Service Association of America.

Specht, H., & Courtney, M. (1994). Unfaithful angels. New York: Free Press.

ATHERTON'S RESPONSE

I am enormously flattered to have drawn so much attention, although I have mixed feelings about accepting Professor Goldstein's thanks for reviving the controversy. I had hoped that, since everyone has said what he or she had to say, we could go on to other things. Clearly, this was wrong. Cynically, I must conclude that authors and editors think that they have found a good thing here and are reluctant to let go of it.

I included Goldstein's question about social work knowledge because a reviewer recommended that I do so. I've been at this game long enough to know that when a reviewer "recommends" including something, an author should jolly well include it. I enjoyed reading Goldstein's 1992 article, but I didn't see how it helped resolve the controversy. And, yes, I did imply that because social work did not identify more with the arts and humanities in its beginnings, it could not turn back. His response clarifies his position for me: While we may not be able to rewrite our past, we can still be informed by it. I agree that reconciliation may not be desirable and thought I made it clear that I believed reconciliation was impossible because of the underlying philosophical differences between the two approaches.

I don't see anything to be gained--except, perhaps, another publication--by engaging Professor Thyer (or anyone else) in a protracted search through each other's literary entrails looking for signs and portents. Nor has the journal allotted me the space to do it! So my response will be limited to this.

Perhaps the place to start is with Professor Thyer's assumption that his 1989 article "provoked Atherton's ire." Had he written "provoked Atherton's amusement" he would have been nearer the mark. He missed my point altogether. I support the use of the scientific method. I am opposed to what I think is the unseemly, peevish tone of the argument and the tediousness of it. Nothing new is being said.

All of my published research (excluding articles based on library research) involves number-crunching, except for one piece done more than 30 years ago that might be considered an ethnographic study. I call attention to two sentences in my article:

Empiricism has had a positive impact on social work research and practice. Perhaps the major contributions of empiricism are clearly defined rules for building professional knowledge and identifying demonstrable outcomes for judging practice effectiveness (p. 618).

And before someone finds it, I wrote elsewhere:

It [the scientific method] is the best means that human beings have devised for providing factual and dependable answers to questions requiring factual and dependable answers (Atherton 6 Klemmack, 1982, p. 4).

And in the same book:

The scientific approach assumes that things can be described in ways that permit some kind of measurement. If a phenomenon cannot be measured, then it is not susceptible to scientific analysis, although it could perhaps be understood in some other frame of reference (p. 13, italics added).

I still believe this. I hedged in the statement above because I believe also that many important things (e.g., the beauty of Mary Cassatt's paintings or Tom Glavine's slow curve) are not susceptible to scientific analysis, but are worth talking about in another frame of reference.

Two of the "misrepresentations" to which Thyer refers are technical points, which I am willing to discuss, just this once. My bemused comment about Professor Thyer's reference to face validity was stimulated by my surprise that he mentioned face validity at all in an article urging us to do scientific research. I didn't "cite several authors supporting the view that the determination of an instrument's face validity is a matter of some subjective appraisal." I cited only one, Anne Anastasi, whose book, now in its sixth edition, has been a standard text in psychological testing since I was a student during the late Medieval period. Anastasi pointed out that face validity isn't verifiable by criteria external to the instrument and that it mainly serves the public relations function of making the subject or respondent feel good about responding. Nearly 25 years ago, Cronbach (1970) made the same point. Thyer should either have omitted discussion of face validity or made it clear that face validity has no scientific basis but is entirely cosmetic in function.

Professor Thyer is quite right in pointing out that genuinely random samples are thin on the ground and that social researchers almost always have to use convenience samples. He is also correct in saying that one can't generalize from a convenience sample, no matter how large the sample. But taking a random sample from a convenience sample is no improvement, because one still can't generalize to the universe in question. What most people do, now that they have computers, is shove in the numbers from the entire convenience sample. Although such data are still not generalizeable, a large convenience sample provides more convincing data than does a trend in a random sample drawn from that convenience sample.

Readers of my article will find that I did not say that "empiricists have the weaker position," but said, "As a tactical issue, empiricists have the weaker position" (emphasis added). I still think so. It would have been a much better tactic to have acknowledged that Heineman-Pieper, Witkin, and others raised some important issues, then address these issues less stridently. Readers will also note that I did not characterize Thyer's entire 1989 article as patronizing. Rather, I was referring to the list of principles presented in the article. I still believe that in listing these elementary points Thyer was "talking down" to those readers with professional education.

I won't comment on Thyer's "Apparently it was not simple enough for Atherton to understand," "I am not sure what periodicals Professor Atherton attempts to publish in," and other such remarks beyond saying that I gather that they were attempts at humor and not intended to be mean spirited.

I found Professor Thyer's frequent accusation that I "misrepresented" his and others' works disconcerting. I prefer to think of my statements as either conclusions or observations. As a graduate of a respectable Presbyterian seminary, I regret that he did not find the theological metaphor helpful in understanding the factional conflict. It works for me.

I wish that both sides of this debate would stop dragging logical positivism into the argument. Logical positivism is not the foundation of science. Clearly, science, including social science, is much older than logical positivism, which was concerned with the meaning of language based upon the central notion that "the meaning of a proposition lies in its method of verification" (Passmore, 1968, p. 368). This notion is not self-evident but must either be believed or not believed, as with a statement in a creed. Logical positivism, derived from Wittgenstein's view of language, was a reaction to Kantian metaphysics (Hanfling [1981] and Passmore [1968] are good introductory discussions). Logical positivism "is dead, or as dead as a philosophical movement ever becomes" (Passmore, 1967, p. 52). All of the important figures in the movement are dead, and I am not aware of any living philosophers who identify themselves as logical positivists. Let the issue rest in peace.

In summary, I think that everything of importance in the controversy has been said at least twice. Proponents of each argument ought to sit down over some warm cocoa and those little English tea biscuits and begin talking quietly about something else. Goldstein is right: Let's take what we can from the argument and call it quits.

Atherton, C. R., & Klemmack, D. L. (198S). Research methods in social work. Lexington, MA: D. C. Heath.

Cronbach, L. 1. (1970). Essentials of psychological resting (3rd ed.). New York: Harper & Row.

Hanfling, O. (1981). Logical positivism. New York: Columbia University Press.

Passmore, J. (1967). Logical positivism. In P. Edwards (Ed.), The encyclopedia of philosophy: Volume 5. New York: Macmillan.

Passmore, J. (1968). A hundred years of philosophy. Harmondsworth, Middlesex, England: Penguin.

Copyright Family Service America May 1994
Provided by ProQuest Information and Learning Company. All rights Reserved