Reflexivity versus rigor

Is a “reflective practitioner” possible?

In one of the most influential organization theory texts of the 1970s, Chris Argyris and Donald Schön (1974) noted that when managers are asked what they do, their answers do not seem to coincide with what researchers observe them to be doing. Thus, even if Kurt Lewin’s (1952) famous dictum “There is nothing more practical than a good theory” (p. 169) is still correct, practitioners may be using two types of theory: “espoused theories” (as presented in interviews with practitioners) and “theories-in-use” (those that can be deduced from observations of actual practices). ¹ Whereas espoused theories tend to be versions of, or at least allude to, rationalist models of decision making, theories-in-use are more or less descriptive of successful tinkering (in the vocabulary of Knorr Cetina, 1981), rarely verbalized by their practitioners:

Managers do reflect-in-action, but they seldom reflect on their reflection-in-action. Hence this crucially important dimension of their art tends to remain private and inaccessible to others. (Schön, 1983: 243)

Thus, Schön (1987) suggested that such theories-in-use should be made explicit, which can be achieved by educating reflexive practitioners.

But can managers be taught to reflect on their reflection-in-action? Only if those practitioners alternate between acting and reflecting; according to Niklas Luhmann (1998), it is impossible to do both simultaneously. The world as seen by actors is necessarily unlike the world seen by observers. Observers are able to see options and distinguish among them—but it is difficult for them to decide how to act. As many studies in our field have shown, the best way to paralyze action is to start a proper decision-making process. Actions stem from resolutions, not from decisions understood as acts of choice. Actors can see diverse options only in the moment of reflection, of observing, of not acting (Czarniawska, 2015).

Schön’s (1983) way out was the postulate that reflective practitioners must therefore develop “a double vision” (p. 164). The very awareness of an alternative would be paralyzing however: “Even Buridan’s ass, placed, as it were, between two equally tempting bales of hay, will survive, even if it notices that it cannot decide, for that is why it decides nevertheless!” (Luhmann, 1995: 360). The donkey is not an observer; if hungry, it will start eating one of the bales, without making a decision. Contradiction—in life and in science—grinds observation to a halt and demands action. Observations can occur only at a distance, establishing distinctions until they become paradoxical. Then it is time to drop the observer’s stance, to come closer and start acting.

Schön would have disagreed, but he did notice the tendency of systems to reproduce themselves, a point strongly emphasized by Luhmann. Thus, the need for the special training—or for consultants’ help. Indeed, a critical perspective on teaching management, as formulated, for example, by Ann L. Cunliffe (2009), offers exactly this possibility: leadership education as an opportunity to reflect over practice in a context separate from that practice.

Whereas management practitioners may have difficulty in reflecting on their practices while practicing, professions do exist in which action is, by design, separated from observation, where reflection on action and reflection on reflection-in-action, to quote Schön again, are practices. One obvious example is research, at least social science research. (I wouldn’t dare to express opinions on natural sciences—at least not in the present context.) Yet, it seems that this part of the scientific trade is often neglected or treated perfunctorily. Why?

Scientific rigor as a way of avoiding self-reflection

My claim is that although researchers suffer from the same difficulty as any other practitioner—the impossibility of acting and observing simultaneously—academic research is constructed in a way that demands and permits reflection. This possibility is not achieved by god-like efforts, but through the institution of scheduled reflection. There are times when, places where, and occasions in which a researcher is allowed—nay, is expected—to withdraw from action and reflect upon it as a matter of self-reflection or a reflection aided by others. Conferences, internal research presentations, peer reviews, graduate education—all these situations are meant to elicit reflection over the way research is being done and what research is being done. Collective reflection should lead to self-reflection and the other way around.

Does it work like that? Not necessarily, because reflection is costly—in terms of time and psychological discomfort—and there is a way to avoid its pains. There are templates, ready-made standards for what is good and what is bad, and it is enough to adhere to those. If there were a list of scientific virtues, it would probably begin with “rigor.” Rigor is the core characteristic of good science, as distinct from literature, art, and music; rigor is a characteristic missing from “bad science,” which is sloppy (“dopey” is the epithet most often used in Swedish), self-indulgent, and unstructured. But the etymology of the term “rigor” indicates that it simply means “stiff”—as opposed to flexible, adaptable, and lively (thus rigor mortis).

What, then, is seen as “rigorous scientific conduct”—not in philosophy but in the practice of science? Mostly, it is understood as following to the letter a set of rules imposed by (some) scientists on other (usually younger) scientists and sometimes even on themselves. Deviations are seen as flaws in self-discipline, as poor professionalism, as incompetence. Yet, as Kuhn (1996 [1962]) made explicit many years ago, all scientific revolutions depend on deviations from and/or breaking these rules. Thus, “normal science” is supported by rigor, whereas “revolutionary science” relies upon the researcher breaking out of rigor.

Two consequences of these assumptions must be examined. One is that the products of “normal science,” although derivative of revolutionary science, are numerically more prevalent and—perhaps—more important. Robert Merton (1985 [1965]) suggested that, contrary to the popular belief that researchers stand “on the shoulders of giants,” scientific giants have always stood on a pyramid of dwarfs and by that action have been able to see further into the distance. He suggested that one could reconsider the relationship between giants and dwarfs in science as symbiotic rather than parasitical. Rigorous research alone would drive science to a dead end; revolutionary research alone would disperse and vanish. Revolutionary research needs rigorous research if it is to leap forward; normal science needs revolutionary research if it is not to stagnate.

The second consequence, and its importance, can be made visible only by introducing a counter-factual. What would have happened if the requirement of scientific rigor were abandoned?

One possibility, often evoked, is that science would lose its grounds for legitimacy. Scientists would begin experimenting with arts, journalism, or astrology, and the society would—rightly so—stop financing this specific activity.

Another possibility is that the requirement of rigor could be replaced by another requirement, which would fulfill the same legitimizing function, without producing the same type of undesirable side effects. This requirement is precisely a habitual and systematic self-reflection. Instead of rigorously following a set of received rules, researchers would need to examine their own doings continuously. Asked, “Why do you do this that way?” they would be forbidden to answer, “Because this is how it is being done in my discipline,” followed by parentheses filled with names and dates. They would have to argue for the approach chosen, for the techniques selected, for the format of the report produced.

In a sense, the necessity of a habitual and systematic self-reflection was one of the postulates of modern science, which refused to accept the appeal to authority—then in the personages of Fathers of the Church, now in the personages of Fathers (yes) of the discipline. But even some postulates of modern science would have to be re-examined. One of those is the forensic model of science (Douglas, 1986), which sees research as a search for evidence. An interesting contemporary variation is Evidence-Based Medicine, which has now proliferated into Evidence-Based Anything. It began as a pragmatic move away from the ideals of high science in medicine. (After all, one cannot wait for an adequate theory in order to start treating a disease; if something works, it should be used, even if we do not know why.) But for some reason, it is now being presented as a high(er) science elsewhere—including the social sciences.

And then, last but not least, there are always some analytical concepts that ceased to do their analytical work and became blunt tools. Such concepts are usually dead metaphors. As Nietzsche (1976 [1873]) said,

… truths are illusions about which one has forgotten that this is what they are; metaphors which are worn out and without sensuous power; coins which have lost their pictures and now matter only as metal, no longer as coins. (pp. 46–47)

In a similar vein, but in less dramatic terms, Umberto Eco (1979) noted that facts, once known, cease to be interesting. It is the live metaphors that aid our understanding of reality. Social scientists should therefore cultivate new metaphors, for they are as fragile as butterflies and may die soon. Researchers should keep the metaphors that work, at least for a time, but the day they begin to restrain thinking, they should be eliminated. And if they merely serve to clutter a text, they should be thrown away. Karl E. Weick’s (1996) famous appeal to “drop your tools” applies not only to fire workers. Perhaps each spring a “conceptual cleaning” should be undertaken (which does not exclude the possibility of attractive vintage additions).

What about some help in reflection, like that given to managers by consultants or by critically minded teachers? Peer reviews can be seen as the closest analogy here. In principle, peer reviews should be an effective tool of professional reflection. In practice, they turn too often into a defense of the bastion of rigor. As excellently formulated by Ingrid Jeacle and Chris Carter (2014), peer reviews

… must consistently push authors to think critically about how they have constructed their arguments and direct them to any weaknesses in theoretical interpretation and application. But in its pursuit of rigour and standards, the review process must also be careful not to curb creativity. The crafting of an academic paper is an inherently creative process and the ultimate published result should still bear testament to the author’s initial inspiration. It should not simply become a mere reflection of how one reviewer would interpret the data or recount the story. If consistently applied, such a process only leads to circular thinking and stagnation in the field. Rigour does not necessarily equate with critical thinking. Hence a perilous tension exists between the constrictive review that ultimately culminates in a bland contribution and the more reflexive review that allows creative risk taking. (p. 1238)

So, my suggestion would be to exchange rigor for a flexible, forever changing, dialogical reflection over the way research is done and what research is done. This suggestion assumes a circuit of collective reflection and self-reflection. One caveat, though: This appeal for self-reflection must not be understood as an encouragement for introspection. As Bruno Latour (1988) once said, reflectivity is best consumed at home, behind closed doors—be those of one’s office or at a conference. The wider public needs to see the results; it does not need to watch the painful process.

Following this course of action can be politically dangerous (in the limited sense of university politics) and psychologically uncomfortable. But learning is always psychologically uncomfortable. It is the results that are pleasant (most of the time).