Theory & psychology

When discussing psychological research, what surprises every economist or physicist is that psychology has no theory. It has many local ones but no overarching theory, not even a provisional one. Yet there is something even more surprising: a lack of awareness of the value of integration. Whereas the unification of theories, such as evolutionary theory and genetics, is a widely shared goal in physics and biology, it is barely visible in psychology. Few psychologists even consider theory integration as an objective. A textbook in economics starts with first principles that lead to an overarching theory and discusses how reality fits into this picture. A textbook in psychology lists dozens of theories in chapters on reasoning, intelligence, problem solving, and judgment and decision making—topics that appear closely related, but are populated by different researchers, published in different journals, and presented as independent enterprises. To the poor student, the relation between the various concepts in these different theories is never made clear. Why is present-day psychology such a patchwork of small territories, resembling, to use a political metaphor, Italy or Germany before unification around 1870? Why are psychologists so content working within the confines of their own small territories?

Few psychologists would open a journal with “theoretical” in the title, and even fewer would submit papers. The sad truth is that these objections were not off the mark. The final compromise was Theory & Psychology, which both distanced psychology from the problematic term and connected the two.1 There appears to be something peculiar about psychology and its relation to theory. Here are some observations and proposals.

Teach Theory Construction ?

A curriculum on theory construction might start with good examples, first principles, and the ability to detect surrogates for theory (see below). First principles could be learned from a dip into the history of science, such as Kuhn’s list of features of good theories: accuracy, consistency, broad scope, simplicity, and fruitfulness. But it should not begin and end with the history and philosophy of science, which for the most part does not deal with psychology. Rather, comparing research practices in the natural sciences with those in psychology is a first step towards getting a graduate student to think.

From Tools To theories

One of the most instructive ways of teaching theory construction is to investigate the origins of existing theories. The neglect of theory construction goes hand in hand with the inductive story that theories emerge from data, yet in spite of being repeatedly asserted, it is too simplistic an explanation (Holton, 1988). Discovery entails more than data and theories; it also involves scientific tools and practice. In fact, one major source of psychological theories is researchers’ tools, such as statistics and computers. When psychologists grow accustomed to a new tool for data processing, the tool is repeatedly proposed as the way the mind works.. Just as in Neyman-Pearson theory, the novel data were hits and false alarms, unlike in the earlier work on sensory discrimination, from Fechner to Thurstone, which was based on measuring thresholds or psychological differences between stimuli. Here, we have a fascinating story about how a major new theory was discovered. After talking with some of today’s major proponents of signal detection theory, I realized that few are aware of the origin of their own theory, with most believing that it stemmed from new experimental data. Yet the tool inspired the theory, and the theory inspired new kinds of data to be generated, which in turn were used to test the new theory.

Circular Restatements as Explanations

• The most primitive means of avoiding theories is to simply restate the phenomenon in question in different words and pretend to have offered an explanation. This technique is also known as re-description or tautology, and has a long tradition. Recall Molière’s parody of the Aristotelian doctrine of substantial forms: Why does opium make us sleepy? Answer: because of its dormative properties. Such circular restatements are not limited to attributing behavior X to an essence or trait X In research on the “belief-bias effect,” participants were instructed to judge the logical validity of syllogisms, such as: “No addictive things are inexpensive. Some cigarettes are inexpensive. Therefore, some addictive things are not cigarettes” (a logically invalid but believable conclusion). The observation was that judgments depended on both the logical validity and the believability of the conclusion. The explanation offered was: “Dual-process accounts propose that although participants attempt to reason logically in accord with the instructions, the influence of prior beliefs is extremely difficult to suppress and effectively competes for control of the response made” . The phenomenon that both logical structure and prior belief influence judgments is “explained” by restating the phenomenon in other words. Dijksterhuis and van Knippenberg (1998) reported the interesting result that priming intelligent behavior by inducing a professor stereotype resulted in more intelligent behavior. This result was explained in this way: “In concrete terms, activation of the professor stereotype is expected to result in intelligent behavior because activation of the professor stereotype leads to activation of intelligence” (p. 872). Katzko (2006) has analyzed this and other surrogate explanations in detail. There is nothing wrong about not having an explanation for how priming (writing down everything that comes to mind about a “professor”) can lead to higher performance in a general knowledge task. Yet circular restatements pretend to have an explanation and thus distract from finding a model about how the specific priming task could have an effect on a general knowledge task. Restatements both create a theoretical void and cover it up.

One-word Explanations

The second technique for avoiding theories is equally abundant. The observation that some factor X influences people’s judgments more than factor Y is explained by saying that X is more “salient,” “available,” “relevant,” “representative,” or “vivid.” One-word explanations are so perfectly flexible that they can, after the fact, account for almost every observed behavior. This technique is also known as the use of labels instead of models. To illustrate, one-word explanations can account for both phenomenon A and its opposite, non-A (see Ayton & Fischer, 2004). Consider the gambler’s fallacy: after a series of n reds on the roulette table, the intuition is that the chance of another red decreases. This intuition was explained by people’s reliance on “representativeness” by saying that “the occurrence of black will result in a more representative sequence than the occurrence of an additional red”. Next consider the hot-hand fallacy, which is the opposite belief: after a basketball player scores a series of n hits, the intuition is that the chance for another hit increases. This intuition was also attributed to representativeness, because “even short random sequences are thought to be highly representative of their generating process”.No formal model of similarity (“representativeness”) can predict a phenomenon and its contrary, but a label can do this by changing its meaning. One-word explanations can be neither proved nor disproved, and hence do not enhance our understanding of how the mind works.

Final Thoughts

Many researchers believe that in order to have a successful career, they need to publish as much as possible, the proliferation of so-called “least-publishable units.” Yet excellent departments want to hire researchers who have made a theoretical contribution and usually discourage mere quantity by asking applicants to submit their six best papers only. This procedure generates incentives to write something better than one has written before, not just more of the same. Similarly, the editors of major journals might consider writing editorials that discourage toothbrush culture and surrogates, and explicitly inviting articles that make advances in theory integration. This will make it easier for young researchers to combine pursuing a career with developing psychology into a theoretical enterprise.