Theories in Psychology – A Gigerenzian View

As I alluded in a prior post, I’m spending August convalescing, which is my excuse for why this post draws largely on quoting someone else’s work. With my right arm in a sling, typing is like putting sunscreen on your own back; it can be done, but it requires uncomfortable contortions.

So, with that, I’m writing today – or, really, mostly cutting and pasting – from a paper that I really like but am guessing most readers haven’t read, “Surrogates for Theories” by Gerd Gigerenzer, published fifteen years ago, in 1998. I kind of love this little paper, and commend it to everyone, especially students. I begin where he ended. Gigerenzer writes:

Several years ago, I spent a day and a night in a library reading through issues of the Journal of Experimental Psychology from the 1920s and 1930s. This was professionally a most depressing experience. Not because these articles were methodologically mediocre. On the contrary, many of them make today’s research pale in comparison to their diversity of methods and statistics, their detailed reporting of single-case data rather than mere averages, and their careful selection of trained subjects. And many topics—such as the influence of the gender of the experimenter on the performance of the participants—were of interest then as now. What depressed me was that almost all of this work is forgotten; it does not seem to have left a trace in the collective memory of our profession. It struck me that most of it involved collecting data without substantive theory. Data without theory are like a baby without a parent: their life expectancy is low.

The short paper takes a broad view of the field of psychology, focusing on what counts as theory in the field, with an emphasis on four strategies that Gigerenzer thinks stand as “surrogates” for real theories; surrogates, he says “are vague, imprecise, and/or practically unfalsifiable, that they often boil down to common sense.” I’ll discuss just two. In essence, Gigerenzer is reacting to the fact that, in psychology, “almost anything passes as a theory.” I leave it to the reader to judge if things have changed in the ensuing fifteen years.

The first surrogate Gigerenzer discusses are “one word explanations.” Of these, he says:

Such a word is a noun, broad in its meaning and chosen to relate to the phenomenon. At the same time, it specifies no underlying mechanism or theoretical structure. The one-word explanation is a label with the virtue of a Rorschach inkblot: a researcher can read into it whatever he or she wishes to see.

Working in the literature in judgment and decision making, Gigernezer has in mind terms like “availability” and “similarity,” words used in the heuristics and biases tradition. A central difficulty is that because they are single words instead of well-specified models, nearly any result can be viewed as consistent with one (or more) of these “explanations.” In this sense, they are not scientific explanations at all. My view is that it remains good practice to be vigilant about putative explanations that are single words, such as “culture,” “learning,” or “plasticity.”

A second issue that Gigerenzer raises is “muddy dichotomies.” He begins by emphasizing that “There is nothing wrong with making distinctions in terms of dichotomies per se,” but rather the issue is “situations in which theoretical thinking gets stuck in binary oppositions beyond which it never seems to move.” Again begging the reader’s indulgence, here is a lengthy quotation from the paper:

Let us consider a case in which false dichotomies have hindered precise theorizing. Some arguments against evolutionary psychology are based on the presumed dichotomy between biology and culture, or genes and environment (Tooby & Cosmides, 1992). One such argument goes: Since cognition is bound to culture, evolution must be irrelevant. But biology and culture are not opposites. For instance, our ability to cooperate with conspecifics to whom we are genetically unrelated— which distinguishes us humans from most other species—is based on mechanisms of both biological and cultural origin. Simply to ask about the relative importance of each in terms of explained variance, such as that 80 percent of intelligence is genetically inherited, is, however, not always an interesting question. The real theoretical question concerns the mechanism that combines what is termed the “biological” and the “cultural.” For biologists, the nature/nurture or biological/cultural dichotomy is a non-starter: genes are influenced by their environment, which can include other genes.

Part of Gigerenzer’s agenda is explaining why surrogates for theories are a particular problem in psychology. He discusses null hypothesis testing as one source of the problem; my perception is that psychology is moving – not as fast as many would like – away from the sorts of practices Gigerenzer is pointing to. I’m not as sure about the second culprit Gigerenzer identifies, the isolation of sub-disciplines within psychology.

Each subdiscipline has its own journals, reviewers, and grant programs, and one can have a career in one of them without ever reading the journals of neighboring subdisciplines. In addition, job searches are often organized according to these categories. This territorial organization of psychology discourages researchers from engaging with psychological knowledge and colleagues outside of their territory… [and] blocks the flow of metaphors and the development of new theories. Distrust and disinterest in anything outside one’s subdiscipline supports surrogates for theory.

Gigerenzer suggests that surrogates for theory “flourish like weeds.” Of course, that was fifteen years ago. Surely things have improved with the passage of time. Right?


Gigerenzer, G. (1998). Surrogates for theories. Theory and Psychology8, 195-204.

08. August 2013 by kurzbanepblog
Categories: Blog | 7 comments

Comments (7)

  1. I would also strongly recommend another Gigerenzer paper regarding theory. This is one of my favorite article.

    Gigerenzer, G. (1991). From tools to theories: A heuristic of discovery in cognitive psychology. Psychological review, 98(2), 254.

  2. Excellent points. I think we don’t get enough input from practical psychology either. I have sat in conferences where evolutionary folk have asked whether there are any data collected on (say) parent child relationships. This is pretty much the meat and drink of most psychotherapy–much of which (like our field) is derived from Bowlby. Very few people seem aware of how much good we could do one another

  3. A related argument of Gigerenzer’s is that psychologists often turn their tools into theories. For instance, with the rise of ANOVA, trait attribution was assumed to be done with an ANOVA-like process.
    Many people claim that viewing the mind as a computer is merely a metaphor (and perhaps one we should have outgrown). I tend to think that in a deep sense the mind really is computer. Do you know what Gigerenzer thinks of this particular “metaphor”?

    • Andy, my view is that the mind literally computes, in the Turing sense, but that like all metaphors, some properties of the source domain map to the target domain, but many don’t. I’m not sure exactly what Gerd would say.

    • I think the confusion stems from conflation of two different definitions or meanings of the word “computer.” In a “deep sense” (Andy) or “the Turing sense” (Rob) a “computer” is something like “an algorithmic information-processing system.” I think we all agree that the brain/mind IS a computer in this sense of the word. There is no metaphor to potentially mislead us.

      However, this is not the definition that most people have in mind when you mention the word “computer”: They are thinking of a physical device, man-made from bits of silicon, metal, and plastic, that they use for various tasks (i.e., a “tool” in Gerd’s perspective). When we say something like “The brain/mind is LIKE a computer,” we not only invite metaphorical thinking with the word “like” (though technically that makes it a simile rather than a metaphor, right?–oh, never mind), but the meaning of the word “computer” has now changed to the more common, more concrete meaning. So the metaphor of the brain/mind as a “computer,” in the sense of the familiar physical, man-made device, comes up against the kinds of limitations to which Rob refers. (For example, the brain/mind does not do all of its information processing serially through a CPU, etc.)

      This distinction between definitions of “computer” at different levels of analysis might be helpful for thinking about whether the computer “metaphor” is or isn’t an example of Gerd’s deriving-theories-from-tools notion. If by “computer” you mean the laptop on your desk, and you “theorize” that the brain/mind works like that thing, you’d be turning your “tool” into a “theory” a la Gerd (and like social psychologists using ANOVA as a “theory” of attribution). But if by “computer” you intend the more abstract sense, I don’t think you’re doing the same thing, because a “computer” in this (deeper, Turing) sense is not a “tool” in the same way as your laptop’s hardware and software is. Now you’re making a more complex argument that certain principles of information-processing generally are applicable to both man-made laptops and brains/minds; this reasoning is further buttressed by arguments about why natural selection is likely to have discovered some of those principles. This seems like “real” theorizing to me, in contrast to simplistic tools->theories reasoning.

  4. Pingback: A week of links - Evolving Economics

  5. I just wanted to say thanks for the lead. That is a fantastic article, and unfortunately the arguments are just as applicable today. It reminded me of Steve Pinker’s response to Tim Wilson’s piece: His critique begins in the first comment after the actual article (right below “Reality Club Discussion”).

Skip to toolbar