Paul Rozin on what psychologists should study

Paul Rozin, one of the founding fathers of cognition-and-culture studies, is a psychologist with a rich set of interests. Even though he’s often known for his work on food, and disgust in particular (cockroach in your drink anyone?), the list of his current projects alone would make many a psychological career look narrowly focused. However, this post will not dwell on the value of having such diverse interests, but on the value for psychology of adopting a richer set of methodologies.

In an insightful series of articles (see below), Rozin highlights some of the shortcomings of modern psychology (while his focus is primarily on social psychology, his remarks apply equally well to most of cognitive psychology). One of these shortcomings is the failure to sufficiently take into account—and study—cultural variability. Even the bulk of cross-cultural psychology only compares undergrads across countries (usually a ‘Western’ sample and an ‘Eastern’ sample). But Rozin draws our attention to the even less forgivable paucity of data regarding presumably less stark cultural variations along ethnic, religious, political or social lines. Understandably, for most Western researchers, a trip to Shanghai or Kyoto to carry out an experiment will be more attractive than one to, say, inner-city Detroit (I plead guilty here). But there also seems to be a publication bias: cross-cultural psychology journals are likely to publish more easily a comparative study of Chinese and American undergraduates rather than one comparing, say, blue and white collar workers in Philadelphia (coda: a publication bias nearly automatically translates into a grant bias which further compounds the problem). But I will not belabor this point, as the lack of cultural variability in the samples of psychologists has already been discussed on this blog.

Rozin’s second important contention is that psychology has become too much hypothesis-driven and that it pays too little attention to the simple study and reporting of phenomena. In order to drive this point home, he engages in a comparison of psychology with her big sister, biology. Rozin points out that when biology was developing, in the 18th and 19th centuries, the study of phenomena—natural history—was playing the major role, while hypothesis driven research only kicked in much later. Of course, someone could point out that psychology does not have to replicate what may have been the errors of a young science and that modern psychology is perfectly justified in adopting the latest research methods from biology. But this criticism would miss the mark, as even modern biology is much less hypothesis-driven than modern psychology. Rozin drives this point home through the comparison of a sample of psychology and biology journal articles. He notes that many of the biology articles are simply driven by the fact that “little is known about X”, so that “informed curiosity seems like quite an acceptable justification for a study.” There is very little room in top psychology journals for the ‘mere’ description of phenomena. By comparison, and drawing only on my recent memory, there have been reports in top biology journals of observations such as coconut carrying octopi and spear throwing chimpanzees.

A related point is that psychologists may now be too bent on minute investigations of mechanisms, again as opposed to making sure that these mechanisms fit within the larger picture of the known phenomena. Sometimes a phenomenon that is only reported in a few (or even a single) studies will give rise to many publications trying to elucidate the precise mechanisms that give rise to it. This is often done without pausing for a second and wondering of the original phenomenon is robust—if it occurs across a reasonable variety of situations—or if it is not part of a larger phenomenon—in which case studying the details of a particular situation is unlikely to yield a deep understanding of the mechanisms.

Why has psychology turned that way? Why does it, maybe in its efforts to differentiate itself from other forms of social science, seem to have gone too far and thrown out the baby with the bathwater? Rozin suggests several explanations, such as the tendency of people at the bottom of a scale—and psychologists are likely to feel that they are at the bottom of the natural sciences—to ‘overdo’ it. However, maybe the most likely explanation is that other people are already doing the work of reporting on phenomena. They are anthropologists, sociologists, demographers, political scientists, historians, economists… and psychologists seem to be wont to stay away from them or their findings—findings that, anathema, did not depend on the experimental method!

Obviously, there are exceptions to this trend. Rozin mentions the work of Richard Nisbett and Dov Cohen on the Culture of Honor. Likewise, the research of Doug Medin and his collaborators on naïve biology relies heavily on careful experimentation, but also on ethnographies, interviews, historical data, etc. But this type of work is altogether too rare.

Much as the usual critique of psychology as being ‘culture-blind’, the present critique might come as old news to some readers of this blog. Still, I thought it might be worth repeating that, while psychology certainly has a lot to bring to other social sciences in terms of methodology (psychologists dream of a time when all anthropologists will know how to do an ANOVA), the reverse is at least as true. And such an avowal has more strength coming from a renowned psychologist.

Before finishing, I couldn’t help but add this wonderful quote of Solomon Asch that captures the present phenomena very well:

“[Psychologists] have, for example, striven to emulate the quantitative exactness of natural sciences without asking whether their own subject matter is always ripe for such treatment, failing to realize that one does not advance time by moving the hands of the clock.” (Social psychology, 1952, pp. xiv–xv, cited in Rozin, 2001)


Cockroach study:

Rozin, P., Millman, L., & Nemeroff , C. (1986). Operation of the laws of sympathetic magic in disgust and other domain. Journal of Personality and Social Psychology, 50(4), 703-712.

Papers on the topic discussed here:

Rozin, P. (2001). Social psychology and science: Some lessons from Solomon Asch. Personality and Social Psychology Review, 5(1), 2. [contains an hilarious–for psychologists–parody of their own field]

Rozin, P. (2007). Exploring the landscape of modern academic psychology: finding and filling the holes. The American psychologist, 62(8), 751.

Rozin, P. (2009). What Kind of Empirical Research Should We Publish, Fund, and Reward?: A Different Perspective. Perspectives on Psychological Science, 4(4), 435.

Nisbett and Cohen on honor:

Nisbett, R. E., & Cohen, D. (1996). Culture of honor: The psychology of violence in the South: Westview Pr.

Medin and Atran on naïve biology:

Medin, D. L., & Atran, S. (1999). Folkbiology: The MIT Press.


  • comment-avatar
    scott howard 2 August 2010 (02:32)

    A major limitation to reporting cultural psychological phenomena is that mere observations will be subjected to calls of racism>> e.g.: It is observed Black people do this and White people do that and Arabs to otherwise etc, etc. Observations appear to be opinions and of course everyone has an opinion. Part 2 is that without statistical analysis then people respond with antecdotes. e.g.: It is observed that the lack of fathers is determintal to Black children>> but Obama seemed to do ok without his daddy. So you have to say: children in this study of 8 year olds, in Texas, do worse (as defined by grade point avg) in fatherless homes 92% of the time with an error of plus or minus 6.3 points. (this statistic was made up on the spot for demonstration) Even the observations will need to have caveats and study limitations. It does not matter if the observation is correct. We still live in a racist society with cultural biases; and those other parties will have a say about the psychological study, regardless of the outcome. Part 3 That all said I have lots of really interesting observations about psychology and culture. As a treating psychologist I can see how the exact same mental disorder manifests itself differently depending upon if the parents are wealthy and educated vs the parents who rent-hop motel rooms. Another example: I work in two different clinics in different towns. In one clinic it is the bullies who are referred for treatment and in the other clinic it is more likely that the bullied kids get referred for treatment. So people’s perception or definition of mental illness depends upon their location? I got thousands of observations… I should write a book.

  • comment-avatar
    Konrad Talmont-Kaminski 2 August 2010 (13:56)

    I have been using Rozin’s work on disgust as a paradigmatic example of first rate psychological research for a number of years now. I particularly appreciate the way in which he carefully examines a number of dimension of the phenomenon and tries to relate them to each other. I think I agree with his identification of a problem with hypothesis-making in psychology but I would have focussed on a different aspect of the cure. As he says, what seems to happen a lot in psychology is that a new methodology is invented, data are gathered and a mental mechanism is proposed to account for that particular set of data. However, while Rozin argues for just gathering data, it seems to me that the problem with hypotheses in psychology isn’t just that there’s too many of them but, also, that they are atomised. The end result is a plethora of mental mechanisms being proposed to account for what are often very closely related methodological results. Clearly efforts to synthesise are called for. Unfortunately, in so far as such efforts have been made, it seems to me that they all too often partake of the problem by attempting to synthesise via label rather than providing any real insight. I am thinking in particular of dual systems approaches. My impression, and I would be glad to hear from others who would differ, is that the idea that the mind contains two systems has very often been used as an all-purpose explanation for any and all distinctions revealed by experimentation. I am not talking about the work of Evans, who is a whole lot more careful, but of the broad tide of articles that have referenced dual systems without any real benefit in understanding of either the data or the hypotheses. Nor, to be added, to the necessary detriment of those articles either. A similar point, it would seem to me, could be made of the use of heuristics in the heuristics and biases tradition. In terms of explanation, I would argue (over a beer, for now) that the problem is not how psychologists self perceive but the lack of an over-arching theory. In the comparison with biology, it cannot be forgotten that work in biology has been able to rely upon evolutionary theory to make sense of the whole picture (as Dobzhansky famously pointed out). Of course, evolutionary theory is now being brought into psychology (not that I need to remind this audience of the fact!) And all to the good. Yet even here I see in some of the ways that evolutionary psychology has been taken up the same kind of synthesis-by-label that I mentioned earlier. Perhaps I think my own discipline too influential but I would blame some of this on philosophers for failing to engage closely enough with the empirical work in psychology to help hypothesis-making take on a more useful form than what looks to me very much like reinventing the wheel every time a new road is discovered. In other words, I think there’s a call for more philosophy of psychology (rather than philosophy of mind). Trust a philosopher to reach this conclusion on the basis of a scientist saying that there’s too much hypothesis-making!

  • comment-avatar
    Bill Benzon 5 August 2010 (01:01)

    Rozin’s second important contention is that psychology has become too much hypothesis-driven and that it pays too little attention to the simple study and reporting of phenomena. In order to drive this point home, he engages in a comparison of psychology with her big sister, biology. It’s quite clear to me that this is the case in literary studies, a point I made in my recent post on cultural evolution at the National Humanities Center (online [url=]here[/url]) , downloadable [url=]here[/url]). What do we need to describe? Literary texts. For example, since the early 20th century it’s been standard to distinguish between story, the incidents in a narrative in temporal order, and plot, the order in which the incidents are entered into a narrative. In many narratives the two orders are the same, but not always (see comments in this essay on evolution of narrative [url=]here[/url]). It seems to me that it would make sense that these two orderings be worked out for every narrative of interest, and that that descriptive work be made available through some standard web-based repository so that every researcher can reference it. The interesting cases, of course, will be those where plot and story diverge. I would expect that we will learn a great deal when we have, say, 100, 150, 300 cases worked out in detail. We’ll see things we’ve never seen before.

  • comment-avatar
    Bill Benzon 5 August 2010 (01:05)

    Alas, the comment engine stripped out my hyperlinks. Online post on cultural evolution: Downloadable version: Paper on narrative evolution:

  • comment-avatar
    Nicolas Baumard 7 August 2010 (12:42)

    Rozin’s second point is not specific to psychology. In all disciplines (at least in human sciences), the trends has been to focus only on one method (such as experiment in psychology) and for a discipline to define itself by its method rather than by its subject. Psychology is the study of human phenomenon by experiments, economics by modelling, philosophy by abstract analysis, sociology by surveys, etc. The best example of such a trends, as many know here, is anthropology. Anthropology has moved from the study of mankind to the study of mankind trough ethnography. When you do not do ethnography, you cannot be an anthropologist! I wonder what are the sociological mechanisms at stake here: is it peer-review? is it the competition among disciplines? is it an effect of social selection that makes modelling people go to economics and economics being more and more about modelling? is there a kind of logic of the field à la Bourdieu? Whatever these mechanisms are, their consequences are of importance of scientific research. For many psychology journals a psychological phenomenon exists and is worth studying only if you can figure out a experiment (even totally trivial) to demonstrate its existence.In a way, we are still [url=]behaviorists[/url]!