Stereotypes are pervasive. Humans tend to see the world in terms of categories. We see a chair, a table, a couch, and we group these items together. The category "furniture" makes our life a bit easier, a bit more effortless when having conversations with others or when confronted with unfamiliar objects (see right; below left) that might otherwise sap valuable cognitive resources.

Furniture often stands on multiple legs. Furniture is something on or at which we sit or lie. Furniture is a hassle to move. Spare furniture stored in one's basement seems to reproduce faster than rabbits. These stereotypes about furniture are uncontroversial, unless you happen to be an avant garde furniture artist, I suppose. One might reasonably argue that they are all, more or less, true.

We have a similar tendency to categorize other humans, but the resulting stereotypes about people are trickier. Whereas furniture itself doesn't mind being stereotyped, people often do. And it is a well-documented research finding that the pervasiveness of social stereotypes in our society does more than offend people, it also has observable, often negative impacts on their performance, domain identification, and long term perseverance on various pursuits.

But as important as these negative effects of stereotypes may be, they are a topic for another time. Today, I'd like to address the idea raised in a series of entries by one of my fellow PT bloggers, namely the provocative claim that all stereotypes are true. Because another important difference is that people are more complicated to evalaute than furniture, a complexity rendering it far more difficult to arrive at generalizable "truths" concerning the former than the latter.

Beginning with the first entry in his series, Kanazawa makes the argument that stereotypes have been given a bum rap. That "stereotype" is nothing more than a synonym for the type of generalization we scientists strive for on a regular basis. That only the frivolous forces of political correctness lead people to resist the application of stereotypes as unadvisable or inaccurate:

I have a number of reactions to this argument. One is to call into question the very analogy at its heart. I'm not convinced that generalizations are the foundation of scientific theory. Often researchers are content to responsibly draw conclusions based on the finite population or stimulus set under observation, leaving to others efforts to generalize. But even when we do seek to make generalizations as scientists, most of us do so in a cautious and data-driven manner that doesn't bear any similarity to the application of societal stereotypes like these people are lazy or those people are deceitful. Contrary to what Kanazawa would have us believe, there are important differences between what we as scientists do and what, say, politically incorrect comedians do, even for those scientists among us willing to devote time to blogging and other enterntainment-related endeavors.

But the stronger objection I have with this argument is that its depiction of stereotypes is too blasé in its absoluteness. The title of these posts is "All Stereotypes are True Except..." and the author goes on to identify 4 or 5 specific stereotypes that he believes are refuted by empirical data (subsequent entries that make the absolute claim regarding the veracity of all other stereotypes all the more dubious—those are the only exceptions to the rule?!). But when you title a series of posts "All Stereotypes are True..." there are logical conclusions that inevitably follow such a claim. For example, it means that Jews are indeed greedy or sneaky. Poles are demonstrably clumsy or unintelligent. African-Americans are immoral or subhuman. White Americans are unathletic, not to mention inescapably racist. For that matter, there exist unflattering stereotypes about evolutionary psychologists—are all of these true as well?

Now to be fair, the blogger's job includes coming up with catchy titles, so one can be forgiven an exaggeration intended to draw in the curious reader. But the sentiment that all stereotypes are true is articulated elsewhere, such as in the above quotation: "If they are not true, they wouldn't be stereotypes." Well, the examples I listed above are verifiable stereotypes in our society, whether via checklist studies of the sort Katz and Braly pioneered decades ago, or recent experiments in social perception and judgment. Where is the empirical evidence that they're all true? I've yet to see it.

Sure, some stereotypes are based in a kernel of truth, or exaggerate a difference supported by empirical data. But many of the stereotypes still pervasive in today's society seem to be vestiges of the past, artifacts of outmoded belief systems once but no longer accepted as social realities. Is Kanazawa really arguing that they, too, are all true?

He should know better. One need look no further than the history of science to see that the popularity or pervasiveness of ideas need not indicate their veracity. Flat Earth Society, anyone? Outside of the realm of science, how about popular conspiracy theories? Superstitions? The pre-war conclusions of international intelligence agencies? In fact, evolutionary theorists have written about the processes through which beliefs, or memes, gain currency in a culture. Truth is not the critical determinant. More important would be whether the idea benefits certain segments of society. And far more important than that is how easily replicated and passed along such ideas are. So take a society that for centuries has legislated inqeuality among its subgroups: In such a context, stereotypes about the traditionally disadvantaged group, true or not, may find fertile ground to spread like wildfire given the cultural landscape already in place.

In the end, I'd like to think my colleague does not really believe the title of his own entries. After all, in the 4-sentence passage I've quoted above, the third sentence makes a much more cautious claim that stands in stark contrast to the absolutism that follows: "Many stereotypes are empirical generalizations with a statistical basis and thus on average tend to be true." That many stereotypes have a statistical basis that is true on average would be a provocative empirical question to debate. So, too, would be the claim that "the stereotypes that have been shown to be false so far have to do with people's physical appearance."

But there is no debate regarding the major conclusion offered, that all stereotypes are true. Such an absolute claim is too easily debunked, even granting the 4 or 5 exceptions described in later posts. And even the doubly qualified conclusion that "most stereotypes are on the whole true" is difficult to assess given the absence of an explanation concerning how wide a net is being cast—are we talking about just those stereotypes regarding gender and mate preference or the entire universe of social stereotypes? Again, I'd like to think that maybe my colleague is focusing on a small subset of stereotypes. But, unfortunately, in his most recent post on politically incorrect humor, he makes it clear that he is not, as he writes: "virtually all stereotypes (ethnic or otherwise) are empirically true."

In the end, the author wishes to argue that stereotypes are neither good nor bad but rather just "are," and that any resistance to stereotype application is simply reactive, sentiment-based posturing that takes place outside the purview of a single-minded scientific fundamentalist. But I'd argue that more pressing fundamentals of science include precision in advancing a thesis and thoroughness in assessing its veracity—especially a thesis with the type of political and social consequences with which Kanazawa, by his own admission, doesn't concern himself. Just as the pervasiveness of a belief in society does not prove its veracity, neither does simply including a sweeping conclusion in a blog make it a true statement.