Is the truth a popularity contest?

I said I was going to discuss some common errors in the science (or culture) wars. Here, I want to deal with one of the claims most irritating to sociologists-- the accusation that sociologists think that "truth is a popularity context" or that "anyone's opinion is just as good as anyone else's". This view is irritating just because it gets things so completely backwards.

If there's anything sociologists agree on, it's this: in any situation, some people count for more than others. The whole field is built on that assumption. Describing and explaining those differences has been the one of the principal tasks of the field since it first appeared.

Thus, some sociology of science is concerned with how and why some ideas and results gain more credence than others. And it isn't adequate to point to the "truth" of the better-accepted results, because that's begging the question. Rather, the problem is to understand how some ideas become established as "truth" in the first place. This process is interesting just because the rate and character and success of the truth-establishing process varies with respect to the nature of the institutional arrangements doing the work of establishing.

Here's an example. Today's New York Times has a story about a study which shows that maternal diet can affect the expression of genes in mice:

Scientists have long known that what pregnant mothers eat — whether they are mice, fruit flies or humans — can profoundly affect the susceptibility of their offspring to disease. But until now they have not understood why, said Dr. Randy Jirtle, a professor of radiation oncology at Duke and senior investigator of the study, which was reported in the Aug. 1 issue of Molecular and Cellular Biology.

The research is a milestone in the relatively new science of epigenetics, the study of how environmental factors like diet, stress and maternal nutrition can change gene function without altering the DNA sequence in any way.

We do have a relatively new line of work in epigenetics; here's a recent collection of papers in the area. But the problem has been around for over a century, usually deeply conflated with the question of how environmental influence can change gene structure (as opposed to expression). It's an important question, with high-stakes practical implications. So how come it's only getting started as a field very recently? Needless to say, the answer is a long, complex story, and we have only the bare outlines of it right now. A new history of the field (especially Chapter 17) provides good background. I also have a working paper which discusses some of the institutional changes involved.

More generally, it's clear that the popularity (or even the legitimacy) of research questions, results, and approaches depends heavily on the ways in which a line of research is connected to the institutions around it. This includes its relations with other (possibly complementary, possibly competitive) specialties.

So when somebody tries to impeach the work of a sociologist by claiming that different opinions count differently, they are missing the point and agreeing with the sociologist. Sociologists try to find out how and why some opinions count for more, and some for less. Very often, they discover that the claims of one side aren't as strong as some claimants would like to believe, or that other claims have some strength too.

Obviously, the people whose claims are challenged this way are going to be disappointed and irritated. There are several ways they might respond to the challenge. One is to discover and point out misunderstandings and technical inadequacies in the sociology. Another (complementary) response is use the sociological findings as a guide to improving the field. A third is to attack the sociologist personally, playing a kind of "shoot the messenger". A fourth is to shoot the messenger's people, by attacking sociology in general, or perhaps lumping sociology with all sorts of other despised phenomena (e.g., creationism, post-modernism, communism, atheism, satanism). The first two approaches lead to better research, and we're all better off. The second two don't.

Comments

[William R. Buckley]

Eli:

I do believe that this is one of the cornerstones of Theodore Nelson's
theory of the operation of the internet: That the value of the information
therein contained will be measured as a function of the number of links
to that information. Hence, belief systems do tend to affect the propagation
of information, and the relative value of that information.

Actually, Ted was talking about how access is focused within hypertext but,
the notion is more generally applicable.

[Rob]

Well, perhaps the number of links is a reasonable proxy to measure the impact of the presentation, crudely and bluntly maybe; a weighted google measure seems somewhat better. But the value of the information is likely different than indicated thereby.

I found this post to genuinely change my understanding of sociology; who knows how much value that will have, in future (perhaps none if I'm run over by a bus) but I'm quite unlikely to link to this.

Admittedly I came to this blog because I read a comment of the author's on another blog I was reading (Evolving Thoughts); but that's still not a link!