Thursday, June 09, 2005

Peer Review, Small Audiences, and The Incredible Shrinking Guilds

A friend recently asked me who I’ve been reading. I had to admit that most of the folks I’ve been reading for the last several years have been from outside my home discipline. (I don’t just mean the management lit, but the stuff I read to keep my ‘scholarly’ side alive.) When asked why, I could only say that the stuff from cognate disciplines is simply better – better focused, less caught up in the fetish of methodology, less imprisoned by footnotes, more willing to take risks.

From what I’m told by people in those cognate disciplines, though, the same problems hold true there, as well; I’m just reading the exceptions.

Without taking my personal taste as normative, I still have to wonder how a field with such wonderfully interesting subject matter became mired in such turgid irrelevance. When I try to think of contributions to the public discourse to emerge from my field in the last thirty years, I can think of exactly one, and that one is iffy.

What the hell? How can so many bright people, applying rigorous (and very, very extended) study to compelling issues, produce such unreadable crap?

I’m beginning to suspect ‘peer review.’ In order to be taken seriously, a scholarly article needs to be ‘peer reviewed,’ or given the stamp of approval by people already in the field. The idea is to prevent quackery, or faddism, and to keep a premium on academic rigor.

Well, okay, but has it worked?

Freakonomics and Moneyball, two of the more interesting popular books of the last few years, are both premised on the (empirically-tested) idea that groupthink can trump evidence. They both tell stories of people using empirical proof to show that articles of faith among ‘experts’ in a given field are either false or badly limited. In both books, the empirical evidence had to become overwhelming before the groupthink was broken.

What would a Freakonomics or Moneyball of academia look like?

Some fields lend themselves easily to reality checks. An engineer friend used to like to tell the story of the first-day lecture by a prof in her first semester of grad school: You Build Bridge. Bridge Fall Down. No Partial Credit! If the bridge falls down, peer review ain’t gonna save it. Other fields, like literature, are impervious almost by definition. If the very subject matter is, by definition, fiction, then empirical evidence can only be of glancing relevance.

(Before the inevitable flaming, let me just admit that I’m using broad strokes here. Certainly, it’s possible to discover new facts about old texts. Although I’m using Manichean language here, I’ll admit that there’s more of a continuum in reality. Polemical license.)

In the social sciences, we’re somewhere in between. There are some basic facts we can use to check interpretations – I recall explaining to one incredulous undergrad that Hamilton couldn’t have been influenced by Marx, because Hamilton died before Marx was born — but there’s also considerable space for interpreting the facts that exist. More basically, there is considerable disagreement over what constitutes an interesting problem.

In theory, I guess, peer review could help make up the difference between what we can verify objectively and what makes for a good interpretation. In practice, though, it seems to fall prey to a really dreary version of groupthink. A typical journal submission will be sent to three ‘peers’ to review. If one of those three doesn’t find the problem interesting, that’s that.

Interesting-ness was never supposed to be a criterion. Peer review is the worst possible mechanism for determining interesting-ness. In a sample size that small, interesting-ness is somewhere between groupthink and whim.

Accordingly, we get journal articles about ever-smaller and more specialized subjects (such as would appeal to, say, three reviewers), using ever more arcane methods (each carefully footnoted), while the world blissfully ignores everything we say.

I think this is part of why so many younger academics have taken to blogging. The gatekeepers are so caught up in the internal fetishes of their respective guilds that they’ve lost sight of the big picture. Blogging gets rid of the gatekeepers, and lets authors find their own peers. If the people who find my stuff interesting come from literature, or engineering, or chemistry, that’s cool.

As a biochemist correspondent noted, guilds were built to ensure that production took place in the interests of the producers. When consumers found other options, the guilds dissolved rapidly. I see the guild system of publishing (and all that goes with publishing) starting to break down. The growth of the University of Phoenix (and of business majors generally) is a signal from the public that we’ve stopped talking about what they care about. The academic job market, as brutal and perverse as it is, is a kind of reality check; the guilds are losing their ability to reproduce themselves. Maybe they’re producing what nobody wants?

Sorry for such a shaggy rant. I haven’t nailed this one down yet. Any thoughts?