What 'Representation' Represents

The Economist magazine's Democracy in America blog cites The Wall Street Journal's Opinion Journal ("An Overrepresentation of Factual Presentations"), which links to the on-line magazine Grist, which features an interview with Al Gore.

(Whew! It's hyperlinks all the way down.)

In the Grist piece, David Roberts asks Gore how he balances fear with hope in talk about global warming. Gore responds that, given denial about the possibility of global warming,

I believe it is appropriate to have an over-representation of factual presentations on how dangerous it is, as a predicate for opening up the audience to listen to what the solutions are, and how hopeful it is that we are going to solve this crisis.

Commentators at Grist, Opinion Journal, and Democracy in America have all seized on the phrase "over-representation of factual presentations," and are debating what it could possibly mean.

James Taranto at Opinion Journal seems to understand the phrase as meaning something like "exaggerate those facts that support one's side of an argument." He compares this to discussion of Weapons of Mass Destruction during the run-up to the war in Iraq.

"An over-representation of factual presentations of how dangerous it is." Isn't that what people accused President Bush of offering vis-à-vis the erstwhile Iraqi regime? Didn't this lead a certain former vice president to thunder, "He betrayed this country! He played on our fears!"?

An anonymous contributor (all contributors to The Economist are anonymous) at Democracy in America sees this as even more ominous.

Whoops. "An over-representation of factual presentations". You can see what he means, and why that may well be the right strategy. But it doesn't sound like the right thing to be caught saying. Better at such times to praise the intelligence of the public and their capacity to recognise the wisdom of your argument, than admit that you plan to bombard them with "over-representations" of your case until you scare the pants off them.

For the Democracy in America writer, "over-representation" seems to be not only frightening ("scare the pants off them") but dishonest (not "the right thing to be caught saying").

Clearly, it is possible to interpret "over-representation" to mean "exaggeration" or even "lies" (of a certain type). I don't think that is the most common definition of over-representation, though.

As a native speaker of English, I would define "over-represent" as something like "include a disproportionate number of a particular type." For instance, we might say that non-Hispanic white students are over-represented at my university, the University of Colorado. While the University says "13.8 percent (4,043) [students] are minorities," the US Census reports that 72.1% of Colorado residents are "White persons not Hispanic." Thus, at 86.2%, they are somewhat over-represented in the student body.

A cursory glance at Google search results for the term "over represent" suggests that (once we throw out references to Al Gore's own usage) this usage is common.

"the overrepresentation of Asian and Jewish students and the underrepresentation of the white, non-Jewish majority [is a problem at Harvard]"

"[Lists of physician groups] tended to over-represent "large" groups (>69 physicians)."

"With the exception of the Hispanic category, books with human characters that feature only one ethnic (non-white) group over-represent that ethnic group’s distribution in the US population."

Translating from the common, statistical sense of "over-represent" to a sense relevant to discourse, what should we assume that "over-representation of factual presentations" means?

If we assume that there is some ideal ratio of problem-talk to solution-talk (an assumption that may be hard to support, but go with me here), then talking more about problems than solutions would be an over-representation. For example, if it is ideal to spend the same amount of time talking about problems and solutions, then spending, say, twenty minutes talking about the threat of global warming and only five minutes talking about potential solutions would be an over-representation of the problem within the 25-minute discourse.

This, in fact, is the meaning that interviewer Roberts attributes to Gore, in the comments section of the Grist piece.

By "an over-representation of factual presentations on how dangerous it is" -- an admittedly inelegant phrase -- I think he meant that at this point, people talking to American audiences need to spend more time convincing them that global warming is dangerous, relative to the amount of time they spend talking about solving the problem.

If this is, indeed, what Gore meant, is he advocating lying? No - at least, not necessarily. He is advocating talking a great deal about the problem of global warming, relative to the amount of talk about solutions. This talk can, and presumably should, be entirely true.

Is Gore advocating exaggeration? Not necessarily. What he seems to advocate is a great deal of "problem-talk," which might be inflated relative to "solution-talk," but which need not be exaggerated at all relative to the actual problem.

At that point, though, we move beyond discourse to empirical facts about global warming - an area in which I have no particular expertise.