How We Know What We Think We Know

One of the pseudo-buzzwords I've helped to coin is "indexical knowledge." Well, check out the always-nimble Louis Menand's New Yorker article on voters' decision-making. He doesn't bust out the phrase (otherwise, I'd have to sue) -- but what he's talking about, from soup to nuts, is indexical reasoning.

Most of the time, the things we want to know -- whether we can trust someone, what the weather will be like tomorrow, whether the economy is doing well or not so well -- are things of which there is no simple measurement or even certain knowledge, even with a great deal of research and effort.

So we rely on an index -- a sign -- that gives us a quick and dirty way to get a maximum amount of information with a minimum amount of effort. We're expert at sizing people up quickly by their age, sex, race, dress, demeanor, and behavior; the best way to tell what the weather will be like in Detroit tomorrow is usually what it's like in Chicago today; and we have a variety of sophisticated and not-so-sophisticated economic "indicators" (GDP, per capita income, employment and unemployment, etc.) which provide simple, bullet-point headlines on the state of the union. None of these are 100% accurate, but that's not their virtue -- they act as a kind of shortcut (and shorthand) to find out what you want to know. This is especially helpful when more detailed or accurate knowledge is either unavailable or unproductive.

Why "index"? What characterizes an indexical sign is that it bears some sort of physical or empirical connection to its referent. "Where there's smoke, there's fire." (This cliché is actually about the value of indices -- their soundness, their generality, their ability to give protection from danger.) Linguistic signs don't have this kind of relationship -- they're arbitrary. Neither do icons -- pictures or copies of the referent. (Shot-out to C.S. Peirce, who, by inventing this terminology, invented semiotics.)

How does indexical reasoning work? Menand gives the example of picking out a stereo. We could spend hours poring over expert reviews, reading technical information, comparing prices, and so on. Or we could walk into Circuit City and pick the one that looks the coolest, as long as we recognize the brand name. I actually enjoy doing the first, but the odds are pretty good that the other guy's going to get nearly as good a deal, without spending nearly as much time doing it. Also, he doesn't believe that the time and investment in acquiring an object adds to its symbolic and psychological value. But let's not get started.

So how does this relate to politics? In two ways. Voters identify candidates they like and don't like through branding (political parties), aesthetics (Bush looks good in a flight suit, Dukakis looked dumb in a helmet), endorsements, and generally, gut judgments over ideology or information. In fact, the more information voters have, the more they need good indices in order to make sense of all of it. Sometimes these are personal touchstones (elites whom people trust or with whom they identify) and sometimes they're empirical but indirect: Florida voters have done pretty well over the past four years, so they're leaning Bush, but more Ohio voters have had it rough, so they're leaning Kerry. Who knows whether voting for Bush or Kerry will make things better or worse; some indices are better than others.

What's more, voters (elites and non-elites alike) need these indices in order to legitimate their choices on the one hand and to signal them to others on the other. We need quick figures, clichés, and nonsense formulae ("I think Kerry seems presidential") just to be able to talk with our friends, families, neighbors, and co-workers.

In short, we need to make the metropolis (the grossstadt) more like the polis (the kleinstadt). We need to make the things of the electronic age sensible for the stone-age minds we carry around. But this isn't really what I want to say -- it just gives you the gist.