Friday, November 25, 2005

Ideological Reasoning

What is an ideology? Douglass North, in Structure and Change in Economic History (1981, p.49), offers a thought-provoking definition:
Ideology is an economizing device by which individuals come to terms with their environment and are provided with a “world view” so that the decision-making process is simplified.
Some reactions. First, this definition implies that having an ideology (though not necessarily any ideology) is rational. A normal human being cannot possibly comprehend the bewildering array of things he perceives in the world without some kind of filter to make sense of it all. There is simply too much data, and the world is simply too complicated. Abstraction – multiple layers of it – is necessary to function in a complex world.

Second, an effective ideology will necessarily fail to explain some aspects of the world accurately. The act of abstraction typically involves focusing on identified salient features of a situation or situations, while deliberately ignoring others. This is what makes a model tractable; it is the reason economists adopt models with assumptions sometimes notably at odds with reality. If you try too hard to make your initial assumptions comprehensive, you will find yourself unable to “solve” the model and make predictions. Yet any model that is not comprehensive – that does not account for every variable or exception or special case – will have blind spots. (For more along these lines, see my article with Roger Koppl on “Rational-Choice Hermeneutics.”) This is not an argument against modeling, and it is not an argument against ideology.

Third, an ideology, like any model, should be subject to testing. But what sort of testing is appropriate? One difficulty is that unless your ideology is subject to modification at the margin, a sufficiently robust refutation means altering your whole world view. (Think of the former Christian who, once having been persuaded of the non-existence of god, finds himself unmoored and unable to deal with the world.) In Kuhnian terms, an ideology is more like a scientific paradigm than a scientific theory; sometimes a new paradigm is required, but people invested in the old paradigm will be resistant to a revolution in thinking.

Fourth, even if someone’s ideology is subject to marginal change, for most people the relevant form of testing is poor results in their personal lives. Yet some aspects of one’s ideology – politics especially – are unlikely to have strong personal effects. Your goofy political views probably won’t stop you from getting a job, unless your zealous proselytizing turns off your boss and coworkers. As a result, we shouldn’t be surprised to see little alteration of people’s ideologies on those fronts relating to matters of great public but little private consequence. (For more on this, see Bryan Caplan’s article on “Rational Irrationality.”) The persistence of unsophisticated ideological views on such matters should be no great surprise.


MT said...

If it's as you say, then what I'd add is that ideology is only one tool. Off the top of my head I probably would have said that chosing who and whose testimony to trust were the biggest factors in political decision making.

MT said...

Here's how my comment meshes with and potentially squashes yours: Testability isn't a one-thinker operation. Einstein didn't test his theory. Marx didn't test his (and I'd say it still hasn't been tested). Socially useful theories are tested and approved by communities. Competing theories implies competing communities. Which testers and which community do you trust? Unless you are an extreme political reductionist who believes history is the result of a few types of behaviors you can observe around the dinner table, nobody can believe they've done any tests themselves of their political theories. Are you suggesting we restrict ourselves to such reductionist ideologies? Must we rule out emergent properties of such large numbers of people over such a long term that one person can't witness them firsthand?

MT said...

By "who do you trust" I mean e.g. where do you get your news, what history textbooks do you read, which economists, do you trust the management or the labor spokesperson, the legislature or the judiciary, the neighbor on the left who is a teacher and the neighbor across the street who teaches at the same school but sees things differently, etc. Before you have any evidence for your ideology--or more likely in parallel with your embrace of an ideology--you must decide what testimony to trust. Only a very small fraction of the evidence is firsthand. According to some theories ("God's running things"), none of it will be firsthand.

Hugin said...

Glen, I read your blog regularly and first of all just have to say that this was a really good post. You are usually insightful, but this was a cut above the rest.

As for the comments by Murky Thoughts, I think the potential misunderstanding can be resolved once you understand that ideologies usually are claimed to exist on two levels, one at the level of the individual as a guide for decision-making, and the second at the level of collective abstraction regarding what people hold what world-views.

Of course, the second level is really only the first level, but the individual making inferences regarding the behavior of others rather than regarding the nature of the world.

If this distinction is properly understood, I take it that you Glen imply that ideology on an individual level is a necessary precondition for human life. However, given that this is the case, you can make normative statements concerning what an ideology ought to be, quickly moving to the second level of abstraction and discussing how people should make abstractions concerning the world. Here we know for a fact that a lot of people employ very stupid ideologies, but this in itself does not prove that ideology in itself is stupid or unnecessary, just that some are better than others.

Or what do you think?

Glen Whitman said...

Anders -- thanks, I appreciate the feedback. And I think your clarification is correct (at least to a first approximation). However, I wouldn't characterize the two levels of ideology as "individual theorizing about the [non-human] world" vs. "individual theorizing about other humans." I think you were closer to the mark the first way you put it: the first level is a single individual's ideology, and the second level is a generalization about many individuals' ideologies. Thus, the label "conservative" actually consists of many individuals whose personal ideologies happen to share certain characteristics. To the extent that people self-consciously adopt the label "conservative," internal or social pressure may influence them to adopt even more of the characteristics common to others under that umbrella.

And yes, I think it is possible to make a normative evaluation of ideologies, at least to some degree. But I would warn against criticizing an ideology solely because it abstracts from certain details and therefore gets some things wrong. Even a good ideology will get things wrong from time to time -- if it didn't, it wouldn't be doing its job of economizing on cognitive processing.

MT said...

Hey, I'm interested in a conversation too! Yes, I hide my identity under a silly psuedonym and I'm acerbic, but I come in peace. Anders spoke as if he were addressing my point, but if he truly did I would appreciate a leg up in connecting the dots.