Sunday, February 3, 2013

Is an Ideological Turing Test Possible?

I have been giving thought to the idea of a Turing test for political ideology, which I first heard about at Daylight Atheism and you can learn more about at The Library of Economics and Liberty. The original Turing test was designed to measure artificial intelligence using a computer’s ability to mimic human thought through a series of questions and answers. An ideological Turing test would do something similar for political ideology to see how well a progressive could explain conservative ideology, or vice versa.

Now that I’m a couple of years into this blogging thing, I see that one of the major obstacles to constructive dialog is the inability of participants to articulate their opponents’ arguments or positions. Progressives are very bad at understanding conservative arguments, and conservatives are very bad at understanding progressive arguments. Given all of this, I find a political ideology Turing test to be a fascinating idea.

The process would work like the original Turing test, with a series of questions that a respondent must respond to about the opposing ideology. For example, you might ask a progressive to explain why most conservatives hold a pro-life stance on abortion, or ask a conservative to explain why most progressives tend to be pro-choice. After a series of specific questions, it would be possible to score someone on a scale of how well they can honestly represent the opposing point of view.

Scoring of the ideological Turing test would involve comparing the answers to those of true ideologues. Using the abortion example, we would examine the progressive response describing the conservative pro-life position and compare it to an actual conservative describing her own pro-life position. If the two answers are indistinguishable, you could say that the progressive truly understands the conservative position. If the progressive scores high on a whole series of questions, you could say he understands the conservative ideology.

The challenge would be to craft questions that force the respondent to truly understand motivations and driving principles instead of simply regurgitating talking points and superficial descriptions. For example, how do you ask the abortion question of a progressive and get them off of the arguments for women’s reproductive rights? And how do you avoid every conservative answer sounding like a soliloquy on limited government?

Another challenge I foresee is that I’m not sure true progressive and conservative ideologues understand their own ideology. My guess is that a significant number of conservatives would fail the Turing test for conservatism, unable to articulate the underlying principles of their ideology. Progressives would have a similar problem, knowing that they are on the left but unable to explain the foundations of their thinking or the reasons why they believe what they believe.

I really love the idea of an ideological Turing test, but I’m not sure any of us are capable of producing the questions, the answers, or the scoring methodology.

4 comments:

  1. That is indeed a fascinating idea, H.R.

    I think most progressives (and conservatives) are too caught up in general talking points and are often trying to use Saul Alinsky tactics to demean or belittle the other's arguments to the point that they really have not tried to see the argument from the other's perspective. I know that I am often guilty of this at times as well.

    This is very true for most conservatives and nearly homogeneous with progressives. Our friend, John Myste, is actually one of the rare exceptions. If you can coax him out of one of his sarcastic or satirical modes, he often times will indeed articulate the conservative viewpoint correctly. The irony of this is that he usually does so when playing devil's advocate to his progressive fellow travelers.

    ReplyDelete
  2. I remember seeing something not too long ago (might have been Reason magazine) where someone did a study of this sort of thing. Turns out that conservatives can emulate progressive answers to ideological questions much better than progressives can emulate conservative answers.

    ReplyDelete
  3. I'd be interested in seeing that, Joseph. In my experience, I spend a lot of time telling conservatives that they're knocking down strawmen instead of addressing my actual point. This, of course, proves nothing. I suppose I don't notice it as much with liberals because I don't argue with them as often.

    I do find this an interesting idea though. One time, while arguing with a Christian, he kept insisting that I wasn't seeing his point. I then wrote out exactly what he was saying from what I understood, and he then had to admit that I did understand what he was saying. What he refused to acknowledge was that even though I understood him perfectly, I still found him to be spectacularly full of crap.

    ReplyDelete
  4. Yes, interesting. This idea seems in line with the tradition of high school debating. You know, you are assigned to a team of three to defend a proposition irrespective of your personal views. (I was always ambivalent about debating, but I can see its educational value.)

    Sometimes, though, I feel as if I waste too much time and effort on left-wing ideas. It's a bit like my knowledge of religion. I'm slightly embarrassed to know as much as I do and to spend as much time as I do thinking about something I have long rejected as being without foundation. (Something inside me is saying, 'Move on, for Christ's sake!')

    ReplyDelete