Take the Ideological Turing Test

My guest blogger alumnus, Leah of Unequally Yoked, has proposed a very interesting challenge for Christians and atheists alike which she calls the “ideological Turing test“:

One of the greatest gifts of my time at Yale has been living, writing, and arguing in a community of smart people with whom I fiercely disagree… The imitation test has helped me make sure I really understood what I was rejecting and, in the end, embracing.

Unless your enemies are purposefully contrarian… there is something they find uniquely compelling about their ideology. To imitate them, you need to know what that is and understand why it moves people. It doesn’t matter if the benefits of an ideology are outweighed by its drawbacks; unless you can recognize the good as good, no partisan will ever trust your analysis of their creed.

And, unless you’re uncommonly brilliant and perceptive, it will do you a lot of good to confront the merits of the other side.

The basic idea is that both Christians, and atheists posing as Christians, will answer a slate of questions aimed at people professing a Christian viewpoint. Then the atheists, and Christians posing as atheists, will answer a similar slate of questions for people professing an atheist viewpoint. Finally, a panel of judges will read all the answers and see if they can tell the difference between the people who genuinely hold each viewpoint and the ones who are merely trying to imitate it. (See here for more detailed rules.)

The point of this exercise is that, if you can convincingly argue the other side’s position, it’s good evidence that you truly understand it and aren’t merely rejecting it out of ignorance. I think this will be a fun game to play, and the outcome, regardless of what it is, should be interesting fodder for discussion and analysis. My only concern is that this may be a difficult game for the Christians – answering convincingly from the atheist viewpoint might require what they’d consider blasphemy – but if they’re willing to play along, that’s up to them.

If you’d like to help out, either as a participant or as a judge, leave a comment here or or on Leah’s blog, or e-mail me or her (leahDOTlibrescoATgmailDOTcom). Leah tells me we particularly need more questions targeted at the atheist viewpoint, so if you have suggestions for those, please propose them.

About Adam Lee

Adam Lee is an atheist writer and speaker living in New York City. His new novel, Broken Ring, is available in paperback and e-book. Read his full bio, or follow him on Twitter.

  • Tom

    An intriguing idea. Is it really true, though, that it’s impossible to explain an idea without understanding it (or, put less strongly, that an attempt to explain an idea without comprehending it will always be less coherent than that of someone who does comprehend it and is otherwise of similar competence?)

    Could a candidate not memorise stock answers to common questions, and regurgitate the words as necessary without even comprehending them, Chinese-room style, and potentially score quite highly? Is it possible to develop an internal model, entirely abstracted from all other experience, of a pattern that itself models parts of the world, yet sufficiently accurate to make successful predictions of its behaviour, without becoming aware that that pattern does in fact accurately model the world as experienced? The experiment seems to require that this be false as one of its axioms; but is it so? That’s a really difficult question; probably something only someone extremely familiar with the best theories of consciousness itself might be able to answer.

    I’ve certainly heard tales, but not verified them, of staunch-creationists (tales sometimes even recounted by themselves, not that that really has any bearing on their truth) who somehow manage to get educational qualifications in evolutionary biology by doing something like this; learning the patterns in the material (or just flat-out rote learning standard answers) sufficiently well to answer questions without ever attaching any worldly relevance or validity to any of it. But perhaps that’s just an idictment of excessive predictability or susceptibility to gaming in current educational methods themselves, not the actual data they attempt to impart.

  • Steffen

    Very interesting experiment. I think I once did something like this.

    In the (deeply catholic) part of Germany where I grew up, there was a blatantly unfair procedure at high-school which totally favoured religious class: For the final exam, one could choose to be examined in either a) Religion or b) History, Political science AND Sociology. So one had the choice of either religion or three extensive topics (and having to learn more than three times more stuff). Of course, nearly everybody choose religion, including me.

    Nevertheless, despite I viewed myself as an Atheist (I became one at around age of 10, and I was already very critical of religion before), I got nearly full points in this exam. This exam didn’t consist of reciting memorized knowledge, but the jury tested whether one had really internalized christian doctrine. For example, I was asked to compare christianity to buddhism, and to prove why christianity is the truth and buddhism is a false religion (I was really asked this; a blatant violation of neutrality of the state towards religions). Of course, all my answers were completely feigned; but I obviously made easily the impression of being a genuine, faithful Catholic.

    This observation would confirm the assumption that Christians will be less successful at the “ideological turing test” trying to pose as Atheists, compared to Atheists posing as Christians. When taking into account that Atheists and Agnostics scored best in a test of religious knowledge some months ago, this should come at no surprise. Speaking for me, I reject Christianity and other religions because I know them and their “holy” scriptures.

  • Alex SL

    Did she get that from Paul Krugman? He recently said something similar about Keynesians vs. Neoliberals.

    The thing is, it may make more sense for a divide like that than Christians vs. atheists, simply because there is not one Christian position, but hundreds, and there are also several variants of us atheists, most obviously weak and strong ones. Many atheists would laugh hollowly at the suggestion that they consider the existence of god to be decisively disproved, and call it a straw man argument, but others would consider the proposition to be a more or less accurate description of their viewpoint. And that is only one issue among many! Compare also the so-called “scientism” of a Jerry Coyne who, like myself, bases his rejection of gods mostly on lack of empirical evidence for their existence, with the philosophical ruminations of fellow atheist Massimo Pigliucci, who views that stance as a naive and uninformed. No matter what a Christian said, there would always be atheists who would call foul, and the other way round.

  • Alex SL

    Damn, looks like I forgot one of the closing tags. Sorry about that!

  • http://rationalcrank.blogspot.com/ rational crank

    This does sound like a fun exercise but I agree with Steffen. This doesn’t seem to be a test on understanding a concept as much as a test on how well you can bullshit other people (and yourself). On that skill the Christians might have the advantage. Wouldn’t this exercise also test the Judges? How can a Christian appraise whether a participant has a good grasp of atheist concepts, and vice versa?

  • Alex Weaver

    Which “the” Christian viewpoint will we be considering here?

  • http://www.commonsensethoughtcontrol.com Tim

    I think Caplan’s plan is flawed simply because of what it means for a person to be a self-identified libertarian (“libertarian social science Ph.D.s”). Get 10 such people in a room, and I’d wager half would argue the other half aren’t what they claim. “No true Scotsman” writ large.

    Similarly, I wonder if any sort of test could objectively compare atheism to any particular sect of Christianity. Even if you were to nail down specific classes of each group, I wonder if those in the group would even be able to self-identify. The value of the Turing test is that almost anyone can identify after a short conversation a non-human test participant AND a human test participant. The ideological Turing test doesn’t appear to have that capability – as even people of supposedly the same ideology would likely not identify their co-believers as such, if such a question weren’t explicitly asked and truthfully answered.

  • Jormungundr

    This is an interesting idea. I’m not sure that the execution would work out all that well though. Whether you do the political one or the religious one you end up with the problem of there being many different versions of Christianity and libertarianism. Atheism is similarly complex in the reasons why people claim to be atheists.
    I am interested in the results of this, but I suspect that the judges will be unable to meaningfully tell the difference between real Christians/atheists/libertarians/whomever vs pretenders. I’m imagining a real life Southern Baptist minister and an actual anarcho-capitalist being determined by the judges to be silly caricatures of real Christians and libertarians. You could write the mildest, most well thought out religious or political views or you could write out the craziest extreme and there would probably be people in real life agreeing with what you wrote.

  • mlg

    I don’t expect the judges to be able to do better than chance. But I’d be happy to be a judge or a participant. I’m an atheist. Submitted my email with the comment form.

    Possible questions for the Atheist viewpoint: depends, how broad do you want them to be?

    “In [one sentence; a few sentences; a short paragraph] describe why you are an atheist.”

    “Have you read any of the popular “Gnu atheists”? How closely do you agree with them? Are there any points of disagreement?”

    “Do you describe yourself an an atheist or an agnostic, and why?”

    Can’t really think of very many… maybe this won’t work so well.

  • http://verbosestoic.wordpress.com/ Verbose Stoic

    I’d be interested in participating as a Christian, although it must be conceded that I’m an odd one. No issues of blasphemy for me.

    I’d offer to judge, but really I don’t think I’d find judging to be in any way fun.

    I don’t really expect this to produce any really meaningful results, but it should be interesting to see the responses.

  • mike

    Ugh, you don’t ask direct questions about religion or atheism. That invites the easy fakery of the gregarious respondent seeming genuine.

    Ask for an opinion on video taping police officers. (Do they revere authority? Do they fear authority? Do they commiserate with authority?)
    Ask for an opinion on police brutality. (Again, you’ll get much about their view on authority but also their supposition of a just-world or a real world)
    Talk about the Catholics’ charity in the third world. (Did they mention Mother Teresa, AIDS, the gilded Vatican. How do they view charity and poverty?)

    Ask around issues to discover their axioms.
    Contraception: Is it regrettable or liberating? Do they know the history?
    Slavery and how it came to an end: Was it ok back then? How did abolition occur?
    Smoking and Drinking: Are they vices? What is vice?
    Science: A useful novelty or mankind’s greatest achievement?
    Scantily clad women: Do they employ slut shame or are they indifferent or enthusiastic?
    Trophy wives, cougars, mistresses: What are the sexes in relation to each other?

    You’re not testing someone’s ability to wear a mask. You’re testing their application of the principles underlying their philosophies to perform common tasks. The Turing Test does not ask “are you human?” and “tell me about your humanity”. Plus there is the danger that adherents to anything do not always know their own history, heroes, hierarchy, or dogma (like theists who love Ayn Rand, who was vilified as a cruel selfish atheist and booed on talk shows).

    If you want a common knowledge base, give them all some recent newspaper articles and place your questions around them. Purposefully balance some sets of articles and unbalance others.

    And the prizes go all around. Everyone is competing. In the Turing Test, there is: most human machine, most human human, least human human (shameful), best judge of true humans, best judge of true machines.

  • http://verbosestoic.wordpress.com/ Verbose Stoic

    Mike,

    The problem is that for all of your questions Christians and atheists could all equually come down on either side. The key when you’re trying to analyze a philosophy is to create questions that allow for a range of answers but analyze the RATIONALE behind the answer to get at if they understand the philosophy or not. So, even questions like “What rationale could an atheist use to restrict contraception?” are good, since even if they go against the person’s personal beliefs the reasoning would indicate an understanding of the atheist position.

  • Jormungundr

    Mike,
    I’m not sure that your proposed questions would be good at separating fake Christians from real ones.
    There are authoritarian atheists and non-authoritarian Christians out there. I’ve heard some accuse atheists of being more authoritarian than most people because of how the Soviet Union and Chinese Communists turned out (yeah, yeah, it’s stupid, but there are people out there quick to claim that atheists are authoritarians).

    There are Christians who are fine with contraception. I read a poll that claimed that most American Catholics claim to use contraception. They are supposed to be officially against it, but people tend to pick and choose which rules to follow.

    For slavery, most anyone would claim that the Civil War ended slavery because the Civil War was about slavery. This is false and ignores the fact that slave states fought for the Union (Delaware, Kentucky, Maryland, and Missouri). But that historical fiction is very commonly believed. It was advocated on this website in fact. I’m afraid that this issue will result in people writing about the Civil War rather than writing about religion.

    Lots and lots of Christians smoke and drink. This isn’t a litmus test for anything. ‘What is a vice’ would likely return responses along the lines of ‘things bad for your health’ by atheists and Christians alike.

    Maybe you could get them on sex questions. But then there are Christians with sexually permissive views and there are very uptight atheists. For all we know a log cabin Republican could take this challenge and write about his right wing Christians views and how he engages in homosexual relationships. It would seem like such a person was fake, but there are people like that. A judge could be easily mixed up on fact vs fiction here.

    Getting them to apply their views to various examples might be a good tactic, but for each matter you could end up with both atheists and Christians going any which way. With all the self-identified Catholics who use contraceptives and whatnot, you could easily get misled by reading someone’s statements.
    Maybe I’m giving the judges too little credit, but I don’t think they’ll beat chance at determining who is real and who is pretending.

  • mike

    Which is why you ask many questions. The odds of someone fooling you on every question is on par with the odds of a coin flip coming up heads on every toss. Each answer is a feature (in the discriminant function sense) and together they may very accurately classify A and B.

    The axioms that I referred to are equivalent to the rationales. Yes, you can ask “What rationale would you use to restrict contraception?”. The respondent is pretending to be an atheist or christian so you don’t necessarily have to put that part in the question.

    Most christians are fine with contraception, but getting them to admit that is interesting. Also, many christians are going to fail at describing contraception when they flat out know nothing about it. I have known people who were looking into getting a vasectomy after their fifth kid because they thought that the rhythm method was legitimate.

    I view vice like most here; as something detrimental to health. Christians view is as against god’s will. Alcohol is prohibited in some circles; and there was that whole Prohibition Era.

    Most atheists actually do (rightly) identify that the Civil War was about slavery. Most christians actually do identify that the Civil War was about states rights.
    Each group does not know or have as much experience defending the opposing view.

    Mention Hitler or Stalin and see how that goes. Most christians do not know that Hitler was Catholic or that Stalin almost became a priest but instead became the head of a church.

    The easiest method to determine that the respondent is pretending will always be that they do not know what they are talking about.

    Some judges may fall for the successful faker, but not all of them will. Computers can beat Chess Grandmasters but they still can’t win at the Turing Test.

  • http://eternalbookshelf.wordpress.com Sharmin

    This is an intriguing idea. A while back, I’d been wondering if I’d be able to convincing write a blog as a religious person. I’d guess that it would be easier for someone who actually used to hold that viewpoint in the past then for someone who was always an atheist or always a Christian. (I used to believe in God, but wasn’t a member of any religion for very long, though I called myself Muslim due to my family and considered Christianity briefly.) I do think, as others have noted, there’s a potential for confusion since there is so much disagreement within groups.