Monday, October 13, 2008

Guilt by association?

Anyone who has taught logic knows that once students have learned about the main logical fallacies they start seeing them everywhere – and mostly where they don’t exist. Not every use of emotional language involves the fallacy of “appeal to emotion.” Not every appeal to some authority counts as a fallacious argumentum ad verecundiam. Calling someone a name does not necessarily make one guilty of committing an ad hominem fallacy. And so forth. It takes subtlety of thought, interpretative fair-mindedness, and experience to be able to determine when some argument truly commits a fallacy, as opposed (say) to being merely incomplete or awkwardly stated, offering a mere illustration (rather than making a hasty generalization), or concluding on the basis of objective, rational grounds that such-and-such a person is morally corrupt or incompetent (rather than simply flinging around harsh rhetoric).

In particular, while fallacies are no doubt very common in political contexts, my view is that they are far less common there than is often supposed. Given the nature of democratic politics, it is difficult or impossible for any politician to make a very sophisticated case for a policy in the course of the standard political stump speech or debate, and unreasonable to expect him to. The average voter simply does not have the time, interest, patience, or expertise required to understand the relevant issues. Hence a politician seeking to persuade his audience has inevitably to rely on slogans, anecdotes, illustrations, jokes, and at most very simple arguments. He has to leave logical gaps in his presentation of his case, but gaps that could easily be filled in by the politician himself, or at least by his more skilled supporters or advisers, if the opportunity presented itself, and which a reasonable and reasonably well-informed listener would also be able to fill in for himself. This is true of conservatives and liberals alike. While it is easy and satisfying to pretend that it is the people whose politics you dislike who commit all the fallacies and right-minded people like yourself whose arguments are always complete, well-thought out, and free of error, this is a self-serving fantasy. Each side is “guilty” of presenting crude simplifications in the standard contexts (debates, political conventions, etc.) but each side (or at least the better informed representatives of each side) could also, if given the chance, make a sophisticated case for his position by appealing (say) to various technical economic or moral-theoretic considerations. One reason they do not do this in such contexts (though of course not the only reason) is that the average voter would tune out in about five seconds.

What, then, of Sarah Palin’s accusation against Barack Obama that his association with likes of William Ayers and Bernardine Dohrn shows him to be a man who “pals around with terrorists”? We are assured by her critics that she has committed a fallacy of “guilt by association.” But has she? She has not. Consider first that the most clearly fallacious attributions of guilt by association are instances of the formal fallacy of the undistributed middle term, as in: Bigots oppose hate crime laws; Jones opposes hate crime laws; so Jones is a bigot. This is an attribution of guilt by association because it alleges that the fact that bigotry is associated with opposition to hate crime laws and that Jones is associated with opposition to hate crime laws shows that Jones is associated with bigotry. It is fallacious because, given that the middle term those who oppose hate crime laws is undistributed, the argument shows no such thing. The fact that bigots oppose hate crime laws doesn’t entail that all those who oppose hate crime laws, such as Jones, are necessarily bigots, any more than the fact that all cats are furry shows that all furry things are cats. Some people who oppose hate crime laws might (indeed, do) oppose them for reasons having nothing to do with bigotry; for example, they might fear that such laws could lead to restrictions on freedom of speech, or that they amount to a totalitarian attempt to punish certain politically incorrect thoughts.

Now Gov. Palin is not guilty of this sort of reasoning. Her reasoning seems instead to go something like this: Ayers and Dohrn are terrorists; Obama pals around with Ayers and Dohrn; therefore Obama pals around with terrorists. Unlike the sort of “guilt by association” embodied by the fallacy of the undistributed middle term, this argument is formally valid. Are its premises true? Well, Ayers and Dohrn certainly were terrorists, apparently by Obama’s own (belated) admission; and while they’ve now left that “line of work,” they are also apparently unrepentant about having engaged in it. So while they are no longer active as terrorists, it is arguably no less implausible to assert now that Ayers and Dohrn are terrorists than it would have been to assert in 2001 (the year before his death) that Abu Nidal is a terrorist, even though he was at the time retired and living unrepentantly in Baghdad.

Regarding the second premise, we know that Obama had a long-standing and friendly professional relationship with Ayers (and, if less directly, with Dohrn, who is Ayers’ wife). It would be silly to quibble over whether this amounts to “palling around”; if we changed the second premise to Obama has a friendly and long-standing professional relationship with Ayers and Dohrn and, accordingly, the conclusion to Obama has a friendly and long-standing professional relationship with terrorists, it would hardly be less damning.

Does the argument nevertheless commit what Antony Flew calls the “masked man fallacy”? This is the sort of fallacy committed in an argument like: Commissioner Gordon knows that Batman wears a bat costume; Batman is Bruce Wayne; therefore Commissioner Gordon knows that Bruce Wayne wears a bat costume. The argument would commit this fallacy (if at all) only if, just as Commissioner Gordon doesn’t know that Batman is Bruce Wayne, neither does Obama know that Ayers and Dohrn are terrorists. But of course Obama not only knows this now, but has known it as long as he has known them, because everyone, especially in the circles in which they and Obama both ran, knew it.

Is the force of the argument undermined by the fact that Obama no longer “pals around” with them? To think so would be to entertain another silly quibble. Their association ended only a few years ago, and evidently for political reasons; indeed, it was only because of political pressure that Obama recently distanced himself from them, just as it was only because of political pressure that he distanced himself from the repulsive demagogue Jeremiah Wright. And again, if we altered the argument to take account of this quibble, the result would hardly be less damning: Ayers and Dohrn are terrorists; Obama had a friendly and long-standing professional relationship with Ayers and Dohrn, at least until political motives led him to distance himself from them; therefore Obama had a friendly and long-standing professional relationship with terrorists, at least until political motives led him to distance himself from them.

So, quibbles aside, Gov. Palin’s accusation seems to be a straightforward statement of easily verifiable fact. But what about the objection that it nevertheless distracts us from the “real issues” (e.g. the economy)? Is she saying something true but unimportant? Surely not. As others have pointed out, were a Republican – Gov. Palin herself, say – known to have had a friendly and long-standing professional relationship with someone like an aged and unrepentant Terry Nichols (Timothy McVeigh’s accomplice), then liberals would quite rightly regard this as very serious evidence that such a Republican is unworthy of the presidency. They would (again quite rightly) have no scruples about saying that such a Republican “pals around” with terrorists, and would not pretend that pointing this out was somehow a distraction from the “real issues.”

In particular, if such a Republican not only “palled around” with the likes of Nichols, but also had a decades-long relationship with someone like David Duke, had Duke officiate at his wedding (supposing Duke got a minister’s license for the occasion), borrowed a line from one of Duke’s speeches for the title of his book, etc., then liberals would – once again, quite rightly – regard this as an outrage, and, more to the point, as very strong evidence that such a Republican secretly held very extreme views even if he portrayed himself as mild and moderate in public.

But of course, this is exactly parallel to what we know of Obama – and yet Obama’s supporters expect us to disregard his associations with Ayers, Dohrn, Wright, et al., and take his public nice-guy persona at face value. It seems to me there can be only two reasons for this attitude, one irrational, the other worse than irrational: either Obama’s supporters, who have so much invested in him politically and emotionally, simply cannot face the ugly implications of his associations; or they do not care about these implications, and perhaps even sympathize themselves with the despicable and extreme views held by these associates of Obama. I have no doubt that a great many of Obama’s supporters, indeed probably most of them, fit the first description. But it is hard not to worry that a significant number of them fit the second.

Finally, some might object – indeed, some have objected – that Gov. Palin’s remarks are inflammatory. Even if they are, that doesn’t show that they are not true. And if they are true, as they certainly seem to be, then if people fear the possibility of an Obama presidency as a result of them, that is Obama’s fault, not Palin’s. Surely voters have a right to know whether a candidate has such associates – whether on the extreme right or the extreme left – and are right to believe that such associations constitute serious (if of course not infallible) evidence concerning the candidate’s true views, and concerning the soundness of his judgment. Again, suppose a Republican were known to have associated himself with the likes of Nichols and Duke, in the manner described above. If liberals were to make a fuss over this, as they would be right to do, then people would, rightly, come to fear such a Republican. And the Republican himself would be to blame in such a case, not the liberals. If a politician does not want people to fear and loathe him, he should not do things to make himself seem worthy of fear and loathing.

Ironically, if anyone has committed a fallacy of guilt by association in this connection, it is Obama’s defenders. Rep. John Lewis recently disgracefully insinuated that Sen. McCain and Gov. Palin are comparable to the segregationist George Wallace, apparently on the basis of the following sort of reasoning: McCain and Palin have stirred up fear; racists like Wallace stirred up fear; therefore McCain and Palin are racists like Wallace. Note that this argument is of the same structure as the “Bigots oppose hate crime laws” example given above, and just as fallacious. As is so often the case with left-wing criticism of conservatives, what we have here seems to be an instance of what the pop psychologists call projection.

1 comment:

  1. Hi Professor Feser,

    Very interesting your comment on fallacies, because I suspect that most people abuse of them. They misconstruct other people's arguments, and then "spot" the fallacies.

    In my view, you can always accuse any person of a logical fallacy, if you force a non-charitable interpretation of his argumentation.

    Also, in my opinion, logical fallacies fans tend to have a "either/or" mentality, because most example of logical fallacies are simplistic.

    In logic, you use the category of "proof"; but in science and daily life, in most cases, you can't prove anything; you can only give good reasons and acumulative evidence for a given hypothesis or opinion.

    So, if I say that you can't believe in a well-known liar, you can say that I'm committing the ad hominem fallacy. But in a practical world, it's very reasonable to disbelieve in the claims of a well known liar (because experience teaches that in most cases, he's lying).

    A logical purist can reply: "But the fact he's a liar doesn't prove that his specific claim is false". It's true in a logical sense of "proof". But in a practical world, if you can't verify his claims by independent means, and he's a well-known liar (or a person with an interest to fool you) you can assume (based upon his precedent of lying as evidence) that he's probably lying.

    Another problem that I see is with informal fallacies. In my view, they aren't logical errors, but errors of content based upon previous (philosophical) assumptions. And most of them can be given a correct logical form.

    For example, take the post hoc fallacy. You can't assume causality because a sucession.

    But is it a logical error? if you reconstruct the post hoc fallacy in a logical form, you can see that it isn't a logical error, but an error of conception or content:

    a: All succession of events implies direct causation between them.

    b: B follows Z

    c: B is caused by Z.

    Given that argument, it's logically correct to accept that B is caused by Z. The problem with that argument isn't formal, it doesn't have to do with the internal coherence of the argument, but with content: the first premise is false.

    But the falsehood of any given belief used as a premiss is a epistemological question, not a logical one. As consequence, informal fallacies assume a priori a certain tacit premise as false.

    In my opinion, informal fallacies aren't intrinsic logical characteristics of specific arguments; but a posteriori judgments or evaluations about them based on previous philosophical (epistemological, metaphysical, etc.) and personal assumptions that implies the falsehood of one of the tacit premisses of the supposed "fallacious" argument.

    Only a ("fallacious"?) opinion...

    Sorry if my english is broken :-)