Friday, April 18, 2008

On the validity of ad hominem attacks

It is generally taken for granted that ad hominem arguments are not valid. That is, arguments of the form “x is y, and x said z so not z”. Classic values of y include capitalist, communist, liberal, conservative, Jew, Christian, atheist and Muslim. However, there are circumstances where ad hominem attacks make sense.

It is well known that ad hominems make sense when one is dealing with the credibility of the claiming. If Bob says he saw Sally deal drugs and Bob is himself a convicted drug dealer, we would be less inclined to listen to Bob than if Bob were say a well-respected businessman who gives much of his time and money to charity. This makes sense since the truth of the statement depends on whether we trust Bob. In contrast, if Bob says “1+1=2” even if Bob is a mass murderer and a pathological liar this does not make the statement less likely to be true.

However, there is another circumstance where ad hominem attacks may have a reasonable place that is frequently overlooked; when they are used as heuristics. Few people have the time or resources to examine every issue in detail. Thus we turn to heuristics to gather information. For example, unless one is a climatologist you are unlikely to be able to examine all evidence for anthropogenic global warming. However, one can say “hmm, the evidence I’ve seen makes sense and the vast majority of people who have looked at the matter have concluded that it is occurring. They are likely correct.” Now, here is where ad hominem attacks may become valid. Since one is substituting the reliability and relevance of authorities over ones own logic.

Similarly, if major proponents of an idea are engaging in unethical or extremely unproductive behavior it may be because the idea lacks enough merit to defend on any substantial grounds.

Consider the example of intelligent design, the modern form of the teleological argument that argues that certain aspects of life and the universe demonstrate that there was an intelligent designer of the universe or life. Major proponents of this idea have included William Dembski who is a known liar who spends a large amount of his time rather than doing ID research writing his blog, removing comments on his blog that make him look stupid, and constructing animations with third grade level insults to people that he dislikes. (1)

Now, no one would be shocked to find the idea that Demski is a proponent of has no validity. (If you think otherwise I suggest you read the decision in Kitzmiller v Dover or if you have time read all the testimony)

Now, from a perspective of formal logic, we cannot dismiss intelligent design just because it was constructed solely to sneak God into public schools. Nor can we dismiss it because the major proponents like William Dembski are known liars who seem to do less scientific research than talented undergraduates. But, if someone does not have the time to examine the issue in detail, these are useful heuristics for whether or not they should take the matter seriously.

It is completely reasonable for the time-pressed individual to look at this sort of event and conclude that this isn’t an idea that is worth their time examining. Real scientists who are proponents of controversial ideas don’t try to push their ideas into public high schools and spend their time making flatulence-filled animations about federal judges. And it isn’t unreasonable for someone to conclude that if Dembski’s ideas had serious merit he’d be spending his time more productively. And this sort of argument may be valid in a variety of circumstances, not just when dealing with fringe science or pseudoscience issues but also in others such as politics.

So given this, I have a few questions that I don’t as of yet have good answers to. When can we use such heuristics? If a proponent of an idea is a drunk, or a jerk does that mean we can dismiss it? More precisely, what exactly are the relevant ad hominem attacks and which are not relevant? And how prominent a proponent of an idea does someone need to for the proponent’s behavior need to be relevant? Finally, is it even clear that getting enough information to reliably use this heuristic will ever take less effort than actually studying the area in question?

(1). See for example (and note also comment 23 there by Dembski)


Elan said...

You're wrong. Ad hominem attacks are invalid because you suck.

Anonymous said...

It occurs to me that it's possible no one will ever read this. It also occurs to me that this is a roundabout way of saying what is already obvious.

But anyway...

To answer when it's appropriate to use ad hominem attacks you first you have to determine when it's appropriate to substitute "the reliability and relevance of authorities over ones[sic] own logic." And just how big said substitution will be.

Here's one solution to this first case:

(The significance of reaching the correct conclusion on an issue) divided by (the extent to which you can efficiently and correctly obtain the answer for yourself) = your reliance on someone else.

where significance is a value ≤ 1, with 1 being life and death; and your ability to obtain information is ≤ 1, with 1 meaning you are the world's most preeminent authority on the issue.

Obviously the greater the output the more you must rely on someone else.

From there you can answer when ad hominem attacks are relevant, since their relevance directly correlates to the degree to which you must rely on someone else's authority. For, the greater your reliance on others, the more careful you must be in selecting who "they" are.

To take your case of the jailhouse drunkard and his contention that 1+1=2:

.0001 (significance) ÷ 1 (your own authority) = .0001, which indicates your reliance on his authority and the relevance of ad hominem attacks (i.e. - not at all relevant).

If your life depends upon you answering an obscure question related to baseball, say, and you had never watched a baseball game in your life, it would look like this

1 (significance) ÷ .0001 (your own authority) = 10000, meaning that you would be almost entirely reliant on someone else, who would, via the significance of his answer, be rightfully subject to ad hominem attacks.

I would also say that this holds true for organizations, where organizations can be substituted for the individual (you). For instance, a potential Supreme Court judge is subject to more ad hominem attacks at his confirmation hearing than a lowly cabinet appointee since the significance of his authority is greater in relation to the Senate's ability to make judicial decisions for themselves - they can't.

The ultimate case of this, is the relationship between the voting public and the President, where the voting public's reliance on the president approaches 1, and it's ability to substitute its own authority in place of the president's approaches 0. In which case, almost all ad hominem attacks on a Presidential candidate are justified.