Wednesday, May 21, 2008

More on the Argument Against Trumping

Alan Rhoda has responded to my criticism of his Trumping argument. (There has also been discussion elsewhere; ocham has some notable links.) First, a clarification:

In addition, it seems to me that Brandon is falsely opposing "authority" and "reason". When I reason, I do on the basis of evidence, of which there different types - empirical, intuitive, and testimonial. An authority is simply a good source of evidence. Thus, in Brandon's example, the physicist is an authority for me because he is, presumably, a very good source of testimonial evidence regarding matters physical. In correcting my thinking he's not giving me evidence over and against my reason. Rather, he's helping me to reason things out more adequately by improving my pool of evidence.


There was, however, no such opposition in my argument; my argument said nothing about the relation between authority and reason. Indeed, given Alan's original post it was impossible to do so; there was no way of identifying from Alan's post what 'Trumping' was supposed to be, if not simply correction of reasoning by authority, so it would have done nothing for the argument to try to pin down the point. All that was in view was the fact that we correct our reasoning in light of authority, and the limits of such correction.

Now to the rest. Consider Alan's further clarification on Trumping:

What I called 'Trumping' is the practice of seizing upon some particular authority or claim, one that is not itself a deliverance of human reason or understanding, and refusing to submit that authority or claim to rational critique. In other words, the Trumper has a cherished theory or dogma of which he says, "I don't care where the evidence and the argument may go from here on out, I'm going to stick to my theory no matter what."


This is helpful; but it was not in the least the impression conveyed by Alan's post, which put forward Anselm and Helm, and then expanded outward. Much as I often disagree with Helm, this is not a plausible characterization of him, and it is an even less plausible characterization of Anselm.

But let's take this further clarification, and examine it more closely, because it is still not so clear. Trumping is characterized in two ways here:

(1) taking an authority or claim that is not itself a deliverance of human reason, and "refusing to submit that authority or claim to rational critique";
(2) having a cherished theory or position and refusing to regard any evidence and argument on the subject from then on out.

I've slightly modified the exact wording, since, again, Alan's phrasing seems to me to be tendentious, and therefore not suitable to an argument that is attempting to show a problem with the position, rather than simply state it or insinuate it. I've left the second clause of (1) in quotations because I have no clear notion what Alan means by it. All of his explicit examples have simply been of deferring to an authority or principle that one has already evaluated as perfectly reliable in a given field of thought. If this were all that were meant, my criticism in the previous post would apply in spades; but clearly it is not. However, the only thing that can be added to it is refusing to regard any additional reasoning or evidence as relevant to evaluation of the authority or principle in the first place. We will get back to this in a second. What I would like to point out first is that by neither (1) nor (2) do Alan's new examples of Trumping, Lewontin and Hume, count as Trumping. This is not fatal to his argument; but I would suggest that they provide a way of seeing a serious flaw in it.

Let's take Hume first, which is the easier case. Hume's conclusion is, first, a deliverance of reason; he has already by this point given an elaborate, if somewhat tangled, analysis of why we don't and shouldn't accept testimony of miracles unless the evidence for them meets a certain standard, and argues that it has never been the case that testimony for religious miracles meets that standard. The point intended in the Jansenist case (among others) is that even here, which occurs in a civilized nation in modern times, where we have the results of close contemporary investigations and a wide variety of witnesses, the evidence fails to meet that standard. So whatever one's view of Hume's position here, it is not an unreasoned one. Second, Hume doesn't absolutely rule out miraculous events or testimony for them; he just argues that the standard of evidence has to be so high, and testimony in religious matters tends to be so unreliable, that testimony does not suffice to support a religious miracle: we can hear of the miracles at the tomb of the Abbé Paris, or of someone rising from the dead, and reasonably dismiss the report (indeed, his argument is that to be reasonable we have to dismiss the report). Hume's argument is, in fact, a cousin of Alan's own: he, too, is making an argument against the use of trumps, at least of a certain kind (quite explicitly, in fact); indeed, given the way Hume understands miracles and the evaluation of testimony, the Humean could be forgiven for regarded Alan's argument as simply an argument against testimony for the miraculous with rational laws substituted for natural ones.

Lewontin is more easily misinterpreted, and it is unsurprising that he has often been so radically misunderstood, but in the context of the review he is actually making a point, not about how we should view materialism, but how materialism actually functions in scientific thought, because he thinks Sagan (whose work he is discussing) has clearly exaggerated the chasm between materialist and supernaturalist thinking, and thus is playing up what he sees as parallels. The argument is not about trumps; the argument is simply that all of Sagan's attacks on beliefs about the supernatural can, with surprisingly slight modification, be turned against scientific inquiry. The supernatural is absurd, contrary to common sense? So are some obviously scientific claims. Science, unlike the supernatural, makes good on its claims? Yet, on investigation, it often does not, if judged by the same standard. The supernatural involves unjustifiable just-so stories? So, it turns out, does some scientific work. The consubstantiality of the Trinity is absurd, but we are to accept as solid science the duality of light. The point of Lewontin's brilliant criticism is not that materialism is a trump; it's that it's much harder to pin down why materialism is rational and supernaturalism isn't than people like Sagan want to make it, and that that's the reason why the struggle turns out to be so hard. It's not a struggle between the obviously rational and the obviously irrational; it's a much more human struggle between the resentment of people who see themselves as dismissed from the discussion and the hubris of intellectuals who dismiss them as stupid. Even if we make it about the intellectuals Lewontin is criticizing rather than Lewontin himself (as we would have to), it's not a matter of appeal to trumps, but of passions and prejudices and foibles.

I've taken the trouble to go this long distance around to discuss the actual examples not because I think these examples are especially important to Alan's position, but because they highlight what I think has been a problem with Alan's argument all along: nothing in the arguments given so far is able to distinguish adequately between (1) reasoned deference to an authority or principle evaluated as extraordinarily reliable, such that the authority or principle can be used to identify facts (and, as the saying goes, against facts there is no reasoning); (2) dogmatism due not to an appeal to something as a trump but to simple, ordinary human reasoning turned into inflexible prejudice by human foible and failing; and (3) simple refusal to reason, supported by an authority or principle. Part of what made Alan's first post so confusing, I think, is that it seemed to slide through all three of these. But it's clear that Alan wants the focus to be on (3). And that makes sense: (1) is obviously not a problem, and that (2) is a problem is recognizable on the basis of entirely different principles. I think it turns out to be very, very, very difficult to find clear, real-life cases of (3). Indeed, I think that none of the actual examples Alan has given are even plausible candidates for it. But let's suppose they are common, and consider such cases.

The first thing to note is that not all such cases are irrational. Harriet Beecher Stowe provides a beautiful set of examples in Uncle Tom's Cabin. A theme that is found throughout the book is the danger of refusing to reach a point where you stop inquiry into a matter; and it does a good job of arguing that there are times when it is justifiable to refuse to reason further. This, presumably, is a trump; but as it is very narrowly confined sort of trumping, we will call it a 'minor Trump', as opposed to the 'major Trumps' Alan seems to have in view. When John Bird tries to rationalize his politicking over the fugitive slave question, Mary Bird quickly calls him on it:

"Mary! Mary! My dear, let me reason with you."

"I hate reasoning, John,--especially reasoning on such subjects. There's a way you political folks have of coming round and round a plain right thing; and you don't believe in it yourselves, when it comes to practice. I know you well enough, John. You don't believe it's right any more than I do; and you wouldn't do it any sooner than I."


What Senator Bird calls 'reasoning' is simply extending an inquiry well past its due date; this is not rationality, it's rationalization. It's reason's form of stalling. And Senator Bird very shortly after his exchange with his wife proves by his actions that she was on to something, because face to face with an actual slave, his conscience won't let him obey the law he had argued and voted for. He had known from the beginning that the fugitive slave law was wrong, and that those who refused to return fugitive slaves were in the moral right; but instead of stopping there, like Mary, and fixing it as a "plain right thing," he weaseled around it like a politician. But when push came to shove, he found he couldn't in good conscience treat it as anything other than a "plain right thing". Uncle Tom's Cabin is in great measure a book arguing against moral wishy-washiness -- and this wishy-washy vice comes about, as the book presents it, from refusing to recognize that some things are just "plain right things" and that when you reach them, you have reached all you need to know; and those who go farther inevitably become "swayed and perverted" by sophistry. In Stowe's view only a sophist refuses to draw an inquiry to a close; rational people will always look for something that they can from then on out take as solid.

Of course, Uncle Tom's Cabin is fiction; although any argument that clearly put the Senator in the right and Mary in the wrong would be just begging to be torn apart. But moral issues are a major problem for those who refuse to allow a rational end to inquiry, a point where reason can just take a conclusion as a "plain right thing" and hold that everything that argues against it is, and must be, merely sophistry. It sounds good to say, "We should never let principle trump inquiry." It's not so clear, though, that it's rational to say, "We should keep an open mind about slavery or genocide, because perhaps, just perhaps, inquiry will turn up sound arguments in its favor." It sounds good in the abstract to refuse to allow Trumps. But on moral matters everyone sooner or later reaches a line they will not cross for all the cleverest and most compelling argumentation in the world. The idea being this: show me someone who allows no minor Trumps and I will show you one who twists and turns on every wind of clever sophistry that comes along, because that's a person who will never recognize anything as a "plain right thing".

Nor do you have to go so far as practical ethics to find such things; every foundationalist of any sort has a Trump in the sense of a principle beyond which no further reasoning is to be had. That's what a first principle is: it's the grounding terminus of inquiry, beyond which it is futile to argue. Descartes reasons down to a first principle; end of the line. Everything else must presuppose that, no argument can possibly overturn it. Spinoza does the same in the Ethics; same thing. If there are foundations there are places where one should just that particular line of inquiry and go on to other things. Nor do you even have to be a rationalist to reach such a starting point; Hume thought, and thought rightly, that it was absurd to ask whether bodies exist. Obviously they do, and anyone who pretends to consider seriously whether they do or not is simply lying. The only questions that are reasonable on the subject of bodies are questions that presuppose that they exist. Minor Trump. And surely none of these are irrational, or inconsistent, or problematic in the way suggested by Alan's argument.

Of course, if one accepts, as any rational person must, the existence of such minor Trumps, the question follows whether it could possibly be rational to accept the existence of a major Trump. And it is not so clear what the answer would be, for we seem to have no clear notion of a major Trump, i.e., of Trumping in Alan's sense. A major Trump couldn't differ from a minor by the size of the ground it covered, because there appears to be no principled cut-off. Taking description (2) from above as wholly adequate, there is no reason why a major Trump should be disallowed if a minor one is allowed. But description (1) adds another element that might perhaps make up the difference: it refers to an authority or claim that is not itself a deliverance of human reason. It isn't quite clear what this would mean. Clearly in some sense every claim is going to be a deliverance of human reason; even if God tells me, "Stop reading blogs and become an itinerant beggar," my taking that command up as a principle would be an act of human reason. The most natural reading of it is to say that it is confining the matter to claims that are inconsistent with human reasoning. This does give you a reasonable enough conclusion -- you should not guide your (human) rational inquiry on the basis of principles inconsistent with human reasoning -- but not, I think, what Alan is going for. One might try to read it as talking about claims that are held irrationally from the get-go. And that, too, gets us something plausible: you should not govern your reasoning on the basis of irrationally-held claims. But that's also not so helpful, and doesn't fit very well with any of Alan's arguments (and certainly not with his examples). So I don't see any way out here.

The whole argument, as I said in the first post, strikes the ear very nicely, and it sounds plausible. Certainly a great many people have found it an attractive line of reasoning. But when we look at it more closely, we find that there does not appear to be any consistent and plausible analysis for it. The words are associated with what seem to be the right words, but when we try to see how they are put together, we don't seem to come up with anything of substance. And it seems to me that the attraction the argument has to many intelligent people says less about the integrity of the argument and more about the muddled values of the time; we're big on words like Reasoning and Inquiry and not so big on reasoning through and inquiring into what they would actually require. The cliché takes so much less effort. As I said before, what rational person wants to argue that there are principles that "Trump" reason and inquiry? But that's just the sound of the words; what we really need to know is what we mean when we are talking about "Trumping", and it seems to me that Alan's clarifications and examples have not made his original argument stronger, but instead shown it more and more to be problematic.