= 27d567bc2d48d9f40e9ccc20ea370b15)
Vaniver comments on Rationality Quotes December 2011 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (577)
Emphasis mine. That's the part that's the result of bias (i.e. primitive and illogical).
However, Wittgenstein is not criticizing Malcolm just for supposedly having wrong factual beliefs, but for mere willingness to use probabilities about beliefs and behavior of people that are conditional on their natonality. He is objecting to the very idea that the probability of the British government commiting a certain act may be different from the probability of some other government committing it, or that certain broader norms that also prohibit such behavior might be a matter of exceptionally strong consensus among the British, which would by itself provide strong evidence that their government is unlikely to exhibit it.
I think we are interpreting Malcolm's position very differently. Malcolm isn't saying "I would be surprised; I put a low probability that the British government would do that." Malcolm appears offended- it is impossible because the British are too decent. You are right that one could, say, be less surprised by an American assassination attempt than a Canadian assassination attempt based on past actions of the governments, but that's not what Malcolm is doing here. He's exhibiting a nationalistic, self-serving bias, which Wittgenstein is right to object to.
I am not concerned with whether Malcolm was correct, and I'm not saying that Wittgenstein had nothing to object to. This is not a situation where we're judging them as symmetrical parties in a debate, but a situation where we discuss whether Wittgenstein's position deserves to be pointed out as an outstanding example of rationality. And it seems tome that even if one takes a much less favorable view of Malcolm, Wittgenstein is still displaying a fair amount of mind-killing biases.
Would you mind naming those biases for me? I'm having a hard time seeing what you're talking about, and suspect that our disagreement may depend mostly on differing interpretations of limited information.
Basically, I think Wittgenstein was too quick to pattern-match every mention of such things as "national character" with propagandistic nationalist ramblings. I suspect this is just an instance of a bias that's been widespread in the Western world for quite a while now, namely the tendency to write off the use of certain kinds of conditional probabilities about people, including most of those conditioned on national origin, as inherently incorrect or immoral. (With a lot of equivocation about which one of these two is actually meant, and how come that the former category just happens to subsume the latter so conveniently.)
Moreover, from the "wouldn't surprise him at all" comment, it does appear that Wittgenstein had, for whatever reason, a biased unfavorable view of the British government. To the best of my knowledge about the state of the world in 1939, this would have definitely been, by all standards, an event far too surprising and shocking to characterize that way, under any reasonable interpretation of that phrase.
Finally, Wittgenstein's reaction is reported as "furious," and he describes himself as "shocked." It seems clear that the very fact that someone got into a shocked and furious state of mind during a conversation about controversial and mind-killing topics makes it very highly probable that at least some sort of bias has kicked in, even if my above guess doesn't identify it correctly.
(Come to think of it, I don't find it implausible that Wittgenstein could have been intentionally baiting Malcolm, hoping for an opportunity to show off some sanctimonious indignation. But lacking any detailed knowledge of his character, this is nothing more than idle speculation.)
You suspect he's too quick to pattern-match every mention from a single example?
I'm aware this is a common bias now, but I don't think it was that widespread in 1939.
Perhaps this is because I have an unfavorable view of governments in general, but it seems that for an even slightly cynical student of history assassination attempts on rival heads of state by a government should not come as a surprise, especially as monarchies were replaced by democracies. It's not clear he was singling out the British, and even if he were singling out the British, it's not clear if that was the result of bias or cool calculation. (The British did have the best spy network in Europe, although whether or not Wittgenstein would have known that is not something I am able to guess.)
I agree that being biased can lead to fury, but I think for someone as passionately logical as Wittgenstein seeing bias, especially in a friend, could also lead to fury. It's not clear to me that his immediate reaction is evidence between those hypotheses, and his persisting fury strikes me as slightly better evidence for the latter. (Background: Almost twenty years earlier, Wittgenstein was rebuked as a teacher because he would also beat the girls if they made mathematical mistakes.)
That is, it is possible that Wittgenstein was biased in pronouncing Malcolm's bias, but it seems to me unlikely. The evidence seems to point the other way, especially the conclusion he draws- that philosophy should help one with the important questions of everyday life.
Admittedly, this is speculative, but from his tone I did get the impression that he was prone to such matching.
Actually, a very strong taboo against such assassinations follows from a very cynical theory. Namely, it provides for a convenient Schelling point for national leaders, where they can otherwise escalate war as much as they like without fear for their personal safety. (As long as they don't let themselves get totally conquered, of course.)
But more importantly, who are all these heads of state supposedly assassinated under orders from rival governments prior to 1939? Can you name any attempts of such assassinations in the period of, say, one hundred years preceding 1939? Or even just cases where the culprit is unknown, but a plot directed by a rival government seems plausible?
(The closest example I can think of is the killing of Engelbert Dollfuss that kicked off the coup attempt in 1934 by the Austrian Nazis, who were clearly acting in concert with Berlin. But even that was an all-out coup attempt accompanied by an armed Nazi uprising across the country, so not really an assassination plot, and also symptomatic of the new and unprecedented wave of political gangsterism of which the British government was not a part.)
Given this history (or rather a lack thereof), do you think that it was possible for a non-biased observer in 1939 to view the accusation against British government plot to assassinate the German head of state as unsurprising if true?
The Schelling Point is stronger for monarchs than for ministers, and so as monarchies disappear or become less relevant one could expect assassinations to increase.
Does the first such assassination have to be a surprise? I think unprecedented events can be unsurprising if they were anticipated. We may be interpreting Wittgenstein's comment very differently: for me, "X would not surprise me" means something like P(X)>.01, not something like P(X)>.5.
But, since you asked, consider the case of Tomas Mac Curtain, assassinated by British forces in 1920. I don't think he qualifies as a head of state, but he was a fairly prominent government official. In 1939, it would have been plausible that Napoleon had been poisoned by the British (though that's a bit outside your hundred year window).
I don't think this is the question you meant to ask, but I think that the accusation should have been entirely unsurprising, regardless of its veracity. I think that most observers would assign a higher probability to the accusation being false than true, but I don't think an unbiased, moderately informed observer could put the chance that the accusation was false at less than ~5% on the day the accusation was reported.
Why? Even an elected ruler who breaks the convention against assassination makes himself fair game for immediate retaliation of the same kind. I don't see why the transitory nature of his office would make the incentives significantly different.
This is not evident in the quote you talk about. Malcolm didn't use probablities, he called it "impossible". He didn't merely condition his guess partly on the nationality, he seems to have based it entirely on said nationality and on nothing else.
Do you know of any act, no matter of how great charity or barbarism that is so incompatible with "national character" that you can find not one person of that nation willing to commit it?
That is not the relevant question here. The relevant question is whether we can think of acts that are so incompatible with the "national character" that it would be inconceivable (i.e. p~0 can be assumed for all practical purposes) that any institutions of a given country's government would commit them, although such acts have been committed by governments in other places and times. The answer is obviously yes.
I can think of only such acts as wouldn't benefit such governments in question. E.g. it wouldn't benefit the US government to cook alive suspected terrorists and use their flesh to feed its troops. Cannibalism isn't part of the American national character -- and it doesn't benefit the US government either, so it doesn't do it.
But I can't think of any acts that would be effectively impossible to be committed by an institution of any government though it would benefit it, merely because it's "not in the national character" to do so. If something is not in the national character, then said institution merely does it in secret.
For example, given the American national character, it would be inconceivable for the U.S. government to kidnap its subjects' daughters to serve as concubines in the president's harem. (Something that many historical governments in fact did openly.) Do you therefore conclude that this is in fact being done in secret? Or maybe that the only reason why it's not being done is the difficulty of keeping it secret?
Primarily the latter. Consider this:
North Korea abducts women for the president's harem.
South Korea does not (neither openly nor secretly, with p~0).
And yet it's people of the same nationality on both sides of the border. Therefore such things don't seem to me to be primarily dependent on "national character". They seem to be primarily about what each leader can get away with doing. South Korea and America are semi-democratic capitalist states. North Korea is a totalitarian regime.
To get back to my comment where I explained what I consider to be a reasonable interpretation of "national character," I defined it thus:
In this discussion, I am not at all interested in the exact connection that these norms have with ethnicity or any other factors. I merely claim that for whatever reason, there is variation in such norms across governments, which sometimes gives very strong information on what they may be capable of doing.
(And anyway, several decades of life under radically different regimes imposed by foreign conquerors, one of which practices extreme isolation, will cause cultural divergences that run deeper than the immediate structure of clear incentives. Moreover, this one example is not conclusive proof that all such differences in governments' behaviors in all places and times are caused by the same factor.)
But I think that such a definition where "national character" are the norms followed by a a national government and which it's expected to be followed by a broad consensus, leads to bizarre ideas such as e.g. the "national character" of the whole of Eastern Europe must be described as having changed at the fall of communism, even though the fall came from within. So the national character suddenly modified itself, just because the norms of government changed themselves. Eh. I don't think that's really how these words are normally used.
And if we return to the subject of actually secret, non-open operations -- if I believe (which I do) that FSB bombed some of Russia's own apartment buildings (for I am a conspiracy theorist in regards to several conspiracy theories), but that the MI5 wouldn't do that against British apartments, nor would CIA do it for American apartments, I don't think it makes much sense to say that the Russian national character enables Russia to blow its own people up, but that the British and American national characters does not. The character of their respective government structures, sure. But not the national characters.
To the extent that there's a "national character" that affects policy, I feel it has primarily, perhaps even solely to do with concepts of self-identification similar in type to the concept of Clash of Civilizations by Huntington. e.g. Greece supported the Serbs in the Yugoslav wars for no more and no less reason than that its "national character" contained a self-identification with Eastern Orthodox significantly more than with Catholics or with Muslims. Now there's predictive power. In any dispute between orthodox and non-orthodox, I know that Greece will back the orthodox. I know that Arab nations will back the Palestinians against Israel. America in the Cold War self-identified as anti-communist, so in any dispute between people identifying as communists and people that didn't , I know America would back the people that didn't.
There's the extent that national character plays in regards to policy. If there's some other element in it with predictive power, I don't see it.
With some noteworthy exceptions, particularly in Africa. I do generally agree that rules of thumb like this generally have decent predictive power though.
This is just the confounding factor of foreign domination, just like in the North/South Korea example. Of course, like with all political categories, the distinctions aren't always clear, since prolonged foreign domination may gradually cause irreversible changes, or even gradually get to be seen as the normal state of affairs. Still, the different national characters of Eastern European countries have been amply demonstrated when comparing their state both before 1990 and since then.
A better example of what you're aiming for would be periods of political instability in which some extremist faction like e.g. the Nazis grabs power and proceeds to implement extremist policies that would have seemed unbelievable coming from that same country shortly before that. Clearly, such black swan events limit the predictability of any model one uses for understanding history and politics. It doesn't mean they have no predictive power during normal times, though.
These things can't be separated from each other. You are speaking as if the system of government is an independent variable. In reality, formally the same system of government imposed in different places will produce very different results, and these results are very much dependent on what is conventionally understood as "national character."
It's hard to make any concrete predictions without offending various nationalities, so I'll limit myself to offending my own kind. For example, suppose I read a story about an affair where vast millions were pillaged in corrupt dealings some years ago and yet the culprits are happy, free, and untouchable despite all this being public knowledge. If it happens in Croatia, I'll shrug my shoulders. But if I heard about this happening in, say, Denmark, I would, like Malcolm, express disbelief because it would, indeed, sound incompatible with their national character. Even though the laws on the books and the theoretical legal consequences are probably similar in both places.
In addition to my previous reply, and to separate the more controversial part from the rest:
Frankly, if you believe that people running the MI5 or the CIA would be willing and capable of doing something like that, I think you have a very distorted view of reality in this regard. Unfortunately, the inferential distances are probably too large for us to have a productive discussion about it in this context.
(In reality, I don't think CIA would be capable of killing my neighbor's cat without it leaking into the press tomorrow. In fact, they'd probably bungle the task so badly that the leak wouldn't even be necessary.)
That would have been news to many anti-communists, but let's better not go there.
Historically, there have at least been some iodine-poor areas within nations that outsiders might have dismissed as being full of cretins without being wholly unjustified...
They have different citizenships, different cultural messages from birth, different access to such messages from the rest of the world (such as the US). They cannot accurately be described as having the same nationality.
Different something, sure.
But North Korea has more in common with South Korea than it has in common with any other country. And South Korea is probably much closer to North Korea than any other country is to North Korea. Anyway, this is all nitpicking, because ArisKatsaris' main point remains: South Korean leaders don't refrain from harem-kidnapping based on national character, but based on the negative incentives they face.
I'm pretty sure that if you had asked me what "national character" means before this thread, I would definitely have included "personality cult around a wacky dictator"!
Which governments did so? I can only think of some that politely asked families to send them their daughters.
Reread that phrase with a cynic's mind in the context of a power struggle.
How would that benefit the US government?
Well, to be fair the proposition that the assassination attempt had British backing implicitly means that it had systematic approval in at least some government division, so it's a rather stronger claim than saying that at least one British person was willing to commit resources to backing the attack. But it did seem to me that Wittgenstein was criticizing his student for naive and sloppy thinking (which he was indeed guilty of) not for having the belief that there can be prevailing trends in national character.