= 27d567bc2d48d9f40e9ccc20ea370b15)
Keith_Coffman comments on Reverse engineering of belief structures - Less Wrong Discussion
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 (34)
Interesting stuff. I am all for trying to improve peoples reasoning skills, and understanding how particular people think initially is a good place to start, but I'm a bit concerned about the way you talked about knowledge in here (and where it comes from).
Frankly, I wouldn't really look to any person as a source of knowledge in the way you seem to be implying here.
Here's how knowledge & experts work: There's a whole bunch of information out there - literally more than any one person could/cares to know - and we simply don't have the time (or often the background) to fully understand certain fields and more importantly to evaluate which claims are true and which aren't. Experts are people who do have the background in a given field, and they usually know what research has been done in their field and can answer questions/make statements with legitimate authority when speaking on the subject with which they are well versed. Once you have a consensus of opinion between many such experts, you have raised the authority of that opinion further because you've reduced the likelihood of one guy misspeaking, making stuff up, being dishonest, etc. Also note that experts talking on subjects outside of their field of study have no more authority than anyone else (though they often are well informed on other subjects) - this is where the argument from authority fallacy comes from, e.g. "Einstein said that the sky is green" ... so what?
I suspect you know all of this already (I don't mean to come off as lecturing too much, just reiterating some baseline stuff)
After all that rambling about experts, the important thing to take away is that the knowledge (and by knowledge in this context I mean being aware of information which corresponds with reality, i.e. the truth) doesn't come from the experts; experts are just the people who go about investigating the truth and report back to the rest of humanity what they've found. In other words, reality is objective and claims should be evaluated based on their evidence, not the person who proposes them.
All of the examples you've used deal with things which actually do have an objective answer, whether or not we have or feasibly can test them empirically. (also, as a side note, that bit about the 50% chance of being true is ridiculous even if you don't have any knowledge going into it - you would simply say "I don't know if these claims are true")
People definitely have biases, and we should be particularly cautious when dealing with any claims that are related to contentious issues. Further, I'd like to stress the point that just because a large majority of the experts in a field say something it doesn't make it true - but it does mean that we should believe that it is true until new information says otherwise, because frankly an expert consensus is one of the highest certainties we can come up with as a species.
I guess the main thing I am trying to say that directly ties into your post is that we shouldn't really care how someone formed their beliefs when evaluating the veracity of a claim; when we should care is:
When we suspect that a bias may have lead to a false reporting of real information (in which case we would want independent, unbiased research/reporting)
When we want to change someone's mind about something
When we want to keep someone's faulty & infectious belief structure from propagating to other people (ex. Dark Side Epistemology) by teaching other people critical thinking/rationality and common mistakes like said structure.
Still, figuring out how people think has always been an interesting area of science that is worth pursuing, and the tools/sample size have gotten a lot bigger since the time of case studies. I hope you find more interesting stuff to share.
"I guess the main thing I am trying to say that directly ties into your post is that we shouldn't really care how someone formed their beliefs when evaluating the veracity of a claim".
This is an absurd proposition on several accounts. Firstly, a great deal of utterance meaning can only be recovered relative to a particular context, for it has complex and variable uses shifting within and across contexts, i.e. the exchange of agreement formalising marriage is not a mono-semantical reference to an internal psychological state, but does something only understandable relative to a particular convention of marriage. The upshot being that a condition of intelligibility is contextual awareness. Secondly, it is important to at least be aware of the structures of understanding through which particular intellectual subcultures and traditions give rise to scholarly output (i.e. you can't satisfactorily understand and evaluate a Marxist-Leninist work independently the sociological reality of post-Cold War vanguard parties, or modern European intellectual history).
The meaning of a claim can, in fact, change based on the context. Moreover, the truth of a claim may change with time (for instance, the claim "Elvis is alive" was at one point true and is now false. Also note that, in the context of me making up a simple example of a claim to demonstrate my point, the meaning is likely referring to the famous performer Elvis Presley rather than any person named Elvis.
Thus we can see how there are a few things that we need to keep in mind when we address a claim, much as you have said above. However, the truth of the claim, given that you understand the meaning and you are evaluating it at a particular time, does not depend on the belief structure.
The reason I said "we shouldn't really care how someone formed their beliefs" is because the words that followed are "when evaluating the veracity of a claim," i.e. whether or not it is accurate. This is entirely independent of the person's reasons for making the claim.
This appears to in one stroke admit qualification:
"Thus we can see how there are a few things that we need to keep in mind when we address a claim, much as you have said above. However, the truth of the claim, given that you understand the meaning and you are evaluating it at a particular time, does not depend on the belief structure."
And in the next revoke it:
"The reason I said "we shouldn't really care how someone formed their beliefs" is because the words that followed are "when evaluating the veracity of a claim," i.e. whether or not it is accurate. This is entirely independent of the person's reasons for making the claim."
The truthful content of a claim is not independent of the utterances which comprise it, such than an understanding of those utterances is a condition of finding intelligible that claim and thus the candidature of that claim for truth/falsity.
Let me distill this and see if you follow:
We need to know what a claim is actually claiming - that can depend on context.
Given that you do know what a claim is claiming, its veracity does not depend on context, nor the belief structure of the person behind the claim.
I understand exactly what you're saying, but the qualification is divergent from your initial statement, from which this discussion arose, and to which you returned in the second paragraph cited above:
"we shouldn't really care how someone formed their beliefs when evaluating the veracity of a claim"
A condition of evaluating the veracity of an utterance is to register the utterance as intelligible, for which the aforementioned considerations to context are necessary, i.e. 'how someone formed their beliefs'.
If it is divergent, then this
is what I meant. To provide an example, (which can quite often help in these situations):
I claim that the earth is approximately round.
You don't need to know how I came to that conclusion in order to evaluate my claim.
Had I claimed something a bit more complex, maybe related to the society that I currently live in, then you would probably need to know something about my society in order to see if my claim was correct. But you actually wouldn't need to know how I came to the conclusion - you just need to know what I'm talking about.
I feel like this is circular: you state your claim, I state my rebuttal, you concede in qualification, and then you return to your original claim.
I need to know how you came to that conclusion, which is slightly ambiguous here, in the sense that I can't understand the claim independently of the linguistic practice in terms of which your intended meaning is given.
In the case of basic and well-worn facts about the natural world, I think I understand their utterance - although I could be unaware of a particular convention or idiom - because I am already very aware of the linguistic practices which endow them which intersubjective force (if I was a peasant in the Holy Roman Empire, I would doubtlessly have no idea what you were attempting to convey or do).
Alright, since you could not verify the Earth being round without knowing my belief structure...
2+2 = 4
You don't know my belief structure. Is it true?
I'm not asking you if you know that off the top of your head, I'm asking if you could go out and check to see if it's actually true!
That's what I mean by evaluating a claim - can you verify it? I'm sorry, but it's asinine to say that you cannot verify it because you don't know how I came to the conclusion. You seem to be arguing something about sharing my language as maintaining your point. I'm past that. If you understand the claim, you can test it.
I don't really understand what your problem is; to evaluate a claim, you have to find it intelligible, for which you have to know contingent things about the empirical practice of the relevant language-game - which, yes, is pretty much equivalent to the ordinary language statement 'if you understand the claim, you can test it'.
It would seem to me that these claims aren't consistent. I agree with the first claim, not with the second. It's true that experts' claims are objectively and directly verifiable, but lots of the time checking that direct evidence is not an optimal use of our time. Instead we're better off deferring to experts (which we actually also do, as you say, on a massive scale).
I wrote a very long post on a related theme - "genetic arguments" - some time ago, by the way.
Well according to the betting interpretation of degrees of belief, this just means that you would, if rational, be willing to accept bets that are based on the claim in question having a 50 % chance of being true (but not bets based on the claim that it has, say, a 51 % chance of being true). But sure, sometimes it can seem a bit contrived to assign a definite probability to claims you know little about.
I don't agree with that. We use others' statements as a source of evidence on a massive scale (i.e. we defer to them. Indeed, experiments show that we do this automatically. But if these statements express beliefs that were produced by unreliable processes - e.g. bias - then that's clearly not a good strategy. Hence we should care very much of whether someone is biased when evaluating the veracity of many claims, for that reason.
Also, as I said, if we find out that someone is biased, then we have little reason to use that person as a source of knowledge.
What I want to stress is the need for cognitive economy. We don't have time to check the direct evidence for different claims lots of the time (as you yourself admit above) and therefore have to use assessments of others' reliability. Knowledge about bias is a vital (but not the only) ingredient in our assessments of reliability, and are hence extremely useful.
I'm making a separate reply for the betting thing, only to try to keep the two conversations clean/simple.
Let's muddle through it: If I have a box containing an unknown (to you) number of gumballs and I claim that there are an odd number of gumballs, you would actually be quite reasonable in assigning a 50% chance to my claim being true.
If I claim that the gumballs in the box are blue, would you say there is a 50% chance of my claim being true?
What if I claimed that I ate pizza last night?
You might have a certain level of confidence in my accuracy and my reliability as a person to not lie to you; and, if someone was taking bets, you would probably bet on how likely I am to tell the truth, rather than assuming there was a 50% chance that I ate pizza last night.
If you you then notice that my friend, who was with me last night, claims that I in fact ate pasta, then you have to weigh their reliability against mine, and more importantly now you have to start looking for reasons that we came to different conclusions about the same dinner. And finally, you have to weigh the effort it takes to vet our claims against how much you really care what I ate last night.
So, assuming you are rational, would you bet 50/50 that I ate pizza? Or would you just say "I don't know" and refuse to bet in the first place?
This is a bit of a side-track. For the Bayesian interpretation of probability, it's important to be able to assign a prior probability to any event (since otherwise you can't calculate the posterior probability, given some piece of evidence that makes the event more or less probable). They do this using, e.g. the much contested principle of indifference. Some people object to this, and argue along your lines that it's just silly to ascribe probabilities to events we know nothing about. Indeed, the frequentists define an event's probability as the limit of its relative frequency in a large number of trials. Hence, to them, we can't ascribe a probability to a one-off event at all.
Hence there is a huge discussion on this already and I don't think that it's meaningful for us to address it here. Anyway, you do have a point that one should be a bit cautious ascribing definite probabilities to events we know very little about. An alternative can be to say that the probability is somewhere in the interval from x to y, where x and y are some real numbers betwen 0 and 1.
I agree that it is largely off-topic and don't feel like discussing it further here - I would like to point out that the principle of indifference specifies that your list of possibilities must be mutually exclusive and exhaustive. In practice, when dealing with multifaceted things such as claims about the effects of changing the minimum wage, an exhaustive list of possible outcomes would result in an assignment of an arbitrarily small probability according to the principle of indifference. The end effect is that it's a meaningless assignment and you may as well ignore it.
I think we are in agreement but my second statement didn't have the caveats it should have; I doubt you would disagree with the first half, that reality is objective. You disagreed with the second half, that claims should be evaluated based on evidence -- not because it's a false statement, but rather that, in practice, we cannot reasonably be expected to do this for every claim we encounter. I agree. The unstated caveat is that we should trust the experts until there is a reason to think that their claims are poorly founded, i.e. they have demonstrated bias in their work or there is a lack of consensus among experts in a similar field.
Hold on now, you did read my bullets right? When we should care is:
Notice that I actually did say suspicion of bias is an exception to the "not caring" statement. In other words, unless we have a reason to suspect a bias, (and/or the second bullet) then we probably won't care. There can be other ways of bad conclusions being drawn; the reason I mention bias is because it is systematic. If we see a trend of a particular person systematically coming to poor conclusions, whatever their reason, then our confidence in their input would fall. On the other hand, experts are human and can make mistakes as well - we should not dismiss someone for being wrong once but for being systematically wrong and unwilling to fix the problem. If we really care about high confidence in something, for instance in the cases where the truth of the claim is important to a lot of people and we want to avoid being mislead if there are a few biased opinions, we seek the consensus.
Now, can we get the consensus all of the time? Unfortunately not. Not even most of the time. So what's our next line of defense? Well, one of them is journalistic integrity; frankly I don't even want to go there, but if done properly there are people whose job it is to sort through these very things - but really let's not go there for now. The last line of defense is yourself and the actual work of checking on things yourself.
If a claim is important enough for you to really care whether or not it's accurate, then you have to be willing to do a little bit of digging yourself. Now I realize that the entire point of this post was to avoid just that thing and to have computers do it automagically; but really, if it is important enough for you to check on it yourself, rather than just trusting your regular sources of information, then would you be willing not to check just because a program said that this guy was unbiased?
That might be a bit of an unfair characterization of what you're discussing, but there is a distinction to be made between using online behavior to measure/understand the general population's belief structure and to check for bias in expert opinions.
I think the idea of understanding the population's belief structures would still be extremely useful in it's own right though, per my second bullet in the exceptions to the "don't care" statement - particularly if someone wants to change a lot of people's minds about something. If you have a campaign (be it political or social), then understanding how people have structured their beliefs would give you a road map for how best to go about changing them in the way you want. To some extent, this is how it's already been done historically, but it was not done via raw data analysis.
I feel that this discussion is getting a bit too multifarious, which no doubt has to do with the very abstract nature of my post. I'm not very happy with it. I should probably have started with more comprehensive and clear examples than an abstract and general discussion like this. Anyway, I do intend to give more examples of reverse-engineering-of-belief-structures-examples in the future. Hopefully that'll make it clearer what I'm trying to do. Here's one example of reverse engineering-reasoning I've already given.
I agree that lots of the time we should "do a bit of digging ourselves"; i.e. look at the direct evidence for P rather than on whether those telling us P or not-P are reliable or not. But I also claim that in many cases deference is extremely cost-efficient and useful. You seem to agree with this - good.
Sure. But reverse engineering reasoning can also be used to infer expert bias (as shown in this post).
Yes. People already perform this kind of reverse engineering reasoning, as I said (cf my reference to Marx). What I want to do is to do it more systematically and efficiently.