Cal: But do you accept that confirmation bias is real, and that if one is wrong because of its effects the one would be irrational?
Because it seems like you're trying to say that confirmation bias isn't real, and that it doesn't affect thinking in a way that would make one irrational. If that's the case, you'd be wrong on both counts.
VR: No, neither. First, what I am questioning isn't the reality of confirmation bias, nor the fact that it can lead to irrationality. Satta M's description seems accurate:
It's more like the Principle of Total Evidence- actively seeking out evidence that disconfirms your hypothesis, not just that which confirms it.
Fine, I get that. People should look for reasons on both sides, and failure to do so leads to mistakes. You could get the right answer while suffering from confirmation bias, but you decrease your chances if you do.
My complaint is not with the idea. But the idea, it seems to me, is best used as a tool for self-evaluation. Insofar as my memory is accurate, I am privy to all of the information that goes into my decision to believe this or that. You are not privy to that information. Consequently, you are not in a position to figure out what I have considered and what I haven't.
It isn't just a matter of what you read, it is also a matter of how charitably you read it. If I trust the authority of someone who tells me "Sure, read this stuff, but make sure you don't buy any of it, because you don't want any of your money going for that guy's royalties, since the stuff he writes sucks so badly," then even if you read stuff from that side you will be reading it not to understand it, but to look for something to pounce on.
It seems to me that, when confirmation bias is used as a tool not for self-analysis, but to explain the disagreement of our opponents, it goes something like this.
"I know I have analyzed this issue correctly, and cannot possibly be wrong. So there must be some explanation for why someone whose intelligence is supposed to be a great or greater than mine has come to the opposite conclusion. I know, it's confirmation bias!" That is, rather than seeking to understand one's opponents position and to figure out whether perhaps there is a parameter of the issue that one has failed to consider, you use the idea of confirmation bias to say "Mistakes were made, but not by me." It just seems to me that the people who yell the loudest about other people's confirmation bias look like people who suffer from it most themselves, and don't realize it. People use the idea of confirmation bias to do exactly what the idea of confirmation bias should teach you NOT to do. It's like the people who listen to sermons some vice in church and start thinking of who other than themselves ought particularly to listen to that message.
I happen to have spent some time in Barnes and Noble the other day, and paged through the Loftus volumes The Christian Delusion and The End of Christianity. Both those volumes include good, interesting, challenging stuff for Christian apologists. The funny thing was that both books contained endorsements by Christians, and even Christian apologists like Randal Rauser and Matt Flanagan. They both said that while, of course, they didn't agree with the conclusions of Loftus et al, they thought that reading the contents of those books would be helpful for anyone trying to think through apologetical issues. What strikes me as odd here is that I don't think John would be caught dead putting an endorsement on some volume by apologetical authors, such as True Reason or Contending with Christianity's Critics, suggesting that the contents of that volume would be a worthwhile challenge for atheists.
43 comments:
Haven't read it all, but I see a major problem right here:
VR: " Insofar as my memory is accurate, I am privy to all of the information that goes into my decision to believe this or that. You are not privy to that information. Consequently, you are not in a position to figure out what I have considered and what I haven't."
If you think you are privy to ALL the information that goes into your decision to believe this or that, then you don't understand psychology, or biases. At all. On a basic, fundamental level.
It's probably that simple; you don't understand modern psychology, and in particular you don't understand biases, and how they influence everyone's (including your) behavior.
I would highly recommend reading Kanneman's Thinking Fast and Slow. There's a lot of other material on the subject, but I think that would be a great primer.
and vice verse
I was simply counting the conscious, reflective arguments I was considering for believing something. Like everyone else, I am only partially conscious of what I am doing.
But without appealing to all sorts of scholarship and reading, can you tell me what bias avoidance looks like. and how it is someone can look at someone else's thinking and charge them with confirmation bias without being unbiased themselves. And, even if you succeeded in doing that, how would you know you were sufficiently free from bias to assess the biases of others.
Since you mention Kahneman, I remember studying an Kahneman-Tversky critique of probabilistic reasoning by people identifying taxicabs in an accident. K and T accused the subjects of failure to properly apply Bayes' theorem, but the Bayesian decision theorist I worked with at the time maintained that their critique was based on a misapplication of Bayesian theory.
VR: "Like everyone else, I am only partially conscious of what I am doing."
Okay. But you can see where that's incompatible with what you said earlier, right?
VR: "Insofar as my memory is accurate, I am privy to all of the information that goes into my decision to believe this or that."
-----
VR: "But without appealing to all sorts of scholarship and reading..."
As a rule I don't like it when people recommend books (instead of arguments), so I sympathize. I just brought up the Kahneman book because it seemed as if you didn't buy into the existence of biases at all, and if that was the case there's too much for a comment to fix.
VR: "...can you tell me what bias avoidance looks like..."
Much of it is reflective, and analytic. Much of it is scientific. In many ways, bias avoidance a) acknowledges that we almost certainly can't entirely avoid our biases (even on reflection and analysis), so b) the best way to avoid biased decisions is to subject our beliefs to testing (which includes intersubjectivity).
VR: "and how it is someone can look at someone else's thinking and charge them with confirmation bias without being unbiased themselves."?
Because observing someone else thinking is different than observing yourself thinking. Although both parties have biases (the thinker and the thinking observer), they bring DIFFERENT biases to a given topic, and because of that difference much information can be teased out.
Here's a way of understanding it. Let's say you and I are both pretty bad golfers. But when you swing, I can see problems in your swing that you can't, and I can tell you about them. And you can watch my swing, and tell me about the problems you see. In that way, even though we're both pretty bad golfers, we can both use each other as observers to improve our golf game.
VR: "Since you mention Kahneman, I remember studying an Kahneman-Tversky critique of probabilistic reasoning by people identifying taxicabs in an accident. K and T accused the subjects of failure to properly apply Bayes' theorem, but the Bayesian decision theorist I worked with at the time maintained that their critique was based on a misapplication of Bayesian theory."
Unless you want to explain what the mistake was, and how it's relevant, I don't know what the point of this is.
Since Bayez was mentioned, I think this is relevant and hilarious.
People on opposite sides of debates about fundamental issues invariably see their opponents as biased. And sometimes, claims of bias are used as a basis for dismissing those on the other side.
A paper on error theories in psychology.
http://www.stat.columbia.edu/~gelman/communication/Gigerenzer1991.pdf
Abstract. Most so-called “errors” in probabilistic reasoning are in fact not violations of probability theory.Examples of such “errors” include overconfi dence bias, conjunction fallacy, and base-rate neglect. Researchers have relied on a very narrow normative view, and have ignored conceptual distinctions—for example, single case versus relative frequency—fundamental to probability theory. By recognizing and using these distinctions, however, we can make apparently stable “errors” disappear, reappear, or even invert. I suggest what a reformed understanding of judgments under uncertainty might look like.
https://erikbuys.files.wordpress.com/2013/12/confirmation-bias.png
Since you mention Kahneman, I remember studying an Kahneman-Tversky critique of probabilistic reasoning by people identifying taxicabs in an accident. K and T accused the subjects of failure to properly apply Bayes' theorem, but the Bayesian decision theorist I worked with at the time maintained that their critique was based on a misapplication of Bayesian theory.
I've studied this very problem extensively and have even delivered presentations on the topic. I would love to hear more about why the Bayesian decision theorist made that claim. It seems to me undeniable that K&T correctly identified the so-called base rate fallacy: that people paid too much attention to the specific information (the evidence to be explained, viz., the witness's testimony about the color of the cab) and too little attention to the general information (the background information, viz., the base rates of blue and green cabs). I wonder if your Bayesian decision theorist was a staunch Bayesian (i.e., someone who rejected frequentism and instead adopted an epistemic interpretation of probability) and built his or her interpretation of probability into the meaning of the terms / parameters of BT?
A paper on error theories in psychology.
As a side note, I really like Gigerenzer's work. He came up with the idea of "natural frequencies" as a way to train people like doctors how to normalize the variables they implicitly use when trying to "naturally" apply Bayes Theorem, without having to actually remember Bayes Theorem.
Patrick Maher is an antifrequentist, who is largely responsible for teaching me the probability background for my two essays on miracles.
Jeffery Jay Lowder said...
A paper on error theories in psychology.
As a side note, I really like Gigerenzer's work. He came up with the idea of "natural frequencies" as a way to train people like doctors how to normalize the variables they implicitly use when trying to "naturally" apply Bayes Theorem, without having to actually remember Bayes Theorem.
Keep working along those lines we are going to be back to good old logic by itself.
I find atheists using that phrase any time o e ha any kind of evidence they can't answer. It's their version of "no one knows the mind of God." A thing ot say when you don't have an answer. It sounds clever seems like an insight until they use it for everything.
VR: "People on opposite sides of debates about fundamental issues invariably see their opponents as biased. And sometimes, claims of bias are used as a basis for dismissing those on the other side."
Yup.
But in any disagreement about fundamental issues, one side is right, and the other is wrong. One side has let their biases influence their belief, and the other side's belief isn't so affected. We are both biased ≠ We are both wrong.
"the other side's belief isn't so affected"
That doesn't follow. Not if people are all biased, anyway. If that's the case, both sides will have biases, and the person on the right side might possibly believe the right thing purely because of bias. "We are both biased" would imply that it's possible that both sides believe the things they do for the wrong reasons, irrespectve of the truth of their beliefs, and thus would have no knowledge -- no justified true belief. As such, if that's the case, even those on the right side would not know that they're right and would not be able to show it; their belief would be accidentally true.
Miguel: "That doesn't follow. Not if people are all biased, anyway."
Many beliefs trigger biases differently depending on the individuals involved.
If the belief were whether or not I, Cal Metzger, am a good person, you would be less affected by the biases that affect the exploration of that belief than I would be. Etc.
And that's my point.
Just because only one of two contradictory positions can be true, it doesn't follow that only one position can be rational to believe. In fact, the true position can be irrational to believe.
VR: "Just because only one of two contradictory positions can be true, it doesn't follow that only one position can be rational to believe. In fact, the true position can be irrational to believe."
Ostensibly, sure.
But we're talking about probabilities here, and aligning beliefs with reality. Let's not throw in the fact of possibility as a way of dismissing the power of probability -- as if we should say that because it's possible that we might go to nuclear war tonight it's rational that we not plan for tomorrow.
And the fact is that our beliefs are more likely to be aligned with reality if we can remove those things, like biases, that we know stand in the way of our apprehending reality.
It seems to me that belief in God can't be a matter of probability,. if it is you can't say God is probable or improbable. I think if Hartshorne is right and God is either necessary or impossible such a sharp dichotomy can't be rendered in terms of chances. If God is impossible there is no chance there's a God, If there is no God it is only because he's impossible. If there is a God there is no chance there might not be.
Then as Tillich says there can't be a God because he's not one of many. I guess we could couch it in terms of human understanding. Can we say "there is a probability might be wrong that God is either necessary or impossible?"
I didn't say biases weren't a bad thing. But what the basis of probability is is a very difficult subject.
http://plato.stanford.edu/entries/epistemology-bayesian/
VR: "I didn't say biases weren't a bad thing. But what the basis of probability is is a very difficult subject. / http://plato.stanford.edu/entries/epistemology-bayesian/ "
I am not following your point. Are you saying there is an epistemic question about who is likely to be biased regarding a kind of belief? Because I don't think that's the case.
My point about probability is that one who takes steps to mitigate those biases that are known to affect relevant beliefs is more likely to align their beliefs to reality than one who does not mitigate those biases. Do you think that this is an epistemic question?
But it is also quite possible that bias-mitigating methods can be constructed in such a way as to produce its own kind of bias. A bias-defense system that is excessively geared toward eliminating false positives may in fact result in a bias in favor of false negatives.
VR: "But it is also quite possible that bias-mitigating methods can be constructed in such a way as to produce its own kind of bias."
What instance do you have in mind?
VR: "A bias-defense system that is excessively geared toward eliminating false positives may in fact result in a bias in favor of false negatives."
A bias-defense system (?) that eliminates positives in a greater proportion than false positives is just another kind of bias.
I honestly have to say that I think you don't really grasp what biases are, and how they're thought of.
I'll give you an example. The outsider test for faith, as conceived by Loftus.
VR: "But it is also quite possible that bias-mitigating methods can be constructed in such a way as to produce its own kind of bias."
Me: "What instance do you have in mind?"
VR: "I'll give you an example. The outsider test for faith, as conceived by Loftus."
That seems like a bad example, then.
The OTF basically challenges the believer to evaluate other religious claims using the same standards as they do when evaluating their own.
Every time I've read a believer trying to explain that their belief passes and the other religions don't, much hilarity ensues.
Discussion concerning the OTF gets really tricky. If I say "Here is a difference between Christianity and some other religion. Christianity has it, the other religion doesn't, so by using this one category, Christianity wins and the other loses," I get told that my criterion is a bad criterion and I shouldn't use it, so the "passing" doesn't count.
Suppose I say that whatever religion has documents who has the greatest support for its scriptures from archaeology is the one I'm going to choose. Christianity is going to win that one. So it's the same standard, different results. But no, that can't be an acceptable criterion.
VR: "Discussion concerning the OTF gets really tricky. If I say "Here is a difference between Christianity and some other religion. Christianity has it, the other religion doesn't, so by using this one category, Christianity wins and the other loses," I get told that my criterion is a bad criterion and I shouldn't use it, so the "passing" doesn't count."
You seem to think that the OTF is a request to create a diagram and weight various differences between the religion and choose the one that scores best. This just shows that you don't understand the OTF -- which really just means basic skepticism, critical thinking, an understanding of history and the world's other religions, past and present, etc.
The OTF is really just the same thing as a very good liberal arts education, although I'd be the first to admit that quite a few people get that and they still struggle as you seem to.
VR: "Suppose I say that whatever religion has documents who has the greatest support for its scriptures from archaeology is the one I'm going to choose. Christianity is going to win that one. So it's the same standard, different results. But no, that can't be an acceptable criterion."
I don't know what you think you're doing there, but as mentioned above, that's not the OTF.
---------
Here's the secret sauce: you know how the little you know about the other world's religions, none of that seems very persuasive to you? That's because all those religious claims also lack what your religion lacks -- good evidence for its religious claims, the kind of evidence that changes minds. The OTF just points out that the lack of persuasive evidence to CHANGE YOUR MIND ABOUT YOUR RELIGION isn't found in the religion you (surprise!) grew up around and became accustomed to believing in. And that's because human beings are complicated (but fairly predictable) things, and we're packed with biases that make the above a very, very common occurrence.
The OTF draws attention to this, but it can't make you understand it. That final step is all up to you.
So, there people don't exist:
https://en.wikipedia.org/wiki/List_of_converts_to_Christianity_from_nontheism
I said that human beings are complicated (but fairly predictable) things.
That's like me saying that humans tend to want to stay alive, and you thinking you disprove this mundane observation by providing a list of famous suicides.
Like I said, I can't make you understand if you don't want to understand.
You say there is no persuasive evidence. I point out that there are plenty of people who at one point were not part of the religion who were persuaded by evidence. And here by evidence I mean some things that convinced them to change their minds.
Admittedly, there is traffic both ways. But if there is nothing that can persuade someone to go from, say, atheist to Christian, then why do so many people go there.
It seems biased to say that all the biases are on one side.
VR: "Admittedly, there is traffic both ways. But if there is nothing that can persuade someone to go from, say, atheist to Christian, then why do so many people go there. / It seems biased to say that all the biases are on one side."
People are persuaded all the time for bad reasons. Look at the success of advertising.
The fact remains that religious claims all lack good evidence for their claims. That some people are persuaded despite this is unremarkable. And that's because some people will always be persuaded without good evidence. And many, many more will believe something because they were raised in it, because it's a fact of psychology that the beliefs we were raised with are harder to shed than they are to keep.
This is all completely uncontroversial.
Religious believers are a) almost entirely those who were raised in their religion, or b) those who were persuaded to join for pragmatic reasons (this group offers me support, benefits, fulfills my psychological needs, etc.), or c) became persuaded despite their not being any good evidence. This is true of every religion. Every. Single. One.
Now you'd like to pretend that people believe as you do because they have good evidence, but you can't provide good evidence. If you could, you'd wouldn't have a religious claim -- the kind that persuades the vast majority of us like knowledge claims do, the ones that have good evidence -- you'd have a fact about reality. You'd have good evidence.
How do we know you'd have good evidence? The stick that we measure good evidence isn't what you seem to think it is -- that some people are persuaded -- no, the standard is that it's objective, reliable, and verifiable. It can be "checked." Now, it is true that everything that is good evidence does persuade almost everyone, but IT'S NOT TRUE that what persuades many people must therefore be objective, reliable, and verifiable. If that were true, then Islam would be true, and Hinduism, and Scientology, etc.
In fact, many things are true despite the fact that most people don't originally believe them. And again, they are true because if we check them, then we see that they are true. (See, for instance, here: http://www.opticalspy.com/opticals/rubiks-cube-optical-illusion1). Do you understand this? Check is good evidence. Not able to check, not good evidence.
I truly get tired of this.
VR: "It seems biased to say that all the biases are on one side."
It would be lying to say that, given a specific instance, one side is not more affected by a bias when they are. For instance, we know that it's harder to adopt a belief that overturns a belief you already hold than if you don't hold a belief that would be overturned. This is basic, basic stuff. And yet you reject it.
I wonder why you can't seem to accept this? Do you think it's because your not prepared to change your belief about god? Nah, that couldn't be it.
I don't think it is harder to overturn a belief than to continue it. But that is not at all irrational. People have to guide their lives by their beliefs, and too many changes undermines reasonable stability.
I cannot check for the existence of electrons, quarks, or strings. I don't think you can either. Are we to conclude that these things are fictional because of this? I don't think so.
The verification principle was thought to be the solution to all problems in philosophy. Only one problem, the principle itself could not be verified, so it had to be dismissed as nonsense itself. If we say "we only have evidence for something if we can check it" then, that statement has to be checkable. And it's not.
There are numerous arguments in support of religious belief, and lots of people have been persuaded by those arguments. I find a few of them very persuasive. For example, if the world were purely material, and everything happened by physical laws, then evidential relations, which do not have no particular location is space or time, would always be causally irrelevant to the formation of any belief. So, in order for evidence to every have anything to do with anyone's belief, theism, or at least panpsychism, would have to be true.
I cannot check for the existence of electrons, quarks, or strings. I don't think you can either. Are we to conclude that these things are fictional because of this? I don't think so.
"I believe in Christianity as I believe that the Sun has risen not only because I see it but because by it I see everything else.” - C. S. Lewis
VR: "I don't think it is harder to overturn a belief than to continue it."
Then you are at odds with modern psychology, and biases like the Anchoring Effect, Confirmation Bias, etc. This position makes you seem arrogant.
VR: "I cannot check for the existence of electrons, quarks, or strings. I don't think you can either. Are we to conclude that these things are fictional because of this? I don't think so."
No, I could check for electrons and quarks. I could use an electron microscope, for starters. I could pursue a career where electrons are the subject, etc. You could do this too. That's what makes electrons and quarks examinable. That if we were to check on them, we'd see that what others know about them is correct.
VR: "The verification principle was thought to be the solution to all problems in philosophy. Only one problem, the principle itself could not be verified, so it had to be dismissed as nonsense itself. If we say "we only have evidence for something if we can check it" then, that statement has to be checkable. And it's not."
This is so lame. You use logic, but can you use logic to verify logic? Etc.
VR: "There are numerous arguments in support of religious belief, and lots of people have been persuaded by those arguments."
Arguments, but no evidence. Hmmmm.
"I believe in Christianity as I believe that the Sun has risen not only because I see it but because by it I see everything else.” - C. S. Lewis
Aah, a deepity.
But still no evidence.
There is a boatload of evidence for theism, and for Christianity. There are facts in existence that are vastly more probable to exist if theism is true than if it is not. In all theistic universes a thinking being exists. In many atheistic universes no thinking beings exists, and it is difficult to see how thinking beings could possibly exist. Therefore, the evident fact that there are thinking beings in our universe is evidence that God exists. It may not be enough evidence for strong atheists, but the fact that it is evidence is indubitable.
CM: No, I could check for electrons and quarks. I could use an electron microscope, for starters. I could pursue a career where electrons are the subject, etc. You could do this too. That's what makes electrons and quarks examinable. That if we were to check on them, we'd see that what others know about them is correct.
VR: You would have to get an education, and the equipment, to do so. I'm guessing you probably won't, and that you will continue to believe that these things exist based on authority, just as Christians believe many things on the authority of the Bible and the Church. I could build a time machine and see if Jesus was resurrected, for that matter. So, the resurrection of Jesus is every bit as checkable, in theory, as quark theory.
If checkability is a requirement, then you have to be able to check the check, and check the check of the check, and check the check of the check of the check, etc. You just started in infinite regress you can't stop.
There a number of historical facts in the founding of Christianity that make sense if Jesus rose from the dead, but don't if Jesus didn't. And these are attested by nonbiblical sources.
When I said I did not believe that it is easier to support a belief than to overturn it, that was a typo.
I also argued that confirmation bias is not necessarily irrational.
What is the evidence for atheism?
VR: "There is a boatload of evidence for theism, and for Christianity."
You can't say that without providing actual evidence. An argument isn't evidence. It's an argument. (I can make an argument for, well, anything. By your thinking, there would be evidence for anything. This makes the word "evidence" practically meaningless.)
VR: "You would have to get an education, and the equipment, to do so. I'm guessing you probably won't, and that you will continue to believe that these things exist based on authority, just as Christians believe many things on the authority of the Bible and the Church."
No. For one, I've seen images taken with electron microscopes. Lots of them. So I don't need anything special to see them. That's because electrons, and the images they can be used to create, are real things.
VR: "I could build a time machine and see if Jesus was resurrected, for that matter. So, the resurrection of Jesus is every bit as checkable, in theory, as quark theory."
Well, time travel appears to be theoretically impossible, kind of like going faster than the speed of light seems impossible. So, by saying you could do something impossible, it seems that you're saying that you're comfortable with no standards about anything. Because apparently, you think that everything is possible, therefore, god!
VR: "If checkability is a requirement, then you have to be able to check the check, and check the check of the check, and check the check of the check of the check, etc. You just started in infinite regress you can't stop."
No.
I can check real evidence. I can't check religious beliefs. That's the difference between things that are real, and religious claims. And that's all that anyone needs to know.
VR: "There a number of historical facts in the founding of Christianity that make sense if Jesus rose from the dead, but don't if Jesus didn't. And these are attested by nonbiblical sources."
No there aren't. You're being silly.
Chris: "What is the evidence for atheism?"
Same as it is for aLeprechaunism. And everything else you don't believe in.
Ok, I see.
Just like aatheism .
Chris: "Ok, I see. / Just like aatheism ."
Pro tip: when you have a double negative, it's simpler (and almost aways preferred) to go with the positive. So, aatheism would be, theism.
--------
But if you wanted to know how we could approach the question of leprechauns from an evidential perspective, we could test for them. We could control for leprechauns, and we'd see that we detect no leprechauns. After awhile, we notice the trend, and (despite some sporadic claims or guesses about leprechauns that never end up being examinable), we gain confidence in the noneness with which leprechauns frequent our world.
You can do the same with religious claims. Those that can be tested (like asking for an appearance, praying for something real, or just observing if a god reliably and verifiably intercedes on behalf of those he's asked to believe in him, and doesn't on those who have rejected him, etc.) reveal, again, the zero-ness of godly intercession. Gods become identical to leprechauns in that if they exist, they are indistinguishable from a world in which they don't. Hence, aleprechaunism, and atheism.
So, you see, you don't even have to rely on the standard of a null hypothesis, and the burden of proof, if you wanted to conduct an evidential approach to leprechauns, or gods. You can actively investigate on your own, and so long as you remain consistent, and apply standards that measure something real.
Cal,
When you talk of leprechauns as a means of illustrating your point, it is hard to take you seriously.
Chris: "When you talk of leprechauns as a means of illustrating your point, it is hard to take you seriously. "
Now you see how it is when I hear people talk about their gods as a means of illustrating their points.
Cheers.
Touche.
Post a Comment