Sunday, September 07, 2008

Guest Post: Epistemic Confessional

[The following is a guest post from Jack.]

Here is a fun experiment in philosophy. But first, let me define my terms.

Philosophers speak frequently about justified and unjustified beliefs. What does it take to believe justifiedly? That is a matter of controversy. But it is often supposed that another way of saying that a belief is (un)justified is saying that it is (in)appropriate for you to believe. It may be appropriate for you to believe that God exists, even if He doesn't, if you trust your parents and they told you that He does. It might be appropriate for you to believe both that all doctors are older than 30 and that some physicians are younger than 25, if you have reason to believe each and you have yet to realize that doctors just are physicians. It is, of course, appropriate for you to believe that the Sun is round, that chairs are frequently wooden, and that there were dinosaurs on Earth once upon a time. Justified beliefs are supposed to be fairly easy to come by.

Now for the experiment: Have you ever believed something unjustifiedly? If so when, and what?

I have done this experiment a couple of times before, always with philosophers and always with the same results. After a bit of attempted remembering, the subject says that they cannot remember a single time that they believed something unjustifiedly. (Justified beliefs aren't supposed to be quite that easy to come by!)

Of course, you will have changed your mind about something at some point during your lifetime. Perhaps you even regret believing as you once did. That would presumably make it unjustified for you to believe today what you believed before. That doesn't make your believing when you so believed unjustified.

In any case, I wanted to do a little online survey to see if others have the same reaction to the experiment. Comment either way. If you think that you have, at some point in your life, believed something unjustifiedly, please say what you believed and why you think it was unjustified of you then to believe as you did. If not, what does it take to believe unjustifiedly?

- Jack.


  1. I want to resist the notion that justification and appropriateness can be conflated, especially as "appropriateness" is used here, which seems to mean nothing more than "understandable." If I'm raised by KKK members, it might be understandable in some sense to believe that non-white people are innately inferior, but it's hardly justified.

    And I'll admit to having unjustified beliefs at the times in my life when I've flirted with religion.

  2. I think my belief in internalism about thought content was unjustified as was my belief in internalism about epistemic justification. My beliefs about the death penalty (for at one point, against at another) were both unjustified.

    [Fwiw, I don't think justified beliefs are quite as easy to come by as many epistemologists think. Following the evidence you have at hand, for example, will not ensure that your beliefs end up justified.]

  3. I'm embarrassed to say that at one time I thought that ethical vegetarianism was not a position to be taken seriously. That was an unjustified belief, since I knew almost nothing about the arguments at the time.

  4. If you mean beliefs that are later discovered to be unjustified then there are lots. Like Clayton, I once held beliefs about the death penalty that I now regard as unjustified, but only because I've become aware of new facts and arguments.

    If you mean beliefs that are unjustified based on facts and arguments known at the time, I am not sure it is even possible. Isn't belief just the end result of processing facts and arguments?

  5. Paul, Clayton, thanks for the responses!

    Paul— A child is raised in a KKK household. On the basis of his parents’ typically reliable testimony, without any evidence to the contrary, he believes that non-white people are innately inferior. This seems to me a paradigm case of a justified false belief. How is this any different from the child who, without evidence to the contrary, believes that Shanghai is the capital of China on the basis of his parents’ typically reliable testimony?

    Save for religious episodes, do you feel that you have had a great number of unjustified beliefs in your life?

    Clayton— Do you think that some people believe in content internalism justifiedly?

    Let me tell you a true story. I used to be a content internalist, partly because it struck me as true and partly because my advisor was a content externalist. As I read through the arguments both pro and con, I continued to be convinced of content internalism. (I even wrote a paper on it.) Sometimes I would wonder whether I was a content internalist merely because it gave me occasion to disagree with my advisor. But then I would remember (i) the considerations that initially and continually led me to content internalism, and (ii) my belief that my advisor was being uncritical in his externalism, and hence likely incorrect.

    Today, I think content internalism is false. But by no means do I think I was unjustified in believing as I did. I truly thought that at the end of the day the arguments for content internalism were stronger than the arguments for content externalism. And, based on my advisor’s demeanor, I believed that Kripke had lulled the philosophical community into erroneous complacency. In short: it seemed to me that the evidence favored content internalism.

    I tell you this story because I very much wonder what it would take to believe in content internalism unjustifiedly. What more could it take to be justified in believing content internalism than to think that the evidence favors content internalism? Take Richard. He knows all of the arguments for and against, and he is a content internalist of a stripe. Does he believe content internalism unjustifiedly? If he doesn’t, and if you don’t mind me asking, what was particular about your case that made your believing content internalism unjustified?

  6. Leo, Nigel, thanks for the responses!

    Leo— Suppose that ethical vegetarianism actually was not a position to be taken seriously. Do you think that your belief would have been unjustified anyway? I ask because I wonder what you think is doing the work: the falsity of the belief or you failure to consult arguments bearing on the matter.

    I have a story for you too. I believe that suicide is morally permissible. But in truth, I don’t really know the arguments on either side. I’ve been too lazy to read around. Why do I believe it? I guess mostly it just seems to me so, but also some jumbled feeling that exercises in self-direction that don’t interfere with others is morally permissible. If pushed, I am sure that I can come up with arguments defending my belief. But those would all be post facto. Do you think I am unjustified in believing that suicide is morally permissible? If yes, whence the difference in your belief that ethical vegetarianism wasn’t a position to be taken seriously.

    Nigel— *If you mean beliefs that are unjustified based on facts and arguments known at the time, I am not sure it is even possible.* I have a similar suspicion. I actually do think that it is possible for one to be unjustified in believing something. But I think it is rare, and a bit difficult to achieve.

  7. If you are indicating that being unjustified is not showing good judgment towards what you believe in, who is to say that your judgment is ever justified? Who determines what is just and unjust. If two plus two equals four, when it really is five... who's to say it can't be six.

    I believe that it is what is popular that determines if the view of that belief is just or unjust. Is the earth the center of the universe? Is it unjustified to believe this because it is wrong, or because it is unpopular?

  8. CaJoh - your position is unpopular, and thus self-defeating. More seriously, read this paper [pdf] for a clear explanation of why the 'who's to say' question is so badly confused.

    Nigel - it should be possible to "process[] facts and arguments" poorly, which would then yield beliefs that are unjustified at the time.

    Jack - if people tend not to notice or remember when they have reasoned poorly, such a bias could explain your results without needing to posit that unjustified beliefs are actually so rare.

    In support of Paul's KKK example, I think it makes a difference whether one has had experiences that provide further (counter-)evidence that swamps the testimonial evidence from one's parents. For another example, I think a child's testimonial justification for believing in God typically expires in early adolescence.

  9. Assuming a given state of knowledge, is there a unique justified level of credence we should assign to any proposition? If so, then almost all of our beliefs are likely to be unjustified almost all of the time. It seems vanishingly unlikely that we've got the exact right level of credence assigned to every proposition.

    If you mean something more like "is there anything you've assigned, say, greater than 50% credence to that should have been assigned less" then I would still say "all the time", primarily on the basis that I'm sure I'm as prone to confirmation bias as anybody else. Just because we don't realise a particular belief is unjustified, doesn't mean it's not. We'll notice far fewer unjustified beliefs than we actually have.

    More generally:

    1. I'll believe lots of things temporarily that I probably shouldn't, because I happen to have adopted one side of an argument, get wedded to it, and not realise I'm wrong as quickly as I should.

    2. I have a tendency to respond too favourably to arguments because they're clever or counterintuitive, rather than because they're good arguments.

    3. I tend to discount positions because the most common arguments I hear for them are bad, rather than because I've actually bothered to weigh up the best arguments.

    These are only the things I've noticed; the ones I haven't are probably much worse. And I can't believe I'm the only flawed human being out here...

  10. I once held unjustified Penrosian style beliefs. I also held unjustified disbelief in my unusualness.

  11. what immediately came to mind for me are situations where i've done something absent-minded and had a temporary brain fart.
    eg suddenly believing my job interview was at 9, instead of 9 30, and then after about 5 minutes realising 'no, i'm being an idiot, it's at 9 30'! many more examples like this..

    but i think this is a potentially very interesting point you're trying to bring out. in so far as we talk of justification in knowledge, I think gettier showed it is not sufficient: the justification must be tied to the truth of the belief in the right way. not any justification will suffice.

  12. Before I came to university, I used to be a natural rights libertarian. I thought that natural rights libertarianism was the only morally defensible position in political philosophy. That belief was unjustified. For, in order to be justified in holding that X is the only morally defensible position in political philosophy, you have to have a fair knowledge of a fair few alternatives (to natural rights libertarianism).

    Now, I'm not sure how you'd systemise that. I certainly wouldn't want to say something like "unless you can rule out all alternatives to P, you can't be justified in believing that P".

    Maybe it's only easy to generate that particular unjustified belief because of its extreme nature "ie. the ONLY morally justified ..."

  13. Richard— Good point about memory biases, although we might be able to sidestep memory altogether. Take your beliefs now. First, notice that you deem at least the vast majority of your current beliefs to be justified. Second, try to come up with something you could learn later on that would convince you that you presently deem incorrectly, that something you presently deem a justified belief is actually an unjustified belief. I have yet to come up with anything that could convince me so. Perhaps you’ll have more luck.

    Say there is an adult who believe that God exists because his family, his preacher and his friends testify to it. He is aware that there is great suffering in the world, and recognizes this as evidence against the existence of God. But he places more trust in the testimony of his loved ones than in the Argument from Evil, surmising that he is simply not sufficiently sophisticated to understand why God permits such suffering. Plus, he thinks he might have had a religious experience a couple years back. That sounds to me, more or less, like a typical theist. And that sounds to me like a justified belief. To you no? Do you think that philosophers who offer what is basically a dressed up version of this very theodicy believe that God exists justifiedly?

    Conchis— You mention some interesting biases in belief formation. Do you think one can believe something justifiedly despite the belief coming about in an way that the believer knows is biased? Let me give you a case. I look out and see a girl, and it seems to me that she is looking at me. I know that people tend to overestimate the amount of attention others pay them. But it does seem to me that she is looking at me. On this basis, I come to believe that she is looking at me. Say that in fact she is not looking at me. Am I unjustified in believing that she is looking at me?

    Max— I am not sure that those brain farts produce unjustified beliefs. You seem to remember the interview being at 9:00. That seems sufficient for justification to me.

    Peter— In some sense, I agree with you. It does seem to me inappropriate to be very confident in some view without ever considering alternative hypothesis. The harder question is what is inappropriate. I am reluctant to call the belief inappropriate. After all, because you haven’t even considered other hypotheses, your evidence strongly suggests (to you) libertarianism. Maybe it is your epistemic strategy that is inappropriate. Maybe it is your refusal to consider opponent views that is inappropriate.

    Say that a child is taught Christianity and comes to believe it on testimonial grounds. The child knows that there are other religions, but has not considered them because the child is rather convinced of Christianity and sees not reason to consider views that must be false if Christianity is true. (Have you considered the view that the world is purple all over?) Does the child unjustifiedly believe the Christian dogma? I think no.

  14. I take it that many of the cases you are soliciting are going to be ones in which your belief is in fact unjustified, but you didn't think or know it at the time to be unjustified. To press the boundaries a bit: A curious (and, it seems to me, not an entirely infrequent) phenomenon with unjustified belief is when we believe something, despite believing (and perhaps knowing) that what we believe is unjustified. Some philosophers seem to think this is psychologically impossible, or that the supposed scenario betrays some sort of deep misunderstanding about either belief or justification. I am doubtful that this is always so.

    For example: I believe that (S): namely, if I spill salt, and then do not take the appropriate precautions (throwing salt over my shoulder), something bad will happen. Now I also believe that this belief is unjustified, and in fact, I would venture to say (even more strangely) that I know it is unjustified. In fact, I can think of no evidence in its favor. It's not, after all, as if I put great stock in the authority of the person who first told me this. Nor do I think that this belief is drawn somehow from received wisdom which I am justified in relying upon. Far from it. I actually have well-founded further beliefs about how this ridiculous superstition arose (scarcity of salt in medieval times, alleged presence of the devil on one's shoulder). Yet this understanding has no effect on my recalcitrant superstitious belief (which it seems nonetheless fair to continue calling a belief). For I would (and in fact do) obsess all day about what bad will happen as the result of my having spilled the salt and not cautiously thrown a pinch over my shoulder, despite thinking all the while that I am unjustified and indeed irrational to do so.

  15. "Do you think one can believe something justifiedly despite the belief coming about in an way that the believer knows is biased?"

    If you advert to the bias and attempt to adjust for it, possibly (though we also tend to undercorrect for biases, even when we are aware of them). But if you don't advert to the bias, or don't attempt to adjust for it (perhaps because you believe you're less biased than others) then I would say you're beliefs are unjustified.

  16. Jack - if my belief is based on certain reasoning, and I later realize that this was actually really bad reasoning -- some fool mistake where I think "Damn, I should have known better!" -- then that would suffice to convince me that the belief was unjustified all along.

    On religion, the point is that experience undermines the grounds for believing others' testimony. The guy should realize his parents were doped the same as he was. And philosophers of all people should know better -- they're certainly unjustified in believing the tenets of any organized religion. (Some weaker deism is a very different matter, but I consider that closer to atheism.)

  17. I (and most others) have an unjustified belief right NOW -- that incest is wrong. There's no rational argument for it whatsoever and yet it persists.

  18. Andrew— A fascinating case! I agree. Let me ask: do you endorse your own believing (S)? Do you feel like it is right of you to believe (S), or is it more that despite yourself you find yourself believing it? How would you defend believing (S)?

    Conchis— That sounds implausible to me. Suppose you come to believe that P in a biased belief forming way, you make no attempt to advert the bias, you have no real evidence against P other than that you came to believe it in a biased way, and P is in fact true. I say you know P. So I say you are justified in believing P.

    Richard— I think you’re on to a similar feeling that I am on to. If I later realize that by my own current standards for belief it is inappropriate for me to believe as I do—if I later realize that I don’t actually endorse believing as I currently do—then that would suffice to convince me that my belief was unjustified too. I think that this is a fairly rare occurrence. Quite typically we do a good job of believing only things that by our own lights we ought to believe, I think, no?

  19. In addition to the sorts of unjustified belief Conchis describes, I tend to form unjustified beliefs to the effect that food in my fridge has gone bad. I'm not sure whether my beliefs that food has gone bad are justified in cases where they're true. I'm also not sure whether my evidence is the same in the cases where the food really has gone bad and the cases where my wonky disgust mechanism is acting up.

    Jack, are you assuming in your 10:24 reply to Conchis that justified belief is a necessary condition for knowledge? You seem to be banking on some sort of reliabilist intuitions, but a reliabilist would deny that all cases of knowledge are cases of justified belief.

  20. Jack-

    I think I would accept your position for cases where correcting for the bias wouldn't make a difference to one's belief. (Say, one's (biased) credence in P is 80%, and correcting for the bias would reduce this to 70%. If we treat belief as assigning greater than 50% credence, then it's seems that one might still be justified in believing P, despite the fact that the belief was biased in its formation - though assigning 80% credence to P would still be unjustified.)

    But in cases where properly correcting for the bias should cause one to stop believing P (say it's actually a 40% bias, or your initial credence was only 55%), your position seems implausible to me. In this case, I would say that one neither is justified in believing P, nor knows P. If it turns out that P is true, then you just got lucky. ;)

    My remaining qualm about accepting your position in the first type of case is that it also seems there that there may be a meta sense in which you "just got lucky" that the bias wasn't strong enough that you should have changed your belief in P. At this point I'm inclined to say that this means that one wouldn't be justified believing that one's belief in P was justified, even though the belief in P might itself be justified.

  21. «Comment either way. If you think that you have, at some point in your life, believed something unjustifiedly, please say what you believed and why you think it was unjustified of you then to believe as you did.»

    In 2003 I unjustifiedly believed there were reasons to preemptively attack Iraq. I was wrong and there weren't weapons of mass destruction in there. Then I unjustifiedly start to believe that Al Qaeda didn't have anything to do with the Iraq question. It turned out to be that Al Qaeda was leading the insurgency and that there wasn't exactly a war, but something different.

    Now I justifiedly or not surely don't know what to think.

  22. Jack, I think maybe we need a notion of temporary justification to track things like the KKK case and the childhood religious indoctrination case -- in each, the evidence available to a person is artificially constrained, and I think it's plausible to say that the child has the justified belief (because the child doesn't really have the capacity to go out and experience additional evidence) but that justification doesn't continue into adulthood -- that adults have an epistemic responsibility to stop basing their beliefs on only the evidence of parental say-so.

    As for non-religious unjustified beliefs, I'm sure I, like everyone else, has had unjustified beliefs arising from wishful thinking. For example, I frequently believe that I'll make it to a destination on time, based on the assumption of minimal traffic delay, even though all my experience is that there is rarely ever minimal traffic delay...

  23. Hmm. I hold the (very unpopular) view that there is insufficient coherent sense to the term 'justification' to make it a useful term for epistemology; people are rarely talking about the same thing. I'm not convinced, for instance, that everyone here is really thinking about the same thing when they type 'justification', for instance.

    I'm interested in this notion of "temporary justification" first suggested by Richard (in his comment about justifications expiring) and then again by Paul. It would seem to run up against a problem (if it is a problem) of indeterminate justification: if, e.g., a child can have a justified belief because she "doesn't really have the capacity to go out and experience additional evidence" then it would appear that it can't be just limited to children (even if adolescents and adults, e.g., have it for different beliefs than children). But also it seems we start getting on shaky ground -- e.g., there are beliefs based only on parental say-so that we don't think adults unreasonable for holding (e.g., beliefs about events that happened in early childhood). On what grounds would one draw the distinctions here?

  24. Most "beliefs about events that happened in early childhood" aren't later swamped by counter-evidence, are they? Unless your parents told you they took you to the moon as a kid... but then that would clearly be unreasonable to keep believing as an adult.

  25. What Richard said, plus a lot of problems get solved if we're willing to talk about epistemic duties and say that there's a duty to seek out additional evidence for certain kinds of propositions -- that is, propositions for which more evidence is readily available. There isn't extra evidence readily available for "funny things Paul said when he was three," whereas there is extra evidence readily available for "evolution is true" and "black people aren't inferior." So we can hold adults accountable to looking beyond their parents for evidence on the latter two domains but not the first.

  26. Religious beliefs, which were being put forward, aren't usually "swamped by counter-evidence", either, so that doesn't really help clarify things. I rather suspect very few beliefs are swamped out, actually, even very wrong ones; and those that are generally become swamped only after we reject them and therefore put ourselves in a position to see all the things that count against them. Most beliefs are reasonably rejected not because they are swamped by counter-evidence but because we find that evidence that can be interpreted a number of ways is most parsimoniously, coherently, and easily (in the psychological sense of 'easily') interpreted in a way that involves rejecting it.

    In any case, the parental say-so comment was in reference to the suggestion by Paul Gowder that adults have an epistemic responsibility to stop basing their beliefs on parental say-so. And in light of this, swamping by counter-evidence doesn't help, because what is in view is the responsibility to gather any (further) evidence on the subject in the first place, so we're still way upstream from any flood of counter-evidence.

  27. 'Religious beliefs, which were being put forward, aren't usually "swamped by counter-evidence", either'

    Well, I think that's precisely what's in dispute between people who disagree about whether religious belief is unjustified.

    "Most beliefs are reasonably rejected not because they are swamped by counter-evidence but because we find that evidence that can be interpreted a number of ways is most parsimoniously, coherently, and easily (in the psychological sense of 'easily') interpreted in a way that involves rejecting it."

    I think parsimony and coherence are very important (they may be what does the epistemic 'swamping'), so I'm not sure we really disagree here either.

  28. Well, I think that's precisely what's in dispute between people who disagree about whether religious belief is unjustified.

    Possibly; as I said, I don't think 'justified' and 'unjustified' is a coherent term, so I generally have no idea what people are disputing when they dispute whether a given type of belief is unjustified.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.