Ethical Realism

January 18, 2012

Beliefs are Innocent Until Proven Guilty

Many people equate “justified” with “justification”—they think beliefs are justified if and only if we give a good justification for them. A sign of this attitude is found in statements such as, “We should only believe something if we can observe it’s true.” I will explain that not all our beliefs require justifications to be justified because (a) we have justified beliefs that we can’t give justifications for, (b) such an assumption is self-defeating, and (c) such an assumption would lead to an infinite regress or vicious circularity.

“Justified” and “Justification”

First, let’s consider the difference between “justified” and “justification.”

Justified – What does it mean for a belief to be justified? It means it’s adequately justified or “rationally permissible.” It’s not wrong to have a belief as long as it’s justified. The belief that “1+1=2” is justified because we know with a high degree of certainty that it’s true. Sometimes two competing beliefs are also justified, such as when scientists debated over which version of string theory is best.

Justification – What does it mean to give a justification for a belief? It means given an argument for the belief that gives us a reason to agree with it. Some arguments are better than others and sometimes we end up giving fallacious (unreasonable) arguments. However, a sufficiently good argument does give us a sufficiently good reason to agree with a belief.

A justification for a belief looks something like “A because B.” For example, I believe I have two hands because I experience that I have two hands.

No one thinks “justified” beliefs are justified merely because we can give a justification for it. Instead, justified beliefs are thought to be adequately justified because of an adequately good argument.

I agree that good justifications can give us justified beliefs. However, I don’t think all justified beliefs require arguments. We can be rationally justified to hold certain beliefs even if we can’t give good arguments for them.

Why some justified beliefs don’t require a justification

We should agree that beliefs are innocent until proven guilty. A belief is rationally permissible unless we have a sufficient reason to reject it rather than the other way around. We don’t need a justification to have a justified belief as long as there’s no reason for us to reject the belief. I reject the that view that all rational beliefs require justifications.

We can realize that all beliefs are innocent until proven guilty for at least three reasons: One, we have justified beliefs that we are unable to give justifications for. Two, it’s self-defeating to assume all justified beliefs require justifications. Three, it’s futile to assume that all justified beliefs require justifications.

Argument 1: We have justified beliefs that we are unable to give justifications for.

Consider the following two counterexamples: One, we know that “1+1=2.” Two, we know that the future will in many respects be predictable based on the past. It is possible that neither of these statements can be adequately argued for, and certainly many people are unable to prove that we should believe them through argumentation. Even so, we all know that these beliefs are justified. We know many facts about mathematics that we can’t prove should be accepted through argumentation, and we know that bread will probably be food rather than poison in the future through past experiences. Predicting the future based on past experiences is known as “inductive reasoning” but there is a “problem of induction” that shows why we can’t know that the future will be like the past in any respect without appealing to induction. We can argue that induction is true because the assumption that it was true worked out well or us in the past, but that’s like saying, “We know induction is true because it’s true.” We can’t appeal to induction to justify the fact that we know something about the future based on the past because it would require us to use circular reasoning.

My solution is that we know induction is justified by merely assuming it’s justified. We don’t need an argument for it.

Argument 2: It’s self-defeating to assume all justified beliefs require justifications.

If all justified beliefs require adequate justifications, then we can give a good justification for the belief that “all justified beliefs require adequate justifications.” However, it’s not clear that we can give an adequate justification for that belief. Anyone who rejects such a belief could consistently assert that not all beliefs require an adequate justification, but it’s not clear that anyone who believes all beliefs require an adequate justification can consistently have such a belief because it’s not clear how such a belief can be argued for.

Do we know that all justified beliefs require adequate justifications because it did so in the past? In that case we are appealing to induction, but I already discussed why we can’t give a good argument that induction is a justified.

Argument 3: It’s futile to assume that all justified beliefs require justifications.

Assume that all justified beliefs require justifications. In that case your beliefs are justified if and only if we can give adequately good arguments for them. Every argument has premises and a conclusion. Every premise is something that we must believe in order to accept the conclusion. So, every belief requires us to accept another belief (the premise) and those beliefs must also be justified. Therefore, the premises we use to justify our beliefs must either justify each other in a circle or we must have infinite arguments to justify our beliefs. However, both of these options are futile. Circular arguments are not informative and it is not possible for us to give infinite arguments to justify our beliefs.

We can show our two options as the following:

  1. Belief A is justified because belief B is justified, and belief B is justified because belief C is justified, and belief C is justified because belief A is justified.
  2. Belief A is justified because belief B is justified, and belief B is justified because belief C is justified, and belief C is justified because belief D is justified, etc. (This never ends.)

Once we see why we have reason to reject the view that “all justified beliefs require justifications” we have a reason to accept the negation—that “not all justified beliefs require justifications.”

Conclusion

Justifying a belief can show a belief to be justified, but not all justified beliefs require good justifications. Beliefs are innocent until proven guilty. That does not imply that it is impossible for us to justify the fact that certain justified beliefs are justified, but the burden of proof is not always on the person who has the belief.

Nonetheless, debates typically require whoever makes an assertion to give a justification for the assertion—and for good reason. If we want other people to agree with us, then they might have no reason to do so without a good argument. There’s a big difference between rationally holding a belief and thinking others should agree with us (or even take us seriously).

Read part 2 — continue reading.

Update (1/23/2012): I deleted a sentence that mentioned verificationism.

Related:

Advertisements

21 Comments »

  1. I agree in part if not entirely with your arguments sans this,
    “We can realize that all beliefs are innocent until proven guilty for at least three reasons: One, we have justified beliefs that we are unable to give justifications for. Two, it’s self-defeating to assume all justified beliefs require justifications. Three, it’s futile to assume that all justified beliefs require justifications.” and I would ask you if the unjustified belief or the one lacking justification ever be anything other then opinion or belief, but never a fact.

    Comment by David — January 18, 2012 @ 8:32 am | Reply

    • I don’t understand your question. Unjustified beliefs are typically taken to be irrational/beliefs we are rationally required to reject. A fact is the reality that makes statements true. Facts exist whether or not we believe anything.

      Comment by JW Gray — January 18, 2012 @ 8:36 am | Reply

  2. Fact not meaning truth, since truth is so subjective. It would seem you have lowered the substantiation bar quite a bit. Leaving enough room under it for someone to levy an unsubstantiated belief and deem it a fact if believed true for no other reason than they think so. There would seem no burden of proof.

    Comment by David — January 18, 2012 @ 8:39 am | Reply

    • Truth is not subjective. The classic Aristotelian view is that a statement is true if it corresponds to a fact. There has to be an appropriate relation between a statement and reality for it to be true.

      You have the burden of proof to convince us of any belief you think is rationally required or irrational. Our beliefs are innocent until you can show them to be inconsistent with rationality.

      Many beliefs are better than alternatives. If you can show your belief to be better than the alternative, then you have made a good use of argumentation. I never denied the importance of argumentation or evidence.

      We think we know lots of things that we might not know how to give arguments for, and so on, as I argued above. If you think my arguments fail, then you should say something about how my arguments fail.

      Comment by JW Gray — January 18, 2012 @ 9:33 am | Reply

    • I think you are concerned that I would allow irrational beliefs to be included in the permissible category. Astrology and ghosts should be rejected. I should not say that people are “justified” to believe in astrology or ghosts. I agree. I think these beliefs are inferior to alternative beliefs (that they are false). We can show that to be the case using argumentation. In particular, these beliefs violate Occam’s Razor. They are overly ambitious claims about reality that are probably false. There is reason to reject them (people probably just made them up, they are far-fetched, etc.), and there is no reason to accept them.

      Comment by JW Gray — January 18, 2012 @ 9:49 am | Reply

  3. There’s a slight mistake in your characterisation of verificationism. It’s very important to remember that the stipulation is verifiable in principle so it’s nowhere near a requirement that you give proof for every belief. The claim is that it’s a requirement of the meaning of a belief that there is some set of observations that could in principle show it to be the case. This would mean that meaning necessarily implies at least some sort of normative constraint which is what your “innocent until proven guilty” clause denies.

    Comment by That Guy Montag — January 22, 2012 @ 3:01 pm | Reply

    • I don’t know that the verification theory of meaning is the same thing as “verificationism” but you are making an important point. I can remove the bit on verificationism since it is a more complex issue and need not be discussed.

      Yes, verificationism (and the verification theory of meaning) require that we can verify everything, but it seems likely that we can’t verify verificationism.

      Comment by JW Gray — January 23, 2012 @ 9:41 am | Reply

      • Verification in the Logical Positivist sense certainly is a theory about meaning hence the old “metaphysics is nonsense” pun. It’s also a theory about meaning if you take the Wittgenstinian private language route, just with more of a pragmatic “meaning as use” sort of emphasis. Given that your argument as a whole is that it’s a practical necessity that we take something as given if we’re going to get our epistemology off the ground it’s really the private language line you need to keep an eye on.

        So let’s just formalise that. Let’s say we’re in a position where we get to just coin a term, say “fibble” for the particular shade of red of the table in front of me. Down the way I find myself seeing a door and I decide that it’s the same shade of red so I declare that “this door is fibble”. There are three possible answers about what fibble means here. The one you’re suggesting is along the lines of “I get to define the term to mean whatever I want” and only when I in some sense try to convince people around me do I need to start justifying the meaning of my terms. The Positivist reply is something along the lines of well that’s all fine, but you’ve not really said anything by saying this door is fibble, it’s just meaningless noise. Meaning for the Positivist would be if you in some sense intended fibble to mean “the same shade of red as I perceive the table in front of me to be” and that this in principle means you could test this meaning by say taking a colour swatch and comparing the shade of the door and the shade of table. The third reply, the Wittgenstinian verificationist challenge, is to both of these ways of defining fibble because the question would be well how do I even know that I’m using my own coinage the same way at all instances. My experience can’t do it because I would never be in a position to guarantee my own experience is consistent. The only recourse would then be to appeal to what other people will let me get away with saying by accepting when I use fibble in certain contexts and by correcting me in others.

        Put like this I’m not sure that you can easily escape the questions around Verificationism because the Wittgenstinian interpretation challenges the idea that you can ever be in a privileged position even about your own coinages and without that, it’s not even clear you’re having anything recognisable as a belief without some kind of normative constraint, let alone whether it’s justified.

        Comment by That Guy Montag — January 23, 2012 @ 3:03 pm

  4. That Guy Montag,

    The positivist challenge assumes various things about epistemology that can’t be verified. It’s self-defeating.

    Wittgenstein is not a positivist, but he also likely has assumptions in mind that are problematic. According to you he would argue that “I can’t know if I use a word properly without evidence.” I would simply reply that I either have a justified true belief that I use the word properly or not. If the belief is justified and true, that’s good enough. What Wittgenstein seems to be denying me is “certainty.” I never said I can be certain about anything. You might say there is some luck involved in knowledge because we lack certainty.

    On the other hand Wittgenstein assumes that other people allowing me to use words a certain way is “evidence” that I am using it right. However, how does he know it’s evidence? Through induction? I already explained we can’t have evidence that induction itself is reliable.

    Comment by JW Gray — January 23, 2012 @ 9:04 pm | Reply

  5. JW

    I don’t think certainty or induction really comes into it. There’s plenty of scope for a verification principle to appeal to theories of evidence that aren’t inductive. The real question is what you’re willing to commit yourself to. Both the positivists and Wittgenstein think that a normative theory of meaning is far less of a commitment than the idea that meaning is just obvious. I’m inclined to agree. The real question I guess is do you have an argument for why we should think that the meaning of our beliefs is just obvious?

    Comment by That Guy Montag — February 4, 2012 @ 4:34 pm | Reply

    • I never said I think meaning is obvious. Perhaps something like foundationalism is an alternative, but I don’t see how induction can be self-evident. And I do think we need to know that induction is reliable. If induction is not meaningful, then we are in trouble. I have already given arguments against verificationism and I don’t think you’ve dealt with them.

      My main arguments are against a view of epistemology and not a view of meaning. Perhaps I made a mistake somewhere because I haven’t studied theories of meanings in quite some time.

      Comment by JW Gray — February 4, 2012 @ 7:25 pm | Reply

      • I’m sorry if I wasn’t as clear as I could be so let’s try again. The goal here is to say that we’re never able to let beliefs off the hook because it’s a necessary feature of them meaning anything at all that they are held to account by either the world or other people. Induction isn’t really important because all I need here is the ability to keep track of my commitments, the commitments of others and a general principle of non-contradiction. There is no kind of additive inductive rule needed, In effect you start with epistemic contraints, and your meaning arises out of it. The alternative you’re offering works the other way: you’re treating meaning as transparent and normative epistemic constraint as being imposed on meaning after the fact.

        Now to get an idea of why I don’t your alternative works just think about a conversation. I don’t know about you, but more often than not when I get into a conversation with someone, particularly about something technical, it’s very rare that we’re able to understand each other right from the start. Instead there’ll be a bit of give and take as we feel around the edges of what we’re both trying to say and then, if we’re lucky, we’ll figure each other out. The challenge then for the idea that beliefs are innocent until proven guilty is this question about how we go about making it clear to other people what it is we believe. If we’re unwilling to accept the constraints of trying to coordinate your belief with other people or with the world then it’s unclear if we could even get across the fact that we actually believe anything at all as opposed to say randomly uttering words without meaning. Take it a step further and Wittgenstein’s Private Language Argument can be read as questioning whether it would even be clear to someone what it is they themselves believe until they’ve gone through the process of coordination.

        Comment by That Guy Montag — February 5, 2012 @ 7:10 pm

  6. That Guy Montag,

    I still don’t understand your argument. It sounds like you’re talking about something practical and it might be that the fact (if it is one) that beliefs are innocent until proven guilty is often not practical. I admit that it isn’t always practical during a debate. But even in a debate people need to share assumptions or it won’t get anywhere.

    Let’s say that the meaning of words has normative constraints. I agree that beliefs do have a normative constraint in the sense that we should reject them (or at least find them to be a lot less justified) when there’s clearly a superior alternative belief available. It is also possible that communication (the meaning of words) has normative constraints that beliefs don’t have in general.

    Even so, I have to wonder if or how my arguments apply to what you are suggesting. How can you track your commitments or the commitments of others without induction?

    Edit:

    Finally, there is the question, “What verifies the verificationist theory of meaning?” If nothing verifies it, then it’s meaningless. (Or replace this argument with whatever normative theory of meaning you have.)

    Comment by JW Gray — February 5, 2012 @ 10:11 pm | Reply

    • JW

      On tracking commitments, I think the problem is that you’re slightly overinterpreting what the problem of induction is. The problem of induction is not a general problem that applies to all evidence. Induction really needs to be understood as the very specific claim that we derive a general rules from finite sets of specific instances *by the application of an inductive rule*. More modern theories of evidence in Philosophy of Science essentially switch things around and assume the general rule, then modify its probilities through appeals to evidence without appealng to anything like an inductive rule. Bayesianism is the most obvious example of this kind of a principle.

      I still think questions like this are a bit of a sideshow though. The question you’re posing is whether we can get away with believing something without being committed to justifying it. The tactic I’m following here is to say that it’s a prerequisite of meaning that it’s publically accountable and so it is in principle always justified to challenge a belief. You’re right that the most obvious way this could get cashed out is verificationism, roughly asking how we would know if it were true, but we could just as easily ask “well what do you mean?” which leaves someone in exactly the same problem of needing justification. This is precisely what my conversation example is supposed to show. The really big argument from Wittgenstein in his Private Language Argument takes this a step further because he points out we can even meaningfully challenge ourselves by asking “well what do I mean” and he seems to believe we can’t give a serious answer that doesn’t appeal to a kind of common meaning.

      As I see it there are two answers that we can give to these challenges as a whole. A very unsatisfying answer would say we’re only interested in what fully fledged language users believe. We get to help ourself to meaning by saying all the hard work of developing what we mean has been done already in our childhood, and there’s no need to go back over that ground. The reason this is unsatisfying is again precisely because in conversation we’re often caught up in the process of explaining what we mean so the idea of what counts as a competent language user seems pretty fuzzy. The other is reply is Jerry Fodor’s but here we’re talking quite literally the David Lewis of counter-intuitive conclusions: meaning is real and hard wired into the brain.

      Comment by That Guy Montag — February 6, 2012 @ 3:31 pm | Reply

      • On tracking commitments, I think the problem is that you’re slightly overinterpreting what the problem of induction is. The problem of induction is not a general problem that applies to all evidence. Induction really needs to be understood as the very specific claim that we derive a general rules from finite sets of specific instances *by the application of an inductive rule*. More modern theories of evidence in Philosophy of Science essentially switch things around and assume the general rule, then modify its probilities through appeals to evidence without appealng to anything like an inductive rule. Bayesianism is the most obvious example of this kind of a principle.

        That sounds compatible with the idea that “beliefs are innocent until proven guilty” but I don’t know that I agree with it anyway. I am quite familiar with Karl Popper’s philosophy of science and I don’t entirely agree with it. At the same time it was Karl Popper’s philosophy of science that greatly shaped my thinking including the idea that beliefs are innocent until proven guilty.

        Can you give an example of where science did not use induction, which is representative of the way science is often done?

        Imagine that someone assumes gravity exists, but then some dropped object doesn’t fall as expected. I would not take that as a falsification of the theory of gravity because it worked so well so far and still seems like a better theory than the alternative. It seems obvious that induction would be involved. I would think it more likely that something went wrong in the experiment than that gravity doesn’t exist.

        I still think questions like this are a bit of a sideshow though. The question you’re posing is whether we can get away with believing something without being committed to justifying it. The tactic I’m following here is to say that it’s a prerequisite of meaning that it’s publically accountable and so it is in principle always justified to challenge a belief. You’re right that the most obvious way this could get cashed out is verificationism, roughly asking how we would know if it were true, but we could just as easily ask “well what do you mean?” which leaves someone in exactly the same problem of needing justification. This is precisely what my conversation example is supposed to show. The really big argument from Wittgenstein in his Private Language Argument takes this a step further because he points out we can even meaningfully challenge ourselves by asking “well what do I mean” and he seems to believe we can’t give a serious answer that doesn’t appeal to a kind of common meaning.

        Questions do not imply a rational requirement of justification. You can question induction, whether other minds exist, and whether an external world exists; but I don’t think such questions imply any requirement for people to know how to defend these beliefs. Are people irrational if they can’t defend them? I think not.

        And my main point is that many beliefs need no pre-existing argument in mind — and the perfect example of such beliefs are those that even the experts don’t agree on. If we have an option of two beliefs and it’s not clear that one belief is superior to another, then I don’t think people have to know how to justify the one they hold.

        Imagine that someone believes that induction in totally unreliable. If I can’t defend the reliability of induction, then perhaps I would have no reason to think this person is being irrational. However, if I wanted to persuade her to agree with me (or prove her belief is inferior), then I’d certainly need to justify the reliability of induction. Perhaps some people can do that even though most people can’t. In that case one belief might be superior to sufficiently knowledgeable people, but other people might lack the expertise to have the superior belief.

        As I see it there are two answers that we can give to these challenges as a whole. A very unsatisfying answer would say we’re only interested in what fully fledged language users believe. We get to help ourself to meaning by saying all the hard work of developing what we mean has been done already in our childhood, and there’s no need to go back over that ground. The reason this is unsatisfying is again precisely because in conversation we’re often caught up in the process of explaining what we mean so the idea of what counts as a competent language user seems pretty fuzzy. The other is reply is Jerry Fodor’s but here we’re talking quite literally the David Lewis of counter-intuitive conclusions: meaning is real and hard wired into the brain.

        I don’t see how that supports your position.

        Comment by JW Gray — February 6, 2012 @ 10:09 pm

  7. I’m not sure how fruitful it would be to go through the arguments for the kind of theory ladenness views of evidence that are contrasted with induction but there are roughly two arguments. The first is the argument that there are in fact an infinite number of different rules that can be inferred from any particular event. The rule that billiard balls move if hit by other moving billiard balls is perfectly consistent with the rule billiard balls move if hit by other moving billiard balls except on Tuesdays. This is a trivial example, but in theory you could build an infinite number of these kinds of perverse rules and induction would be unable to distinguish between them. The argument is then that the fact that we do draw inferences means that we privilege certain kinds of rules over others and that privileging has to reflect prior theory. This particular principle is even more stark if you buy into a Bayesian theory of evidence which builds theory into the prior probabilities you bring to the table. The second argument is just to point to the fact that theoretical work is often necessary in order for science to progress. Most of the hard work that Gelileo did for instance to help overthrow the Aristotelean physics wasn’t actually experimental, it was building thought experiments for example playing with the concept of the motion of ideal bodies in ideal space. As I understand it he never even did the experiment of dropping an object from the mast of a moving ship even though that was one of his strongest arguments against Aristotelianism. The Galilean Revolution was a revolution of consciousness raising, not of experiment, of induction.

    But you’re right, none of this does anything the decide the question that seems to be at issue between us. As I read it, the idea that beliefs can be innocent until proven guilty can seem to leave a very large gap open for relativism which is the real target of my argument. My argument is that a pre-requisite of relativism will be a kind of transparency of meaning because if your meanings needs to be justified, either by verification or appeals to convention or maybe even use, then you can’t claim that your beliefs can be relative. To be fair though, it looks like you’re not arguing for relativism, but maybe it would help if you pointed out how you think your “beliefs are innocent until proven guilty” claim differs from relativism?

    Comment by That Guy Montag — February 22, 2012 @ 11:04 am | Reply

    • That Guy Montag,

      I agree that some beliefs are better than others. Some are more justified or rational than others. If one belief is clearly better than another, then it might no longer be “innocent.” Perhaps it would then be “unjustified” and “irrational.” That could depend (in part) on the person actually understanding why one belief is better than another. What is “rationally permissible” is based on our knowledge. Even so, one belief can be proven false or inferior or unjustified by “the experts.” They would then have the better belief and we might need to know what the experts think to have better beliefs.

      I did not fully explain when a belief is “proven guilty” (inferior or unjustified) here. That is open to debate.

      You can tell me how you think all that would relate to relativism. What form of relativism are you concerned about?

      Comment by JW Gray — February 22, 2012 @ 4:36 pm | Reply

  8. […] Beliefs are Innocent Until Proven Guilty […]

    Pingback by What is the Burden of Proof? « Ethical Realism — May 6, 2012 @ 3:53 am | Reply

  9. “it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence.”

    http://www.infidels.org/library/historical/w_k_clifford/ethics_of_belief.html

    Comment by Luis Tovar (@Luis_Tovar) — May 14, 2012 @ 11:15 pm | Reply

    • What about all the arguments I presented? Merely stating the opposite of my conclusion doesn’t make your claim true. Can you explain what exactly the argument is on the infidels website?

      Comment by JW Gray — May 15, 2012 @ 12:59 am | Reply

    • It is possible that sufficient evidence is sometimes no evidence at all.

      Edit: In the conclusion he says, “We may believe what goes beyond our experience, only when it is inferred from that experience by the assumption that what we do not know is like what we know.”

      Notice the word “assumption” here. Can he prove that what we do not know is like what we know? Of course not. He is referring to something like induction and realizes that induction can’t be rationally proven to be reliable.

      Comment by JW Gray — May 15, 2012 @ 1:05 am | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: