December 20, 2009
One reason it bothers us so to concede the fundamental uncertainty of our epistemology is the questions it raises about how we mary deal with dissenters. One thing the uncertainty ought to do is ground a commitment to some version of libertarianism. If I cannot be entirely sure about what is for my own good, then I should think more than twice about making other people -- or asking the government to make them -- do anything for their own good.
There is probably no discoverable algorithm for picking even a minimal set of epistemological axioms, certainly not more than a handful that every one would agree to. The actual reality of the world we perceive is probably one such axiom. That world includes other people, and the basic similarity of their minds to ours is another candidate axiom. Related to the first is the general reliability of our sensory impressions. Then there is memory and our sense of time. We assume the existence of something we call the past, and we assume that what we remember happening in it did actually happen, for the most part. We assume further that this thing we call time will continue to go on into something we call the future. We can't prove any of these things, but we feel compelled in some way to believe all of it, and no one has ever suggested a good reason for us to believe anything different.
We continually validate these assumptions by acting as if they were so and observing the consequences. We expect certain consequences to transpire if they are true, and we expect other consequences not to transpire if they are not true. In general, the expected consequences are what do transpire, and this reinforces our justification in assuming that the axioms are true. We also observe certain consequences happening to people who act as if these assumptions were not true. Such observations imply that those people were mistaken to act as if the axioms were false. This is the sort of thing Hume referred to as the universal experience of mankind.
None of this would make any sense if there were not something right about inductive reasoning, or at least something that has been right so far. Strict deductive logic informs us unequivocally that we can infer nothing about the future from the past, but it tells us just as unequivocally that no reason exists for doubting that some things, if they have never changed yet, are not now about to change.
The proposition "There exist some characteristics of nature that are constant" is either true or false. We may put on hold the question how we identify those particular characteristics. We are not strictly obliged to affirm either that contants exist or that they do not exist. We can say we just do not know. Logic permits us to have no opinion either way. But life compels us to act either as if there were or as if there were not some constants. There is no way to act undecided. The behavior of indecision is indistinguishable from the behavior of denial.
It seems like another axiom that we should act as if there are constants, since the alternative offers no basis for any decision at all.
October 4, 2009
According to Armstrong and just about every other progressive theist in world, science can tell us what the world is like but cannot say a word about how it ought to be. Scientific rationalism (logos) can give us facts; for values, we need to go somewhere else -- a place called mythos.
It was a failure to observe this distinction that got Christianity into trouble, says Armstrong. When the church started peddling its teachings as a set of facts that all right-thinking people were obliged to believe, then it lost its moral compass along with its credibility, she thinks.
Obviously, a religion with no credibility will have no moral authority, either. Somehow, though, no matter how many skeptics are laughing at the church, anybody who wanted to burn witches, torture heretics, own slaves, terrorize abortion providers, or anything of that sort has managed to find the church well able to give him all the moral authority he might have thought he needed.
Armstrong would have us think that it's not, as Steven Weinberg opined, religion that makes good people do bad things. Rather, she suggests, it is religion trying to be rational that makes good people do bad things. But the suggestion comes without much justification. We're told that the application of reason to religion seems to ruin whatever is good about religion. Why this is so, we never learn from this book.
But if reason and religion make a bad mix, just maybe it's because religion ruins reason. Armstrong passes on the canard about scientific rationalism guiding the likes of Stalin, Mao, and Pol Pot, ostensibly proving that logos doesn't need any help from religion to run up its own body count. Leaving aside the occasional claim that communism was turned into a religion without a god, it appears that these political movements did in fact have a mythos. They had a story that their followers told themselves. From that story, they learned what the world was like and what it should become like; they learned how to live and what to tell other people about how they too should live.
*Karen Armstrong, The Case for God (New York: Knopf, 2009).
October 2, 2009
George wins a lottery against odds of 1 in a billion. Martha wins three consecutive lotteries, each with odds of 1 in a thousand. (Assume that all the lotteries are conducted so as to guarantee a winner with every drawing, and that every contestant buys only one ticket.) Schlesinger observes that Martha's three-lottery sweep ought to surprise us, but not George's winning of one lottery, even though the odds are identical. This is because somebody had to win the single lottery, but there was no guarantee anybody would win all the little lotteries.
Just how is it relevant that in one case there was a guaranteed winner and in the other case there was not? Surely, George himself ought to be no less surprised at his good fortune than Martha should have been at hers. At the individual level, she was no luckier than he was.
Schlesinger's point is that we are justified in suspecting that the small lotteries were rigged to ensure, or at least make it very probable, that Martha would win all three, whereas we have no reason to suspect that George benefited from anything but sheer good luck. Our intuition says his point is well taken. Let's see if we can check that intuition.
In the one case we are not all surprised that somebody won. We are surprised only that it happened to be George. We knew antecedently that one person would win. We just did not know who it would be. In the other case, we knew antecedently that it was nearly certain nobody would win all three lotteries, and so we are very surprised that one person did win all three. If we apportion our surprise according to the odds, then we are equally surprised by "George won the big lottery" and "Somebody won all three little lotteries.'' Why is it that one surprise arouses our suspicion and the other does not? There is a clue in the degree of specificity in the subject of each statement. Let's equalize the specificity and see what happens to our surprise level.
Big lottery: Somebody won -- no surprise; George won -- big surprise.
Little lotteries (all three): Somebody won -- big surprise; Martha won -- big surprise.
Before going any further, we need to get clear on what we mean by "surprise." A surprise is an event contrary to expectations. And on what do we base our expectations? We base them on our assumptions about how the lottery is conducted. If we assume in either case that the lottery was fixed -- either by the winner or by someone acting in their behalf -- then we are not at all surprised by the outcome. We we surprised only because an assumption of honest conduct led us to expect a different outcome. Assuming an honest lottery leads us to expect that neither George nor Martha will win anything.
So then, what do we expect instead?
In George's case, we expect someone else to win. OK, but who else? Well, any of the other 999,999,999 players. Very well. Suppose that Tom, Dick, and Harry were among those others. Would we have been less surprised if Tom had won instead of George? No. If Dick had won? No. Harry? No.
We are surprised that George won because we assumed an honest lottery, but does our surprise justify our questioning that assumption? No, not so long as we would have been equally surprised by any other particular outcome, equally specific. It was slightly more probable that one of the three others -- Tom or Dick or Harry -- would have won, but the disjunction is not equally specific to "George won." Individually, each had the same chance George did.
In Martha's case, what we expected instead of her winning all three lotteries was not that somebody else would but that nobody would. It was certain that three persons, x, y, and z, would win, but it was practically certain that they would not all three be the same person.
Statistically, there was only one alternative to George's winning the big lottery, and that was for exactly one other person to win it. That person, no matter who it had been, had no better chance to win than George did. Thus, if George's winning is sufficient reason for us to question whether the lottery was conducted honestly, then ho would anybody else's winning. But, then we're saying we would be suspicions no matter what the outcome had been. In that case, though, we contradict our claim that we expected someone other than George to win. If we're going to be incredulous no matter who wins, then we must have expected nobody to win.
There were several possible alternatives to Martha's winning all the three lotteries. One person other than she could have won the first lottery or the second or the third. Two people other than she could have won the first and second, or first and third, or second and third. Either of those outcomes was more probable than Martha's winning all three.
We can put it another way. There were at most 2,999 other people who could win have prevented Martha's winning streak, whereas there were 999,999,999 other people who could have prevented George's single win. Whoever won the big lottery instead of George had to beat the same odds he had to beat, whereas whoever blocked Martha's streak needed to beat much lower odds -- just 1 in a thousand.
*George N. Schlesinger, New Perspectives on Old-time Religion (Oxford: Clarendon Press, 1988), pp. 124-38, 141-4, 147-8.
September 23, 2009
It is the ultimate ugly question: Why should I believe that? We don't want to hear it, and it is nearly always considered rude to ask it.
This is partly because we have conflated intellectual rights with civil rights. People say the have a right to believe whatever they believe, as if that were all the justification anyone should want. Of course it is a fact that no one can be legally compelled to change their opinion about anything, but "There is no law against it" is not always a good excuse for doing something.
With occasional and highly interesting exceptions, the law is clear as to whether anyone has a right to engage in some particular behavior, and there is a broad consensus that the law has authority to decide what the citizens' rights are. With regard to epistemic rights, nothing is clear, partly because there is no consensus about who or whether anyone can or should decide what those rights are.
September 18, 2009
One of the more popular skeptical mantras is "Extraordinary claims require extraordinary evidence" (hereafter ECREE). It is commonly attributed to Carl Sagan, and he did indeed popularize it, but he did not coin it. Whatever its origin, though, it is often criticized as an epistemological sham, serving no more purpose than to provide skeptics an excuse for dismissing any evidence at all for any claim to which they are undisposed to give credence. It certainly could be so used, and we can even stipulate that it is so used on some occasions. But no principle or theory is invalidated merely by misuse or misapplication.
The problem of assessing ECREE's validity hinges on the meaning of "extraordinary." What makes any claim extraordinary, and what constitutes extraordinary evidence?
There is a trivial sense in which everyone agrees with ECREE. Would we not all agree that, whatever else "extraordinary" might mean, when it comes to evidence, testimony is about as ordinary as anything gets? If so, then we must note: For every one of us, there are some propositions that we will not believe solely on the basis of uncorroborated testimony, no matter who provides the testimony. We may differ as to which propositions we should place in that category, but the category is not empty for anybody.
For all of us then, some claims do require more or better evidence than other claims do. It simply is not the case, for anyone, that whatever constitutes sufficient evidence for some propositions must be sufficient for all propositions.
But then, how do we, or how should we, decide which claims need better evidence, and how do we decide just what makes certain kinds of evidence better than other kinds?
There is no answer to either question that satisfies all epistemologists. We lay philosophers are on our own here. Regarding ECREE in particular, any evaluation must depend on a definition of "extraordinary," and I have see no non-question-begging definition. Its defenders tend to end up saying in effect that if ordinary evidence fails to convince them that a particular claim is true, then it is an extraordinary claim.
Trying as best I can not to beg too many questions myself, I observe that any given person will call a claim extraordinary whenever he or she regards it -- for any reason at all -- as extremely unlikely to be true. Now, assuming that this person is rational, they will have some reason for having reached this judgment. Maybe we agree it is a good reason, maybe we disagree. It does not matter what we think. It's their reason, and it makes them think that the claim at issue is very unlikely to be true, and therefore they cannot believe it just because somebody says it is so. In such a case, extraordinary evidence will be simply whatever it takes to convince them that the reason they had for making that judgment is wrong.
We can get a little more specific. For some of us, a claim will look unlikely to be true if, were it true, it would contradict some set of beliefs that we antecedently consider very well justified. The claimant is thus telling us that all those beliefs, notwithstanding our justification for believing them, are wrong. And, nobody is obliged to believe such a thing just because somebody says so. Any evidence offered in support of such a claim has to be of such a nature as to provide us with good reason to think that those antecedent beliefs were mistaken, that our justification just was not as good as we thought it was.
This too, I would suggest, is an epistemic position that everybody takes. It is contrary to our human nature to suddenly decide, for no reason but somebody's say-so, that a large number of our fundamental beliefs are in error. Nor do I regard this as a defect in our cognitive nature. Most of the time, it is a good thing if we find it difficult to change our minds about matters of great import.
September 3, 2009
One reason it bothers us so to concede the fundamental uncertainty of our epistemology is the questions it raises about how we mary deal with dissenters. One thing the uncertainty ought to do is ground a commitment to some version of libertarianism. If I cannot be entirely sure about what is for my own good, then I should think more than twice about making other people -- or asking the government to make them -- do anything for their own good.
There is probably no discoverable algorithm for picking even a minimal set of epistemological axioms, certainly not more than a handful that every one would agree to. The actual reality of the world we perceive is probably one such axiom. That world includes other people, and the basic similarity of their minds to ours is another candidate axiom. Related to the first is the general reliability of our sensory impressions. Then there is memory and our sense of time. We assume the existence of something we call the past, and we assume that what we remember happening in it did actually happen, for the most part. We assume further that this thing we call time will continue to go on into something we call the future. We can't prove any of these things, but we feel compelled in some way to believe all of it, and no one has ever suggested a good reason for us to believe anything different.
We continually validate these assumptions by acting as if they were so and observing the consequences. We expect certain consequences to transpire if they are true, and we expect other consequences not to transpire if they are not true. In general, the expected consequences are what do transpire, and this reinforces our justification in assuming that the axioms are true. We also observe certain consequences happening to people who act as if these assumptions were not true. Such observations imply that those people were mistaken to act as if the axioms were false. This is the sort of thing Hume referred to as the universal experience of mankind.
None of this would make any sense if there were not something right about inductive reasoning, or at least something that has been right so far. Strict deductive logic informs us unequivocally that we can infer nothing about the future from the past, but it tells us just as unequivocally that no reason exists for doubting that some things, if they have never changed yet, are not now about to change.
The proposition "There exist some characteristics of nature that are constant" is either true or false. We may put on hold the question how we identify those particular characteristics. We are not strictly obliged to affirm either that contants exist or that they do not exist. We can say we just do not know. Logic permits us to have no opinion either way. But life compels us to act either as if there were or as if there were not some constants. There is no way to act undecided. The of indecision is indistinguishable from the behaviour of denial.
It seems like another axiom that we should act as if there are constants, since the alternative offers no basis for any decision at all.
August 27, 2009
Probably most of what we believe, skeptic and believer alike, we believe on the basis of testimony. If we learned it in school, we believe it on the testimony of our teachers. If we read it in a book or other document, we believe it on the testimony of the author. (Or not; most of us actually do not believe everything we read.)
We believe so much just because somebody says it is so. But none of us believes all testimony. One reason is that we cannot, because there is always contradictory testimony. If one person says X happened and another says X did not happen, or one says Y exists and another says Y does not exist, we know they cannot both be right.
There is much about which no contrary testimony exists. Propositions of this nature, if almost everyone is aware of them, are often called common knowledge. For much else, though, from some people we hear one thing and from others we hear the contrary. In many cases, we are inclined to believe one testimony rather than the other. This is often more or less automatic. Just knowing that some say X and others say not-X, we don't go through any analysis. We suppose either that X is the truth or that not-X is the truth, and that's the end of it. We all do this, all the time.
We have so far said nothing about the nature of any particular testimony. The only thing about any proposition P that makes it testimony is that somebody affirms P, either orally or in writing. And in most cases, if we believe it with little or not thought, then we are presupposing two things. One is that the testifier actually believes what he says. The other is that he has good reason to believe it. Without those two presuppositions, we have no rational justification for supposing that if so-and-so says P, then P must be true.
We may suppose as a generality that these assumptions are warranted, but exceptions to both are numerous, especially the assumption of good reason. Unfortunately, there is no algorithm for determining when we should reject one presupposition or the other, certainly none that is even close to foolproof. Any rule we follow will lead as inevitably to believe some false testimony and disbelieve some true testimony. The very attempt to formulate rules presupposes a prior knowledge of what kinds of testimony tend to be reliable and what kind tend to be unreliable. Thus, if anyone asks What kind, of testimony should I trust? the unavoidable answer is It depends on whom you ask.
Partly for this reason, we and our detractors seem to each other to be guilty of inconsistency. It looks to you as if I reject testimony if it supports your position when I would that testimony in any other case. Likewise, it looks to me as if you believe testimony that you would reject in any other case, provided only that it supports your position. Undeniably, we're all guilty at least once in a while of inconsistency if not outright hypocrisy. That doesn't make it right, but it should make it forgivable.
But if we're inconsistent in applying some principle, it could be because the principle the principle is wrong and we are right. When every rule seems to have an exception, that could be telling us something about human diversity. There are constants of human nature, but for a lot of X's, "Anyone in situation X always tells the truth" is not one of them.
Now some comments on the two presuppositions.
Does the testifier believe it himself? The answer is not always straightforward. Sometimes it depends on what you mean by believe. More than once I have seen a writer introduce or conclude an anecdote with the observation, ''If it isn't true, it ought to be." This seems to be a concise way of saying, "This story might be true or it might not be true, but in either case, (a) it is believable and (b) it contains a Very Important Lesson and so it must be told."
What if the writer is not so committed to literal truth? What if he thinks that, since the story is credible and not known to be false, and since it would benefit people to think it was true, then it might just as well be true. Such a person would be convinced that any disclaimer was an unnecessary and nitpicky distraction, perhaps even counterproductive to the goal of making the world a better place. T0 such a way of thinking, if a story is not provably false and should be true, then it is morally equivalent to being actually true. For these people, the distinction between possibly true and actually true is just for pedants to worry about. There are causes to be served and crusades to be won, and they believe everything they say in the sense that, in their opinion, it is true, but their construal of "true" might differ a bit from yours and mine.
As for having good reason to believe, we can assume that everyone's reasons are good enough in their own judgment. What we should want to know is whether we ourselves would judge those reasons to be good enough. The problem arises when someone asserts P without giving any reasons. We must then ask whether we are justified in supposing that the testifier would not have asserted P without having what we would agree is a good reason.
The answer has to depend on at least two considerations. Let us call the testifier Sam. First, we have to know something about Sam besides his name. An unknown source can not be known t0 be either reliable or unreliable. An unknown source is of unknown reliability. Possibly we will have other reasons to believe the testimony, but if we know nothing about Sam, then we cannot argue, ''It's probably true because Sam said it."
We must also consider how consistent Sam's assertion is with everything else that we thought we knew before we encountered his testimony. Every one of us has a worldview and an epistemic right to use it when evaluating testimony, any testimony. If what Sam tells us implies that our world view is fundamentally in error, then we are under no obligation to believe his testimony for no reason other than
his say-so. We need corroborating evidence that we can evalulate independently -- meaning that we can assess it without giving any consideration to Sam's credibility.
What if lots of other people believe Sam? Then they have a worldview that Sam's testimony is consistent with, and so they are within their epistemic rights to think that if Sam says it, then it must be true. But their rights don't impose any epistemic obligations on the rest of us. If we have a different worldview, then we have every bit as much right to apply it to Sam's testimony as they do to apply their worldview to Sam's testimony.
If it happens to be the case that Sam's testimony is true in fact, then of course we who disbelieve him are making a mistake. And if his testimony happens to be false, then those who believe him are making a mistake. So be it. We all makes mistakes of these sorts all the time. It comes with being human.
What I just said is not to excuse pure obstinacy. My point is solely that nobody has to change his worldview just on someone's say-so. That doesn't mean we're entitled to act like we're infallible. It means that when somebody tells us, "Your worldview is totally messed up," it is always appropriate for us to respond, "OK, you say so. Can you give us any other reason to think so?" And if no more reasons are forthcoming, then it is appropriate to end the debate with that. If, on the other hand, Sam or one of his supporters does produce additional reasons, then we may be obliged to undertake a serious and thorough evaluation of those reasons to see how cogent they are.
August 6, 2009
Reality is the set of all things that exist, and existence implies reality. A thing is real if and only if it has consequences, and a consequence is an observable state of the universe. That is to say, an entity E is a real entity if and only if a universe in which E exists is observably different from a universe in which E does not exist. Such a difference must be implied by any definition of E. If no such difference can be identified, then the existence of E is meaningless, considering that a universe in which it exists is indistinguishable from a universe in which it does not exist.