You are here

Fact and value

A correspondent criticized my statement in Science, Rationality and the Good to the effect that “ ‘rationality’ is itself an evaluative term, and to say a proposition is rational is to say that it should be believed—an evaluative statement”:

In science, a proposition worthy of belief needs to comport with reality. A proposition comports with reality when it can be used repeatedly to make predictions which come true. What this means is that a proposition can be shown to be valid if it can be used to make successful predictions. Predicting the future is very useful and that is why science is so powerful. For example, if I give penicillin to 100 persons with a life-threatening, wild-type gram-positive infection, at least 90 of them will recover whereas if I do not, fewer than 30 will.

My response:

A problem with what you say is that “comporting with reality” is the thing that requires evaluation to determine.

To take your example: no two instances of penicillin or infection are identical. Nor do we know everything about the particular things we are dealing with. The drug might just look like penicillin, or there might be something very slightly different about this batch that unfortunately causes patients with this particular infection to drop dead, or we might have become confused and the previous trials really involved amoxicillin, or the infection might have been misdiagnosed—the tests didn’t work properly, or the results got mixed up—or it might be some new bacterial strain that the test can’t distinguish and thrives on penicillin, or we might be preoccupied or insane or whatever and unable to think straight so we’ve completely misconstrued what’s going on.

You might tell me that in a particular case all those things are silly and can be dismissed, and you might be right, but to make that determination necessarily involves evaluation. Reality isn’t what we start with in science, it’s what we arrive at—or believe we arrive at—after investigating and making judgments that can’t be completely formalized and so necessarily involve subjective evaluation. That doesn’t show that the judgments are arbitrary or that we can’t arrive at truth. What it does show is that our knowledge of truth depends on our ability to judge correctly what is better and what is worse. Knowledge always has a personal element.

Up to now we’ve been discussing a case dealing with a very simple statement like “if you do A then B will follow.” When we’re dealing with more general theories the element of evaluation becomes greater. For example, the theory might be wrong that it’s bacteria that cause the disease and killing the bacteria is what cures it. Arguments like that have been advanced by serious scientists in connection with AIDS. The way they are dealt with is by looking at opposing arguments and deciding which ones are stronger. “Stronger” is, of course, an evaluative term—we’re not talking about formalizable mathematical proof—and not all people who appear rational and competent are convinced. So of necessity what’s done, when a decision has to be made, is that the consensus evaluation overrides the minority evaluation.

All this might seem like quibbling, but it seems to me the basic point is important. It’s still routine today, certainly in practice, to believe that there are facts, things scientists and experts know about, that are objective so they can be relied on in public decisions. Those facts are to be contrasted with values, which are personal and subjective, so public decisions should as much as possible treat one the same as another. It seems to me that kind of view is behind liberalism, and should be fought at every point. (I think the contrasting view that since science is value laden it’s as arbitrary as liberals believe substantive values are is mostly an academic game although one that may in the end have serious consequences.)



Philosophically speaking: No scientific fact should ever be “believed.” On the other hand, a great many scientific facts are as reliable as a statement like “the sky is blue.” Which is a scientific enough statement itself. So no one should believe that the sky is blue. But there is so little doubt to be had over such statements that it would be an abuse of English to insist the verb not be used for such things.

Not so philosophically: The practical objections against scientific fact are far more important than any abstract arguments about the nature of truth. False consensus is a problem. So are the limitations of human reason and institutions when dealing with complex systems. The facts of chemistry are for more “fact” than the facts of medicine. Still, the only way to deal with all that is with more science. I hardly see any realm of reasoned investigation that could improve itself by claiming to be above science.

The scientific institutions of journals and universities are not all there is to science, though. Where they fail, as they often do in the social sciences, for example, it is far more scientific to look elsewhere for answers. (It should not be a surprise that the institutional methods that work great for studying stars have problems with studying the human psyche.) Evolution is an example. It is a commonplace that evolution forms biological solutions to problems of survival that engineering could hardly think up. Economists might claim the free market as something similar.

Tradition also forms from a complex process requiring input from countless sources. There is no reason not to think that the truths formed from such a process are not more valuable than the ones that we might form with simple reason. After all, it is when dealing with some of the most basic human questions that our scientific institutions have been weakest. There is good reason to think that tradition may be better at it.

Hmm… A proposition comports with reality when it can be used to make successful predictions?

Well, there are no propositions without people who speak them or write them. There are no ‘predictions which are successful’ unless someone makes an honest measure of the results when events are constructed to test the prediction. One has to be ‘honest’ about what the experiment shows, what the results are. No ‘cold fusion’ episodes here.

IOW, the ability to measure whether a proposition makes a ‘successful prediction’ requires an Observer who possesses morals, a standard of good and bad behavior, and knowledge of it, prior the experiment…

Thus, there are no ‘valueless facts’.

If no scientific fact should ever be “believed” it seems that questions of truth and falsity fall outside science. I can understand why someone might want to say that for the sake of defining some domain in which the standards of investigation are defined as rigorously as possible. After all, a judgement that something is true always goes beyond the evidence and so includes an element of faith.

Still, it seems that the effect of purifying science in that way is to export the difficult questions outside of science. Those questions—truth and belief, for example—still remain and must be dealt with somehow. So it seems that in addition to science there must be some realm of reasoned investigation that claims to be if not above science then at least beyond science because it draws on science as on other things in studying what is true.

I’d add that if science is defined to exclude questions of truth and belief then it seems that the actual practice of science includes a great deal of non-science. For example, scientists no doubt routinely believe that they can read their instruments correctly, that their memories work well enough for their reasonings to make sense, that they can understand what other scientists are saying and so on.

“questions of truth and falsity fall outside science.”

Say rather, they ultimately fall out of the realm of human capability. Individuals can have ultimate faith in something or other, to be sure, (like, perhaps, faith in Vishnu) but there is no way to evaluate whether that ultimate faith corresponds to ultimate truth. Those who have not been granted such faith must settle for being provisional creatures of practical truths rather than philosophical.

In dealing with questions of whether instruments have been read correctly, memories and reason work, or that the communications of language can be trusted, it is not a question of ultimate truth. It is a question of provisional truth. Investigate the things that one is least sure of first. Eventually, there may come a point where a person may have investigated enough fruitless possibilities to begin to question instruments and such. There is no promise against that.

There is no promise, even, that humans are up to the task of finding out even what can be known. That may take some better creation.

But I think that there are few people who would be willing to throw away the truths, that though not ultimate, we are rather sure of. That the earth goes about the sun. Evolution through natural selection. The efficacy of Penicillin. If these facts are, in the end, only almost certain, that seems good enough for most things that they are needed for.

It seems to me that if “truth” means anything at all then we must sometimes be justified in judging something to be true. If that’s so then it doesn’t make sense to say that questions of truth and falsity fall out of the realm of human capability. If they do then what sense does it make to say we are “rather sure” of some truth? Also, if we can’t justifiably make judgments about truth I don’t see how we can make judgments about practical truth.

My own take on this is that we simply can’t get by without the concept of truth and without the conviction that we are sometimes justified and even correct in saying something or other is true. The problem is that once that’s recognized the modern dream of a self-contained and so fully verifiable system of thought that’s sufficient for all human purposes has to be given up.

A ‘truth’ is a quality of a proposition. Propositions which have the quality ‘truth’ are descriptions of things as they actually are.

For the proposition ‘the sky is blue’ to have the quality of truth, the words ‘sky’ and ‘blue’ must be understandable as names of things. The context of the proposition’s utterance could be that of a father alerting his child to a beautiful visual event, and naming the event to the child. It could be that of two scientists who are overly scrupulous about the word ‘is’ thinking that they cannot speak of such a complex atmospheric system with a term which generalizes this system into one of its aspects, that of color.

The former uses words in their generally understood sense to describe a thing as it is; the latter uses the same words to describe the same event, but feels uneasy that the use of the word ‘is’ seems to equate this complexity to one of its aspects, the color that the sky appears to be to the naked eye, but which is much more than color alone. Our scientists speak out of a different subcontext, tradition, vantage point.

If the former must use words in everyday life in the latter’s context of the use of words, how could he ever say, “I’m hungry?” or “I’m cold?” He would be forever getting at the particulars of exactly “how” he is hungry, or “how” he is cold, and doubt whether he should eat a hamburger or put on a sweater.

Nobody says “Its provisionally true I’m hungry” or “Whether its true that I’m cold or not falls out of the realm of human capability.”

Why? Because people all over know that language conveys things as they really are to other human minds. The sense in which these meanings are conveyed are easily understood by folks sensitive to the context in which they are spoken. Most folks confused by the context ask for clarification. And it is clear that a child is observing the truth or that a scientist is observing it.

One crippled by thinking that to admit ‘Truth Exists’ is to admit that Mind (God) exists has to so break his general experiences into smaller and smaller ‘provable by experiment’ particulars that we view him odd to say the least. His scruples lead him to become a practical nominalist…he cannot trust that general terms/definitions have meaning. He becomes lost in the cosmos…lost to his own experience.

And, it is self refuting to deny that Truth Exists, because the denial itself is a proposition offered as a truth.