Aus: Fred Dretske: The Pragmatic Dimension of Knowledge
aus Fred Dretske Perception, Knowledge, Belief. Selected Essays. Cambridge 2000. S. 48ff
- Reprinted from Philosophical Studies 40 (1981), 363-378, copyright © 1981 by Kluwer Academic Publishers, with kind permission from Kluwer Academic Publishers
Knowing that something is so, unlike being wealthy or reasonable, is not a matter of degree. Two people can both be wealthy, yet one can be wealthier than the other; both be reasonable, yet one be more reasonable than the other. When talking about people, places, and topics (things rather than facts), it makes sense to say that one person knows something better than another. He knows the city better than we do, knows more Russian history than any of his colleagues, but doesn't know his wife as well as do his friends. But factual knowledge, the knowledge that something is so, does not admit of such comparisons. If we both know that today is Friday, it makes no sense to say that you know this better than I. A rich man can become richer by acquiring more money, and a person's belief (that today is Saturday, for example) can be made more reasonable by the accumulation of additional evidence, but if a person already knows that today is Friday, there is nothing he can acquire that will make him know it better. Additional evidence will not promote him to a loftier form of knowledge — although it may make him more certain of something he already knew. You can boil water beyond its boiling point (e.g., at 300° F) but you are not, thereby, boiling it better. You are simply boiling it at a higher temperature.
In this respect factual knowledge is absolute. It is like being pregnant: an all-or-nothing affair. One person cannot be more pregnant or pregnant better than someone else. Those who view knowledge as a form of justified (true) belief typically acknowledge this fact by speaking not simply of justification, but of full, complete, or adequate justification. Those qualifications on the sort of justification required to know something constitute an admission that knowledge is, whereas justification is not, an absolute idea. For these qualifiers are meant to reflect the fact that there is a certain threshold of justification that must be equaled or exceeded if knowledge is to be obtained, and equaling or exceeding this threshold is, of course, an absolute idea. I can have a better justification than you, but my justification cannot be more adequate (more sufficient, more full) than yours. If my justification is complete in the intended sense, then your justification cannot be more complete.
Philosophers who view knowledge as some form of justified true belief are generally reluctant to talk about this implied threshold of justification. Just how much evidence or justification, one wants to ask, is enough to qualify as an adequate, a full, or a complete justification? If the level or degree of justification is represented by real numbers between 0 and 1 (indicating the conditional probability of that for which one has evidence or justification), any threshold less than 1 seems arbitrary. Why, for example, should a justification of 0.95 be good enough to know something when a justification of 0.94 is not adequate? And if one can know P because one's justification is 0.95 and know Q because one's justification is similarly high, is one excluded from knowing P and Q because the justification for their joint occurrence has (in accordance with the multiplicative rule in probability theory) dropped below 0.95?
Aside, though, from its arbitrariness, any threshold of justification less than 1 seems to be too low. For examples can easily be given in which such thresholds are exceeded without the justification being good enough (by ordinary intuitive standards) for knowledge. For example, if the threshold is set at 0.95, one need only think of a bag with 96 white balls and 4 black balls in it. If someone draws a ball at random from this bag, the justification for believing it to be white exceeds the 0.95 threshold. Yet, it seems clear (to me at least) that such a justification (for believing that a white ball has been drawn) is not good enough. Someone who happened to draw a white ball, and believed he drew a white ball on the basis of this justification, would not know that he drew a white ball.
Examples such as this suggest (although they do not, of course, prove) that the absolute, noncomparative character of knowledge derives from the absoluteness, or conclusiveness, of the justification required to know. If I know that the Russians invaded Afghanistan, you can't know this better than I know it because in order to know it I must already have an optimal, or conclusive, justification (a justification at the level of 1), and you can't do better than that. I have explored this possibility in other essays, and I do not intend to pursue it here. What I want to develop in this essay is a different theme, one that (I hope) helps to illuminate our concept of knowledge by showing how this absolute idea can, despite its absoluteness, remain sensitive to the shifting interests, concerns, and factors influencing its everyday application. In short, I want to explore the way, and the extent to which, this absolute notion exhibits a degree of contextual relativity in its ordinary use.
To do this it will be useful to briefly recapitulate Peter Unger's discussion of absolute concepts. Although he misinterprets its significance, Unger does, I think, locate the important characteristic of this class of concepts. He illustrates the point with the term flat. This, he argues, is an absolute term in the sense that a surface is flat only if it is not at all bumpy or irregular. Any bumps or irregularities, however small and insignificant they may be (from a practical point of view), mean that the surface on which they occur is not really flat. It may be almost flat, or very nearly flat, but (as both these expressions imply) it is not really flat. We do, it seems, compare surfaces with respect to their degree of flatness (e.g., West Texas is flatter than Wisconsin), but Unger argues that this must be understood as a comparison of the degree to which these surfaces approximate flatness. They cannot both be flat and, yet, one be flatter than the other. Hence, if A is flatter than B, then B (perhaps also A) is not really flat. Flatness does not admit of degrees, although a surface's nearness to being flat does, and it is this latter magnitude that we are comparing when we speak of one surface being flatter than another.
Unger concludes from this analysis' that not many things are really flat. For under powerful enough magnification almost any surface will exhibit some irregularities. Hence, contrary to what we commonly say (and, presumably, believe), these surfaces are not really flat. When we describe them as being flat, what we say is literally false. Probably nothing is really flat. So be it. This, according to Unger, is the price we pay for having absolute concepts.
If knowledge is absolute in this way, then there should be similar objections to its widespread application to everyday situations. Powerful magnification (i.e., critical inquiry) should, and with the help of the skeptic has, revealed "bumps" and "irregularities" in our evidential posture with respect to most of the things we say we know. There are always, it seems, possibilities that our evidence is powerless to eliminate, possibilities that until eliminated, block the road to knowledge. For if knowledge, being an absolute concept, requires the elimination of all competing possibilities (possibilities that contrast with what is known), then, clearly we seldom, if ever, satisfy the conditions for applying the concept.
This skeptical conclusion is unpalatable to most philosophers. Unger endorses it. Knowledge, according to him, is an absolute concept that, like flatness, has very little application to our bumpy, irregular world.
I have in one respect already indicated my agreement with Unger. Knowledge is an absolute concept (I disagree with him, however, about the source of this absoluteness; Unger finds it in the certainty required for knowledge; I find it in the justification required for knowledge). Unlike Unger, though, I do not derive skeptical conclusions from this fact. I will happily admit that flat is an absolute concept, and absolute in roughly the way Unger says it is, but I do not think this shows that nothing is really flat. For although nothing can be flat if it has any bumps and irregularities, what counts as a bump or irregularity depends on the type of surface being described. Something is empty (another absolute concept, according to Unger) if it has nothing in it, but this does not mean that an abandoned warehouse is not really empty because it has light bulbs or molecules in it. Light bulbs and molecules do not count as things when determining the emptiness of warehouses. For purposes of determining the emptiness of a warehouse, molecules (dust, light bulbs, etc.) are irrelevant. This isn't to say that, if we changed the way we used warehouses (e.g., if we started using, or trying to use, ware-houses as giant vacuum chambers), they still wouldn't count. It is only 10 say that, given the way they are now used, air molecules (dust ',articles, etc.) don't count.
Similarly, a road can be perfectly flat even though one can feel and see irregularities in its surface, irregularities that, were they to be found on i he surface of, say, a mirror would mean that the mirror's surface was not really flat. Large mice are not large animals and flat roads are not necessarily flat surfaces. The Flat Earth Society is certainly an anachronism, but it is not denying the existence of ant hills and gopher holes.
Absolute concepts depict a situation as being completely devoid of a certain sort of thing: bumps in the case of flatness and objects in the case of emptiness. The fact that there can be nothing of this sort present for the concept to be satisfied is what makes it an absolute concept. It is why if X is empty, Y cannot be emptier. Nonetheless, when it comes to determining what counts as a thing of this sort (a bump or an object), and hence what counts against a correct application of the concept, we find the criteria or standards peculiarly spongy and relative. What counts as a thing for assessing the emptiness of my pocket may not count as a thing for assessing the emptiness of a park, a warehouse, or a football stadium. Such concepts, we might say, are relationally absolute; absolute, yes, but only relative to a certain standard. We might put the point this way: to be empty is to be devoid of all relevant things, thereby exhibiting, simultaneously, the absolute (in the word "all") and relative (in the word "relevant") character of this concept.
If, as I have suggested, knowledge is an absolute concept, we should expect it to exhibit this kind of relationally absolute character. This, indeed, is the possibility I mean to explore in this essay. What I propose to do is to use what I have called relationally absolute concepts as a model for understanding knowledge. In accordance with this approach (and in harmony with an earlier suggestion) I propose to think of knowledge as an evidential state in which all relevant alternatives (to what is known) are eliminated. This makes knowledge an absolute concept, but the restriction to relevant alternatives makes it, like empty and flat, applicable to this epistemically bumpy world we live in.
zurück zu Die Jugendsünde
<root><h level="3" i="1">=== Kontext ===</h>