萨拜因：什么是政治理论？（下）

It is true, of course, that every political theory is a fact, a quite substantial fact that occurs in the gamut of facts that makes up a particular political situation. As such, it had its causes and may, no doubt, have its effects. Moreover, it has its effects whether it be true or false, because in either case it exists in a quite objective sense, as a thing that may affect men's conduct. It is always possible that men be have differently in any given situation, merely because they entertain some theory about their own existence and the situation in which they find themselves. This is a curious involution that occurs in all social theories and that has no precise an alogue in the theories of natural science, unless it be in those cases, recently brought to light by the principle of indeterminism, where the mere fact of observation operates to change the very state of affairs which is under observation. Where this occurs, the natural scientist admits with all modesty that he has reached a limit beyond which he cannot conceive are finement of his theory. And the social scientist must surely, in allintellectual honesty, do the same. In so far as theories figure as facts,standing in causal relations with other facts, and in so far as they appear as the data of human behavior which a theorist must himself count among the dataof the situation that he is studying, they must of course be accepted as all data are, simply as elements of the reality studied. Their effects are in noway correlated with their truth, for even false theories may influence men's conduct. Their causal influence as existing facts is simply irrelevant to their truth or falsity. But in any given time and place one must make up his mind which language he elects to speak. If he accepts a theory as itself a bona fide effort to speak the truth, he must accord it that respect which belongs to such an effort. He must meet it on the plane of logic, must confirm or refute it onthat level, by showing its consistency or inconsistency and its ability orinability to explain the facts. When he begins to discuss its influence, he puts it among existing things in the world of events and objects, and event sare not themselves true or false; they simply occur.

An example will make the meaning clearer. Acritic may deal in two quite different ways with the doctrine that Jeffersonwrote into the Declaration of Independence. It is possible to discuss the validity of those propositions about inalienable and indefeasible naturalrights, to apply a rational criticism to the assertion that all men are createdequal, to analyse their meaning, to show how they agreed with a prevalent conception of scientific method, and to point out where in they fall short ofthe self-evidence that Jefferson attributed to them. But such criticism is possible only so far as the critic is willing to discuss the factual truth or logical consistency of the theory examined. Quite apart from all such questions, however, it is still a fact that Jefferson and his fellow-members ofthe Continental Congress did believe in the theory of natural rights. It isquite possible that they would not have acted as they did had their beliefs been otherwise. It is credible that in so believing they may have been unconsciously the agents of a militant middle- class, intent upon rising to the political power that their economic importance warranted. If such causal influences swayed their action, it is of no consequence whatever whether whatthey believed was true or false. As Bishop Butler said, "Everything iswhat it is, and not another thing," and beliefs may have their effect showever false they are. But surely no critic can apply both criteria at once.He may be concerned to assess the correctness of a doctrine, and if so its consequences are irrelevant; or he may be concerned with its actual effects andinfluences, and then its truth is irrelevant.

Political theories, therefore, live on two planes or play a double role. They are theories, or logical entities belonging to the abstract world of thought, but they are also beliefs, events in people's minds and factors in their conduct. In this latter role they are influential (if they are) not because they are true but because they are believed. On this plane they operate as events, or as actual factors in historical situations,and as such are part of the data which the historian of politics has to deal with. But this historical reality is obviously not what interests those persons who sincerely believe a theory to be true; such persons are not interested in a theory because it exists but because they believe it to be a valid explanation of something else. What the framers of the Declaration of Independence meant to do was "to declare the causes" that impelled them to dissolve the political bands which had connected the Colonies with England, an explanation required by "a decent respect to the opinions of mankind." In this they set down as a major premise the claim of indefeasible natural rights and as a minor premise a long list of aggressions,which they attributed to the King of England and interpreted as evidence of asettled determination to tyrannize. For them these claims were not merely beliefs; they were parts of what purported to be a correct statement of factsand a valid inference from them. A rational criticism, as distinguished from a study of historical causes, would have to take these claims as bona fide, even though it might end with the conclusion that they were utterly fallacious.

To return now to the beginning of this essay, it will be apparent that the questions there raised referred to the rational criticism of political theories, the question whether, or in what sense, they can claim the logical attributes of truth or validity. It will be remembered also that in the description there given of political theories itwas said that they regularly unite two kinds of factors. In the first place,there are elements of a factual and causal nature: the apprehension of a state of affairs actually existing, an estimate of the relative importance ofdifferent factors in this situation, and a weighing of future possibilities. Inthe second place, there are elements of valuation: an estimation of importance,not in the sense of what is likely to happen, but of what ought to happen, the discrimination of a better from a worse way, the conviction that some courses of action are morally obligatory, an expression of choice or preference growing from an attitude of desire, or fear, or confidence toward what the presentholds and what the future may bring forth. The question, then, is whether a theory uniting these two kinds of factors can be rationally adjudged to be trueor false; in short, whether there is any common measure that can extend overand validate the theory as a whole.

Now the only absolutely general standard of rational criticism is the rule that a theory must not contain propositions that are mutually contradictory. A person who thinks about politics is under the same obligation to think consistently as one who thinks about any other subject, and to be convicted of an inconsistency is as damaging to a political theorist as to any other kind of theorist. Moreover, the standard of straight,coherent thinking is applicable both to thought which has facts for its subject-matter and to thought which has values for its subject-matter. Athinker can argue for mutually contradictory obligations as easily as he can attribute mutually incompatible properties to objects, and when he does the first he is as certainly wrong as when he does the second, for the avoidance of contradiction is a general principle that applies to all valid intellectual operations what so ever. Nevertheless, the mere absence of contradiction can not be regarded as equivalent to truth, except perhaps in pure logic and mathematics. For even if a theory were altogether self-consistent, there would still be the question whether what actually happens is the same as what the theory contemplates, and even if a theory of values were entirely coherent,there would still be the question whether the values which it contemplates are really acceptable as ends to be striven for and, if possible, attained. After making every admission possible to the binding- force of logical consistency,one must still agree that it goes only a little way toward validating a theory of any kind, whether in politics or any other subject.

If non-contradiction, though indispensable,is still not a sufficient principle of criticism, is there any other principlethat can bridge the two kinds of propositions-allegations of fact andascriptions of value that occur together in every political theory? Apparently the answer must be, No. In combining these two kinds of factor a political theory puts together propositions for which there is no common logical measure and which all the dictates of clear thinking require to be distinguished. In so far as a political theory depends on the assertion, expressed or implied, that some state of the facts is so and so, the only test applicable to it consistsin inquiring whether the facts really were as alleged or different. In so faras it presumes that one course of events is more likely to occur than another,it can be tested only in the light of the actual probabilities and perhaps inpart by seeing whether the event seems to justify the expectation. In either case the assertion that an event has happened and the assertion that it ought to have happened are simply different and therefore ought not to be confused.And similarly, to say that a future event is probable is quite different from saying that it is desirable, or good, or the reverse. The two kinds of propositions are logically disparate in the sense that any statement containing such a copulative verb as "ought to be" requires the assumption of a standard of value which is never present as such in any purely actual situation or any purely causal sequence of events. When the two kinds of statement occurin conjunction, as they continually do in political theories, the beginning of critical judgment is analysis, the discrimination of the two kinds and the application to each of the tests appropriate to it.

Analysis and discrimination in this matter do not imply the superficial idea that political theories can be made"scientific" by the omission of references to moral and other forms of valuation. This idea usually depends not at all on discrimination of values as one element in a theory, but only on a simple-minded unconsciousness of valuations that have become habitual. It depends upon the kind of intellectual simplicity that Schopenhauer once attributed to an opponent: heimagined, Schopenhauer said, that whatever he had learned before he was fifteen years old was an innate principle of human reason. In truth it is humanly impossible even to describe a political or social situation without at leastimplicit assumptions about the importance of the elements that are to go into the description; the choice is between implicit assumptions and the explicit avowal of what is assumed. Moreover, there is no objection, at least on thescore of logic, to making explicit assumptions about what is desirable; apolicy or an end can be discussed as reasonably as anything else. It isprobably not true even that men disagree more about values than they do about other matters. In any case there is no logical reason why a social philosopher should not postulate any value he chooses, provided only that he avows what he is doing and does not pretend to prove what he is merely taking for granted. What he cannot do logically (or even honestly, if he knows what he is doing) is to pass off his valuations as if they were inescapable facts.

The practical question, of course, remains,whether it is really possible to perform this act of analysis, at least so long as a political theory is still an element in a living situation. Looking back to the past one easily perceives how often men's judgment of facts is swayed by their interests or misled by the intensity of their moral convictions, but in one's own thinking it must be admitted that one does not, and probably cannot,always avoid the same kind of error. The common usages of language conspireto make such confusions. The most ordinary words, like is and must be, have regularly a two fold use, to signify indifferently logical or moral necessity,existence or predication, and the precise meaning must be gathered, if at all,from the context. Thus, to refer again to the Declaration of Independence, when Jefferson declared it to be "self-evident that all men are created equal," he may have thought that the proposition was an alogous to the alleged self-evident propositions that stand on the opening pages of Euclid. In the light of an exact analysis of those propositions, however, no one can imagine that he was merely giving a rule for handling symbols. It is hardly likely that Jefferson thought that all men are as a matter of fact equal;certainly the moral effect of the sentence is spoiled if one takes it to beparallel with some literally true statement about the way men are created, such as, all men are created babies. As everyone knows, Jefferson was really expressing a moral conviction to the effect that, in some matters of vital human importance, it is wrong to deprive men of their freedom of choice. One may accept or reject this assertion, but he cannot intelligently do eitherunless he sees what is really intended. In a sense the inevitability ofconfusion or error is irrelevant, even if it is a fact. No one wholly avoid sinconsistency, but inconsistency is an error just the same. If there is a confusion inherent in the conflation of facts and values, it is still a confusion even if the whole world conspires to do it. Of course, no one doubts that, in this as in other respects, men do think more clearly when they try resolutely to avoid confusion.

It would be altogether unfair, however, to imply that the coalescence of judgments of value with judgments of fact or of logical implication has only the' standing of a frequent, but admitted,popular confusion. On the contrary this coalescence is undertaken systematically in certain philosophies which, together, cover a considerablepart of current philosophical opinion. A representative of one of these views would enter an exception against the statement made above, that propositionsstating facts and propositions ascribing values are logically disparate and would hold that it is possible to include both within a single logicalsyn thesis. Historically this contention goes back to Hegel, who believed that the idea of a self-developing totality in logic could sublate the duality of rationalism and empiricism and refute at once the revolutionary doctrine of natural rights and the conventionalism or positivism implied by Hume's critique of natural rights. This was the purpose which Hegel thought that dialectic could fulfill. By means of dialectic he supposed it possible to show that certain values must emerge in the course of history and, conversely, that thecausal processes of history are regulated by an inherent tendency to realizeand conserve values. The dialectic was at once, therefore, a causal explorationand an immanent ethical criticism. However it may be formulated, the belief that some such dovetailing of value and fact is a soluble problem remains thebest index of Hegel's influence over later philosophy. It continued to characterize the English NeoHegelians, and with all their differences it remains the fundamental claim of the Marxists, whose dialectical materialism is still in essence a claim that causal and moral necessity can be synthetized. Ina milder form the purposes, if not the apparatus, of Hegel's dialectic perpetuated themselves in the pragmatism of Professor Dewey and Professor Mead,already referred to. For from the allegation that meanings can occur only in the fulfillment of purposes and that reflective thought is only a directiveagency in behavior, it appears to follow that logical adequacy must include both factual efficiency and the fulfillment of purpose.

It would be silly to embark upon athumb-nail refutation of Hegelianism, with all its ramifications, at the end of an essay already too long. The purpose has been to outline a problem and to suggest a type of solution but not to offer a refutation of other types of solution that have been attempted. Its intention has been to suggest that hereis one of the systematic differences between philosophical points of view,rooted as such differences are likely to be in diverse theories of knowledge.Descriptively one finds in human thinking about a concrete problem-say theproblem attacked by a political theory-what seems to be a variety of factors answering to a variety of critical standards. There are allegations of fact and cause; there are imputations of value or obligation; there are the consequences, for human behavior, of believing or disbelieving the theory. Now must it be the case that there is some criterion of truth large enough to stretch over all these factors in the problem? If some such criterion is proposed-say "coherence"-does it cover the ground because it affords a really applicable standard for all types of problem, or does it seem to do somerely because it is so vague and ambiguous that no one knows with certainty what it means? Or, on the other hand, is it possible that the drawing of distinctions is both the beginning and the end of wisdom? In short, is it possible that truth is a word with several different meanings and that no one cansay what it means unless he is allowed to discriminate at least what Leibnizcalled truths of reason and truths of fact, and perhaps several other kinds beside? With this reference to Leibniz it will be well to stop. For it suggests the obvious line of criticism, namely, that this paper illustrates a kind of philosophic atavism, the no stalgia for clear and distinct ideas that was more typical of the seventeenth than of the nineteenth century.