Chapter 66 Chapter 17 Hume

David Hume (1711-76) is one of the most important figures among philosophers, because he developed the empiricist philosophy of Locke and Berkeley to its logical end, by making it self-consistent , making it something unbelievable.In a sense, he represented a dead end: in his direction, no further progress was possible.Contradicting him has been a fashionable pastime among metaphysicians since he wrote his books.For my part, I find none of their rebuttals convincing enough; however, I have to hope to find some system that is less skeptical than Hume's system. Hume's major philosophical work, Treatise of Human Nature, was written while he was living in France between 1734 and 1737.The first two volumes were published in 1739 and the third in 1740.He was young then, not yet thirty; he was unknown, and his conclusions were such that almost all schools of thought would disapprove.He was expecting a violent attack, intending to meet it with a dignified rebuttal.But no one paid any attention to the book; as he himself said, "it was stillborn from the printing press." He went on: "However, I soon recovered from the blow, because of my naturally jovial disposition. ” He devoted himself to the writing of prose, and published the first collection of prose in 1741. In 1744 he attempted unsuccessfully to obtain a professorship at the University of Edinburgh; failing in this, he became tutor to some madman, and then secretary to a general.Emboldened by these pledges, he ventured again into philosophy.He condensed the book by omitting the best parts of the Treatise of Human Nature and most of the grounds for his conclusions, and the result was Inquiryin to Human Under stunding, a book that was by a long time shorter than the A Treatise of Human Nature is much more famous.It was this book that awakened Kant from his "dogmatic sleep"; Kant does not seem to have known the Treatise of Human Nature.

Hume also wrote Dialogues Concern ing Natural Religion, which he did not publish during his lifetime.In accordance with his instructions, the book was published as a posthumous work in 1779.His Essay on Miracles became famous, arguing that such events as miracles can never have proper historical evidence. His History of England, published in 1755 and in subsequent years, was eager to prove that the Tories were superior to the Whigs, and the Scots superior to the English; he did not think that history deserved a philosophical detachment. In 1763 he visited Paris and was highly regarded by philosophes (sages).Unfortunately, he befriended Rousseau, with whom he had a famous quarrel.Hume showed admirable tolerance, but Rousseau, who suffered from persecution paranoia, insisted on breaking with him.

Hume once described his own character in a self-conceived obituary, which he called "apocalypse": "I am a mild-natured person, restrained temper, cheerful, sociable and pleasant; I can have affection, But hardly any hatred; and great moderation in all my affections. Even my chief passion, my desire for literary fame, never made me sour, though I was often disappointed." All these words Everything we know about him is confirmed. Hume's Treatise of Human Nature is divided into three volumes, each dealing with reason, emotion, and morality. What is new and important in his teaching is in the first volume, so I will confine myself to the first volume.

He started by talking about the difference between "impression" and "idea".They are two classes of perceptions, of which the impression is the one of greater force and violence. "By ideas I mean the vague mental images of impressions in thinking and reasoning." Ideas, at least in the case of mere ideas, are like impressions, but vaguer than them. "Every simple idea has a simple impression, like it; and every simple impression has a corresponding idea." "All our simple ideas, when they first appear, are derived from simple impressions, which have the same The simple idea corresponds exactly to this simple impression." On the contrary, the compound idea does not necessarily resemble the impression.We have never seen a winged horse and can imagine a winged horse, but the composition, composition, and elements of this compound concept are all derived from impressions.Impressions come first, and the evidence for this is empirical; for example, people who are born blind have no idea of ​​colour.

Among the various ideas, those that retain a considerable degree of vividness of the original impression belong to memory, and others belong to imagination. There is a section (Volume I, Book I, Section VII) "On Abstract Ideas", which begins with a passage which is in striking agreement with Berkeley's following doctrine: "All general ideas are nothing but superimposed on a certain term This term gives this idea a wider meaning, and makes it recall other individuals similar to itself at the corresponding time." Hume maintains that when we hold the idea of ​​​​"person", this idea has " all the individual qualities of the impressions of man." "The mind cannot form any concept of quantity or quality without forming exact concepts of degrees of quantity or quality." "Abstract ideas, however general they may become in representing [impressions], are themselves individual." This theory is A modern nominalism, it has two shortcomings, one is logical and the other is psychological.Let me talk about the logical shortcomings first.Hume says: "When we find a resemblance among several objects, we attach the same name to all of them." All nominalists would agree.But in fact, generic names like "cat" are as unreal as the universal cat.The nominalist solution to the problem of universals thus fails because of an incomplete application of its principles; it is a mistake to apply this principle only to "things" and not also to words.

The psychological shortcomings are serious, at least as far as Hume is concerned.The whole set of theories he talked about, which regards ideas as copies of impressions, has the disadvantage of ignorance, ambiguity, and confusion.For example, I have seen a flower of what color, and when I later remembered its mental image, which lacked precision, meaning that there were several shades very similar to each other, it might be its mental image, to use Hume's terminology. That is, "concept". It is not true to say that the mind cannot form any concept of quantity or quality unless it forms precise concepts of the degrees of either quantity or quality.Suppose you ever met a man who was six feet one inch tall.You keep an image of him, but that image would probably work for someone an inch taller or shorter.Ambiguity is different from generality, but shares some of the same characteristics.Hume's failure to notice the ambiguity led him into needless difficulties, such as the difficulty concerning whether it was possible to imagine a never-before-seen hue between two very similar hues seen middle.If the two shades are sufficiently similar, any mental image you can form will equally apply to both shades and tones in between.When Hume said that ideas arise from impressions exactly represented by them, he escaped from the reality of psychology.

Just as Berkeley drove the concept of substance from physics, so Hume drove the concept of substance from psychology.He says that there is no such thing as an impression of a "self," and therefore no idea of ​​a "self" (Vol. I, Part IV, Section VI). "As far as I am concerned, when I observe very closely what I call myself, I always encounter some particular perception, cold or hot, light or dark, love or hate, pain or pleasure. Perceptions. At no time have I grasped myself apart from perceptions, and I have never observed anything but perceptions." He concedes ironically that perhaps some philosophers perceive their "But apart from some metaphysicians of this kind, I may boldly assert to the rest of the human race that the self is nothing but a cluster or group of distinct perceptions succeeding each other with inconceivable rapidity, and in an uninterrupted cycle. In flux and in motion."

This denial of self-concept is very important.Let's see exactly what it claims, and how much it holds up.First, there is no such thing as a self, if there is one, that is never perceived, so we cannot have a self-concept.If this argument is to be admitted, it must be carefully stated.No one is aware of his own brain, yet in an important sense he has the "idea" of a brain.This kind of "idea" is the inference of perception, which does not belong to the basic ideas of logical meaning; it is compound ideas and descriptive - if Hume's principle that all simple ideas are derived from impressions is correct, it must be so; and If this principle is denied, we will have to go back to the concept of "birth".In modern parlance, it can be said that ideas of unperceived things or events can always be defined in terms of perceived things or events, so that by substituting definitions for terms that are defined, we can never introduce any unperceived things or events to describe what we know from experience.As far as our present problem is concerned, all psychological knowledge can be stated without introducing the "self".Moreover, the "self" defined in this way can only be a cluster of perceptions, not a new simple "thing".I think that any thorough empiricist must agree with Hume on this point.

But it does not follow that the mere self does not exist; only that we cannot know whether it exists or not, and that the self cannot form any part of our knowledge except as a cluster of perceptions.This conclusion, which weeds out the last remaining use of "substance," is metaphysically important.It is important in that it abolishes all supposed knowledge of the "soul" in theology.It is also important in the analysis of cognition because it shows that the categories subject and object are not fundamental.On this question of the self, Hume made a significant advance over Berkeley.

The most important part of the whole Treatise of Human Nature is the section called "On Knowledge and Probability".What Hume called "probability" did not refer to the kind of knowledge contained in theoretical probability theory, such as the chance of rolling a double six with two dice is one in thirty-six.This knowledge itself is not probabilistic in any special sense; it is certain as far as knowledge can be.Hume was dealing with indeterminate knowledge of the kind obtained from empirical data by non-demonstrative inferences.This includes all our knowledge about the future and all our knowledge about the unobserved parts of the past and present.In fact, apart from direct observation on the one hand, and logic and mathematics on the other, it includes everything else.By analyzing this "probable" knowledge, Hume draws some skeptical conclusions that are as difficult to refute as they are to accept.It turned out to be a table for philosophers, which, in my opinion, has not been enough to meet the opponent's challenge so far.

Hume begins by distinguishing seven philosophical relations: resemblance, identity, relation of time and place, ratio of magnitude or number, degree of any quality, reciprocity, and causality.These relations, he says, may be divided into two kinds, namely, those which depend only on ideas, and those which are capable of changing ideas without changing them.In the first category are likeness, oppositeness, degree of quality, and ratio of quantity or number.But space-time relationships and causality fall into the second category.Only relations of the first kind give definite knowledge; our knowledge of the others is only anecdotal.Algebra and arithmetic are the only sciences in which we can make a long chain of reasoning without loss of certainty. Geometry is not as certain as algebra and arithmetic, because we cannot be sure that the axioms of geometry are correct.There are many philosophers who erroneously assume that ideas in mathematics "must be apprehended from pure and intellectual viewpoints peculiar to the higher faculties of the soul."Hume says that the error of this opinion is immediately apparent as soon as it is remembered that "all our ideas are copied from our impressions." The three relations that do not depend solely on ideas are identity, space-time relations, and causality.In the first two relationships, the mind does not go beyond what is directly present to the senses. (Hume argued that spatial-temporal relations can be perceived and form part of impressions).Only causation enables us to infer from one thing or event to some other: "what convinces us from the existence or action of an object that it follows, or precedes, some other existence or action which produces this connection There is only cause, effect, relationship." A difficulty arises from Hume's assertion that there is no such-called impression of causality.Just by observing A and B, we can perceive that A is above B, or on the right of B, but we cannot perceive "Because of A, the result of B".In the past, causality has been more or less compared to the relationship between ground and assertion in logic, but Hume rightly recognized that this comparison is wrong. In Cartesian philosophy, as well as in the philosophy of the scholastics, the connection between cause and effect is held to be as necessary as a logical connection.The first really serious challenge to this view came from Hume, with whom the modern philosophy of causality begins.Like almost all philosophers up to and including Bergson, he thinks that the law of causality is a proposition of the form "because A results in B", where A and B are two types of events; Philosophers do not seem to know the fact that laws are not found in any developed science.But much that has been said by philosophers can be transformed to apply to the laws of causality as they actually arise; and we may therefore disregard this for the present. Hume began by saying that the force that causes one object to produce another is not obtained from the discovery of the concepts of these two objects, so we can only know the cause and effect by experience, not by reasoning or introspection.He said that the statement "everything that happens has a cause" is not a statement of intuitive certainty like a proposition in logic.According to him: "If we consider objects in their own right, never looking beyond the ideas we have formed about them, then there are no objects which imply the existence of other objects." According to Hume, it must be experience that makes Man has knowledge of cause and effect, but not merely the experience of events A and B that are causally related to each other.It must be experience, because the connection is not logical; And since we find nothing in A alone that would cause A to produce B, it cannot be just the experience of individual events of A and B.The necessary experience, he said, is that Type A events and Type B events are often linked to this experience.He points out that in experience when two objects are often connected, we do in fact infer one from the other. (By "inference" he means that the perception of one leads us to anticipate the other; he does not mean formal or definite inferences.) "Presumably, necessary connections depend on inferences," but not the other way around.In other words, seeing A makes people expect B, so we believe that there is a necessary relationship between A and B.This inference is not determined by reason, for that would require us to assume the uniformity of nature, which itself is not necessary, but is inferred from experience. Hume then came to this point of view: we say "because of A, the result of B" means that A and B are in fact often connected, not that there is a certain relationship between them. "We have no notion of cause and effect, except that of certain objects which have always been connected together. . . . We have no insight into the reason for this connection." Hume supported his theory with a definition of "belief", which he believed to be "a distinct idea related or associated with a present impression".If A and B have been constantly conjoined in past experience, the impression of A produces this vivid idea of ​​B, and constitutes a belief in B, by virtue of the association.This explains why we believe that A and B are related: A's perceptual representations and B's ideas are related, so we think that A and B are related, although this opinion is really unfounded. "There is no found relation of unity between objects; and that we can infer from the presence of one object that another will be experienced, from no other principle than habit acting on the imagination." In our What appears to be a necessary connection between objects is really only a connection between the ideas of those objects. Hume repeated this assertion many times; , it is this impression, that is, this determination." The repetition of the instances which give us the belief "Because of A, consequently of B" imparts nothing new to the object, but creates the idea in the mind. the union of things; thus "necessity is not something that resides in objects, but in the mind." Now to the question of how we should look at Hume's doctrine.There are two parts to this doctrine, one is the objective part and the other is the subjective part.The objective part is: when we conclude that "because of A, the result of B", as far as A and B are concerned, what actually happened is that it has been observed repeatedly that the two are connected, that is to say, A is always immediately followed by B, or soon followed by B; we have absolutely no reason to say that A must be followed by B, or will be followed by B in the future. And no matter how often A is followed by B, we have no reason to suppose that there is any relationship beyond "sequential order".In fact, causality can be defined by "sequence", which is not an independent concept. The subjective part of Hume's theory says: Because of the repeated observation that A and B are joined, the result is: because of [the impression of] A, the result of [the idea of] B.However, if we want to define "because..., as a result..." in terms of the objective part of the doctrine, then the above statement must be rephrased.Substituting the definition of "because..., the result..." into the above sentence becomes: "It has been observed many times: the connection between the two objects A and B that has been observed many times has always been followed by such occasions: A's impression is followed by B's idea." We may as well admit that this statement is true, but it is hardly to the extent that Hume ascribes to the subjective part of his doctrine.He maintains again and again that the repeated association of A and B is not a reason for expecting that the two will also be associated in the future, but only a cause of such an expectation.That is to say, the experience of repeatedly linking this event is repeatedly linked to a linking habit.But, if the objective part of Hume's doctrine is admitted, the fact that unions have been repeatedly formed in such cases in the past is no reason to suppose that such unions will continue, or that new unions will be formed in similar circumstances.The fact is that, with respect to psychology, Hume affirms the existence of causal relations in the sense in which he generally denounces them.Give an example.I see an apple and anticipate that if I eat it, I will experience a certain taste.In Hume's opinion, there is no reason why I should always experience this taste: the law of custom accounts for the existence of my anticipation, but is not sufficient to justify it.But the law of habit itself is a law of causality.So, if we take Hume's opinion seriously, we must say this: Although seeing an apple has been associated with anticipating a certain taste in the past, there is no reason why it should continue to be so.Maybe next time I see an apple I'll expect it to taste like roast beef.Right now, you may think it's unlikely; but that's no reason to expect that five minutes from now you'll think it's unlikely.If Hume's objective doctrine is correct, our expectations in the realm of psychology are as unjustified as in the realm of physics.Hume's theory might as well be jokingly described as follows: "The proposition 'because A results in B' means 'because [there is] the impression of A, consequently [there is] the idea of ​​B'".By definition, it's not a masterpiece. We must therefore look more closely at Hume's doctrine of objectivity.There are two parts to this doctrine: (1) When we say, "Because of A, consequently of B," we are entitled to say only that, in past experience, A and B have frequently occurred together or in rapid succession. , A is not followed by B or A is not accompanied by B, and it has never been observed. (2) However many instances we may observe of the connection of A and B, that is no reason for expecting the connection of the two at some time in the future, although that is the cause of this expectation, that is, Said that it has been repeatedly observed that it is connected with this expectation.These two parts of the doctrine may be stated as follows: (1) In the causal relationship, there is no undefinable relationship except for "connection" or "succession"; (2) Simple enumeration and induction are not a sound form of argument.Ordinary empiricists have always accepted the first of these two arguments and denied the second. When I say they always deny the second argument, I mean that they always believe that if a fairly large number of instances of a connection are known, the probability of the connection occurring in the next instance is more than half; or, even if They don't claim exactly that, they also claim something to the same conclusion. At present I do not wish to discuss induction, which is a great and difficult subject; now I would like to say that, even admitting the first half of Hume's doctrine, the denial of induction makes all predictions about the future, even the prediction that we shall continue to expect unreasonable stuff.I do not mean merely that our predictions may be wrong; this must be admitted anyway.What I'm saying is that even in the case of our strongest predictions like the sun coming out tomorrow, there's not the slightest reason to assume that it's more likely to be confirmed than it won't be.With this condition attached, I will come back and talk about the meaning of "cause and effect". People who disagree with Hume maintain that "cause and effect" is a special relationship. If there is such a relationship, there must be a certain sequence, but if there is a certain sequence, there may not be such a relationship.Recalling the Cartesian clock: two perfectly accurate clocks may strike one after the other invariably, but neither is the cause of the other.In general, those who hold this view maintain that, although in most cases we have to infer, somewhat dangerously, causality from the frequent connection of events, we can sometimes perceive causality.On this point, let us see what are the reasons for and against Hume's views. Hume summed up his argument in the following words: "I realize that, of all the bizarre paradoxes I have held up to this point in this treatise, or have had to propose hereafter, the present one is the most extreme, relying on sound proof and reasoning, and I We can hope that it will be recognized and break the deep-seated prejudices of men. How often must we repeat these words to ourselves before we are convinced of this doctrine: Let two objects or effects, however related to each other, be simply The seeing of them never gives us an idea of ​​the force or connection between the two, the one; this idea arises from the repetition of the combination of the two, the two; Repetition neither reveals nor causes anything in its object, but by means of the change of routine which it manifests affects only the mind, this third; Forces and inevitability that are imperceptible to objects from without are one and the same.” Hume is often accused of having an overly atomistic view of perception, but he does admit that certain relationships are perceivable.He says: "We must not take any part of our observations of identity, of the relation of time and place, to be interpreted as inferences; What is presented to the senses." The difference between causality, he says, is that it takes us beyond sensory impressions, telling us of the unperceived.As an argument, this seems inappropriate.We believe that there are many relationships between time and place that we cannot perceive: for example, we think that time extends forward and backward, and space extends beyond the four walls of the living room.Hume's real argument is that while we sometimes perceive the relationship of time and place, we never perceive causality, so even if causality is admitted, it must be inferred from the relationships that are perceived.The debate then becomes a debate about empirical facts: do we sometimes perceive a relationship that could be called causal?Hume said "no" and his adversaries said "yes", it is not easy to see how either side could produce evidence. I think perhaps the strongest arguments on Hume's side can be drawn from the properties of causal laws in physics.It seems that a simple rule of the form "because of A, result of B" is never allowed in science, except as a premature formulation in the early stages.In the most advanced sciences, the laws of causality, which take the place of such simple laws, are so complex that no one can suppose them to have arisen in perception; they are evidently subtle inferences drawn from observed natural tendencies.I have not included modern quantum theory, which further confirms the above conclusions.As far as natural science is concerned, Hume is absolutely right: "Because of A, the result of B" will never be recognized, and our tendency to recognize it can be explained by the law of habit and the law of association.These two laws are themselves, in strict form, delicate statements about the nervous tissue—first about its physiology, second about its chemistry, and finally about its physics. But Hume's opponents, while admitting all that has been said about the natural sciences, may not yet admit that they are utterly refuted.He might say that there is no shortage of instances in psychology where causality can be perceived.The whole concept of cause is mostly derived from the action of the will. It can be said that between a certain action of will and the subsequent action, we can perceive a relationship beyond a certain sequence.The same could be said of the relation between sudden pain and screaming.However, this opinion becomes a very difficult opinion in terms of physiology.Between the will to move the arm and the movement that follows it, there is a long chain of causal mediations consisting of operations in the nerves and muscles.We perceive only the two ends of this process, volition and action, and we are mistaken if we think we see a direct causal connection between these two ends.Although this set of principles is not conclusive on current general issues, it also shows that it would be rash for us to expect to perceive causality if we think we perceive causality.So, weighing the two sides, Hume's view that there is nothing in cause and effect that is anything but a certain succession prevails.But the evidence is not as conclusive as Hume thought. Hume was not content to reduce the evidence of causal connection to the experience of repeated connections of events; he went on to argue that such experience was no reason to expect similar connections in the future.For example, (recalling an earlier instance) when I see an apple, past experience leads me to expect that it will taste like an apple rather than roast beef; but there is no rational reason for this expectation.If there were such a reason, it would have to proceed from the principle: "The cases which we have never experienced are similar to those which we have experienced." This principle is not logically necessary, since we can at least conceive Natural processes change.Therefore, it should be a principle of probability.But all probable arguments presuppose this principle, so that it cannot itself be proved by any probable argument, nor can any such argument even render it probable truth. "The assumption that the future is similar to the past is not based on any kind of argument, but is entirely a matter of habit." The conclusion is a thoroughly skeptical one: "All probabilistic reasoning is nothing but a kind of sensibility. Not only in poetry and music we must follow our own taste and emotion, but also in philosophy. If I am convinced of a principle, it is only an idea, which is stronger. It is imprinted on my mind. If I think one set of arguments preferable to another, it is only a decision of my personal feeling as to the superiority of appeal of one set. There is no detectable bond of unity between the objects; and From the appearance of one object to the existence of another we draw any inferences, from no other principle than habit acting on the imagination." The end result of Hume's investigations of what is commonly thought of as knowledge is not what we might expect him to be looking for.The subtitle of his book is "A Probe into the Introduction of Experimental Reasoning in Psychiatry".It is evident that he began with the conviction that the scientific method yields truth, all truth, and only truth; yet he ended up convinced that, because we know nothing, belief is never a rational thing. Having presented the arguments in support of skepticism (Book I, Book IV, Section I), he does not proceed to refute them, but appeals to the natural credulity of man. "Nature, by an absolute and uncontrollable necessity, determines not only our breathing and feeling, but also our judgment. So long as we are awake, we cannot prevent ourselves from thinking; turning our eyes to the surrounding objects in broad daylight, We cannot prevent ourselves from seeing these objects; in the same way, we cannot resist having a clearer and more complete view of certain objects because of their habitual connection with present impressions. Whoever has taken pains to refute this absolute The skeptic, who actually argues without adversary, endeavors to establish by argument a faculty which nature has previously instilled in man and made inevitable. I, then, Therefore, the purpose of developing the arguments of the school of the absurd so carefully is only to make the reader understand that my hypothesis is correct: concerning the cause and the effect of ours. ·cut·deduction·argument·no·non·is·from·habit·habit·come;·believe·read·and·it·say·is·we·us·nature·nature·in·think·consider· Part, part, action, behavior, not, like, say, is, feel, feel, part, part, action, action, comparison, comparison, appropriate." He goes on to write (Vol. I, Part IV, Section II): "Even if the skeptic asserts that he cannot justify his reason with reason, he still continues to reason and believe; Nor can one pretend to assert that the principle of the existence of bodies is true, yet one must agree to the principle of the existence of bodies.  … We may ask, what prompts us to believe Thing, body, being? But it is vain to ask yes, no, yes, thing, body. In all our inferences, this must be taken as unquestionable." The above is the beginning of the section "On Skepticism Concerning the Senses".After a lengthy discussion, this section concludes with the following conclusions: "This skeptical doubt about both reason and feeling is a disease that can never be cured, a disease that, however much we drive it away, and sometimes seems to get rid of it altogether, it just happens to invade us every moment. chronic illness. . . . Only indifference and inattention can be our little remedy.For this reason, I trust both points completely; and think that whatever the reader's opinion may be at the moment, an hour from now he must be convinced that both the outer world and the inner world exist. " The study of philosophy is a pleasant way of passing the time for a certain temperament, and there is no reason to study it—so Hume argued. "In all the events of life we ​​should still retain our skepticism. If we believe that fire warms a man, or water refreshes the spirit, it is only because we shall suffer too much otherwise. No, if we If he is a philosopher, it should only be based on the principle of skepticism, a tendency of our feelings to think that way." If he gives up thinking, "I feel that I have a loss in pleasure; source of philosophy." Whether Hume's philosophy is right or wrong, it represents the bankruptcy of the reasoning spirit in the eighteenth century.Like Locke, he set out with this intention: to be rational, to experience, to believe in nothing, but to pursue whatever knowledge can be obtained from experience and observation.But because he possessed an intellect superior to that of Locke, with a greater acuity for analysis and a smaller measure for accommodating inconsistencies, he came to the unfortunate conclusion that nothing could be known from experience and observation. 所谓理性的信念这种东西是没有的；“我们如果相信火使人温暖，或相信水让人精神振作，那无非因为不这样想我们要吃太大的苦头。”我们不得不抱有信念，但是任何信念都不会根据理性。而且，一个行为方针也不会比另一个方针更合理，因为一切方针同样都以不理性的信念为基础。不过这个最后结论休谟似乎并没有得出来。甚至在他总结第一卷的各个结论的那一章，怀疑主义最甚的一章中，他说道：“一般讲，宗教里的错误是危险的；哲学里的错误只是荒谬而已。”他完全没资格讲这话。“危险的”是个表示因果的词，一个对因果关系抱怀疑的怀疑论者不可能知道任何事情是“危险的”。 实际上，在《人性论》后面一些部分，休谟把他的根本怀疑全忘到九霄云外，写出的笔调和当时任何其他开明的道德家会写出的笔调几乎一样。他把他推赏的救治方剂即“不关心和不留意”用到了自己的怀疑上。从某种意义上讲，他的怀疑主义是不真诚的，因为他在实践中不能坚持它。可是，它倒有这样的尴尬后果：让企图证明一种行为方针优于另一种行为方针的一切努力化为泡影。 在这样的自我否定理性精神的后面跟随着非理性信念大爆发，是必不可免的事。休谟和卢梭之间的争吵成了象征：卢梭癫狂，但是有影响；休谟神志正常，却没有追随者。后来的英国经验主义者未加反驳就否定了他的怀疑论；卢梭和他的信徒们同意休谟所说的任何信念都不是以理性为基础的，然而却认为情胜于理，让感情引导他们产生一些和休谟在实践上保持的信念迥然不同的信念。德国哲学家们，从康德到黑格尔，都没消化了休谟的议论。我特意这样讲，尽管不少哲学家和康德有同见，相信对休谟的议论作了解答。其实，这些哲学家们——至少康德和黑格尔—— 代表着一种休谟前型式的理性主义，用休谟的议论是能够把他们驳倒的。凭休谟的议论驳不倒的哲学家是那种不以合理性自居的哲学家，类如卢梭、叔本华和尼采。整个十九世纪内以及二十世纪到此为止的非理性的发展，是休谟破坏经验主义的当然后果。 所以，重要的是揭明在一种完全属于、或大体属于经验主义的哲学的范围之内，是否存在对休谟的解答。若不存在，那么神志正常和精神错乱之间就没有理智上的差别了。一个相信自己是“水煮荷包蛋”的疯人，也只可能以他属于少数派为理由而指责他，或者更不如说（因为我们不可先假定民主主义），以政府不跟他意见一致为理由而指责他。这是一种无可奈何的观点，人不得不希望有个什么逃避开它的方法才好。 休谟的怀疑论完全以他否定归纳原理为根据。就应用于因果关系而言，归纳原理讲：如果一向发现甲极经常地伴随有乙，或后面跟着有乙，而且不知道甲不伴随有乙或后面不跟着有乙的任何实例，那么大概下次观察到甲的时候，它要伴随有乙或后面跟着有乙。要想使这条原理妥当，那么必须有相当多的实例来使得这个盖然性离确实性不太远。这个原理，或其他推得出这个原理的任何一个原理，如果是对的，那么休谟所排斥的因果推理便妥实有据，这固然并不在于它能得出确实性，而在于它能得出对实际目的说来充分的盖然性。 假如这个原理不正确，则一切打算从个别观察结果得出普遍科学规律的事都是谬误的，而休谟的怀疑论对经验主义者说来便是逃避不开的理论。当然，若不犯循环论法，这原理本身从观察到的齐一性是推论不出来的，因为任何这种推论都需要有这个原理才算正当。所以，它必定是一个不基于经验的独立原理，或由这种独立原理推出来的原理。在这个限度内，休谟证明了纯粹经验主义不是科学的充足基础。但是，只要承认这一个原理，其它一切都能按照我们的全部知识基于经验这个理论往下进行。必须承认，这是严重违反纯粹经验主义，非经验主义者的人或许问，如果一种违反是许可的，为什么旁的违反就得禁止。不过这些都是由休谟的议论非直接引起的问题。他的议论所证明的是——我以为这证明无法辩驳——归纳是一个独立的逻辑原理，是从经验或从其它逻辑原理都推论不出来的，没有这个原理，便不会有科学。