Will the Sun Rise Tomorrow? 明天太陽會升起嗎？

🔊Hume begins by sorting everything we can know into two kinds. Relations of ideas are truths such as “three times five is equal to the half of thirty.” They can be established by thinking alone, and to deny one is to contradict oneself. Matters of fact are truths such as “the sun is shining.” They can be learned only from experience, and the denial of any of them is perfectly conceivable. His example has become famous: “That the sun will not rise to-morrow is no less intelligible a proposition, and implies no more contradiction than the affirmation, that it will rise.” So no matter of fact can be proved by pure reasoning. If we know the sun will rise, we know it from experience.

休謨一開始，把我們能知道的一切分成兩類。「觀念的關係」是像「三乘五等於三十的一半」這樣的真理；它們光靠思考就能確立，否定其中任何一個就會自相矛盾。「實際的事情」是像「太陽正照耀著」這樣的真理；它們只能從經驗得知，而它們當中任何一個的否定，都完全可以設想。他舉的例子後來很有名：「『明天太陽不會升起』這個命題，並不比『它會升起』這個斷言更難理解，也沒有蘊含更多的矛盾。」所以，沒有任何實際的事情可以靠純粹的推理來證明。如果我們知道太陽會升起，那是從經驗知道的。

🔊Experience tells us directly only about what we are now perceiving or remember. To get beyond that, to the fire in the next room or the eclipse next year, we reason from cause to effect. Hume therefore asks what we observe when we observe one thing causing another. One billiard ball rolls toward a second; they touch; the second moves. We see the first motion, the contact, and the second motion. We do not see any power in the first ball that forces the second to move. A person who had never seen bodies collide could not tell, from looking at the first ball, what would happen. What experience adds is repetition: events of the first kind have always been followed by events of the second. He calls this constant conjunction. The idea that the effect must follow is something we supply ourselves.

經驗直接告訴我們的，只有我們此刻知覺到的、或記得的事。要越過這個範圍，推到隔壁房間的火、明年的日食，我們靠的是從原因到結果的推理。所以休謨問：當我們觀察到一件事「造成」另一件事時，我們到底觀察到了什麼？一顆撞球滾向第二顆；兩者接觸；第二顆動了。我們看見第一個運動、接觸、第二個運動。我們並沒有看見第一顆球裡有什麼「力量」，逼得第二顆非動不可。一個從沒見過物體碰撞的人，光看第一顆球，是說不出會發生什麼的。經驗添上的是重複：第一類事件之後，總是跟著第二類事件。他把這稱為「恆常連結」。至於「結果必定跟著發生」這個觀念，是我們自己添上去的。

🔊Now the problem can be stated exactly. All our conclusions from experience take the form: such things have always gone together, so they will go together again. That step is valid only with an extra premise, that the future will resemble the past. How might the premise be established? It is not a relation of ideas, since a change in the course of nature is conceivable. It must then be a matter of fact, known from experience. But “all inferences from experience suppose, as their foundation, that the future will resemble the past.” To argue that nature will be uniform because it has been uniform so far is to use the principle in order to prove the principle. In Hume’s words, it “is begging the question.”

現在可以把問題精確地說出來了。我們從經驗得出的一切結論，都是這個形式：這些事一向相伴出現，所以它們會再度相伴出現。這一步要有效，需要一個額外的前提：未來會與過去相似。這個前提要怎麼確立？它不是觀念的關係，因為自然進程的改變是可以設想的。那它就必定是實際的事情，要從經驗得知。可是「一切從經驗出發的推論，都預設未來會與過去相似，以此作為基礎」。主張「自然會保持齊一，因為它到目前為止都是齊一的」，就是用這個原則來證明這個原則。用休謨的話說，這「是乞題」。

🔊Notice what the argument does and does not claim. Hume does not say that the sun will fail to rise, or that expecting it is foolish. He says that the expectation is not produced by any argument, because no good argument for it exists.

注意這個論證主張了什麼、沒有主張什麼。休謨沒有說太陽將不會升起，也沒有說期待它升起是愚蠢的。他說的是：這個期待不是由任何論證產生的，因為根本沒有支持它的好論證。

🔊If reasoning does not produce our expectations, what does? Hume’s answer is habit. After seeing two kinds of event joined many times, the mind passes from one to the other automatically, and feels that passage as necessity. “Custom, then, is the great guide of human life. It is that principle alone which renders our experience useful to us.” Infants and animals learn from experience in just this way, and nobody supposes that they do it by constructing arguments. Hume calls this a skeptical solution, because it explains why we believe and leaves the question of justification where it was. He was content with that. Nature, he observed, has not left so important a matter to our fallible reasoning.

如果產生我們期待的不是推理，那是什麼？休謨的答案是習慣。看見兩類事件多次相連之後，心靈就自動地從一個過渡到另一個，並且把這個過渡感受為必然。「所以，習慣是人類生活的偉大嚮導。唯獨這個原則，使我們的經驗對我們有用。」嬰兒與動物正是這樣從經驗學習的，而沒有人認為牠們是靠建構論證辦到的。休謨把這稱為「懷疑論的解決」，因為它解釋了我們為什麼相信，卻讓證成的問題留在原地。他對此感到滿意。他說，這麼重要的事，自然並沒有交給我們那容易出錯的推理。

🔊Bertrand Russell gave the problem its most memorable form. A chicken is fed by the farmer every morning of its life and comes to expect its food when he appears. “The man who has fed the chicken every day throughout its life at last wrings its neck instead, showing that more refined views as to the uniformity of nature would have been useful to the chicken.” The chicken’s reasoning was the same as ours about the sunrise. Europeans were in a similar position with “all swans are white,” on the evidence of every swan anyone had seen, until black swans were found in Australia in 1697.

羅素給了這個問題最令人難忘的形式。一隻雞，一輩子每天早上都由農夫餵食，於是一見他出現就期待食物。「那個在雞的一生中每天餵牠的人，最後卻扭斷了牠的脖子；這表明，對自然的齊一性若有更精細的看法，對這隻雞會很有用。」這隻雞的推理，跟我們對日出的推理是一樣的。歐洲人對「所有天鵝都是白的」也曾處在類似的位置上，憑的是任何人見過的每一隻天鵝；直到 1697 年在澳洲發現了黑天鵝。

🔊Chinese tradition supplies a pair of stories that mark the two ways of going wrong. In the Han Feizi a farmer sees a hare run into a tree stump and break its neck, and he puts down his plow and waits by the stump for another (守株待兔). He has generalized from a single case. In the Liezi a man of Qi cannot eat or sleep for fear that the sky will fall (杞人憂天). He refuses to rely on any regularity at all. Everyone agrees that good sense lies somewhere between the two. Hume’s point is that we cannot prove where.

中國的傳統提供了一對故事，正好標出兩種走錯的方式。在《韓非子》裡，一個農夫看見一隻兔子撞上樹樁、折斷了脖子，於是放下犁，守在樹樁旁等下一隻（守株待兔）。他是從單一個案做了概括。在《列子》裡，一個杞國人因為害怕天會塌下來而寢食難安（杞人憂天）。他拒絕倚賴任何規律。人人都同意，明智的做法在兩者之間的某處。休謨的重點是：我們無法證明是在哪裡。

🔊Hume’s problem has been called the scandal of philosophy, and the attempts to answer it fall into four groups. Kant held that causality is not learned from experience at all. It is a form the mind imposes, a condition of having any ordered experience, and so it is guaranteed to hold of whatever we can experience. Karl Popper accepted Hume’s argument and concluded that science does not use induction. Scientists make bold guesses and try to refute them, and a theory that survives is never thereby shown to be true or even probable. Hans Reichenbach offered a practical defense. We cannot know that nature is uniform, but if it is, induction will find the pattern, and if it is not, no method will work, so we lose nothing by using it. And P. F. Strawson argued that the demand is confused. To ask whether it is reasonable to rely on induction is like asking whether the law is legal: relying on past experience is part of what “reasonable” means.

休謨的問題被稱為「哲學的醜聞」；回答它的嘗試可以分成四組。康德主張，因果性根本不是從經驗學來的。它是心靈加上去的形式，是擁有任何有秩序的經驗的條件；所以對我們所能經驗的一切，它保證成立。波普接受休謨的論證，並且推出：科學並不使用歸納。科學家提出大膽的猜測，再設法反駁它們；一個存活下來的理論，並不因此就被證明為真，甚至也沒有被證明為很可能為真。萊辛巴哈提出一種實用的辯護：我們無法知道自然是不是齊一的；但如果它是，歸納法會找到那個規律；如果它不是，任何方法都不管用；所以用歸納法，我們什麼也不會損失。史陶生則論證，這個要求本身就搞混了。問「倚賴歸納是否合理」，就像問「法律是否合法」：倚賴過往經驗，本來就是「合理」這個詞的一部分意思。

🔊Each reply has its critics, and in 1955 Nelson Goodman made the problem harder. Call an object “grue” if it is first examined before the year 2100 and is green, or is not examined before then and is blue. Every emerald seen so far has been green, and every one has also been grue. Yet we expect future emeralds to be green and not grue. Even if induction is granted, then, something must tell us which patterns to project, and that further rule cannot come from the evidence, which fits both.

每一種回應都有批評者；而在 1955 年，古德曼把問題弄得更難了。把一個東西稱為「綠藍」（grue），如果它在 2100 年之前首次被檢視而且是綠的，或者在那之前沒被檢視過而且是藍的。迄今見過的每一顆祖母綠都是綠的，也都是「綠藍」的。然而我們預期未來的祖母綠是綠的，不是「綠藍」的。所以，就算承認了歸納法，還是得有什麼東西告訴我們該投射哪些規律；而這條進一步的規則，不可能來自證據，因為證據跟兩者都相符。

🔊What should one conclude? Not that anything goes. Hume himself drew a sober moral: “A wise man … proportions his belief to the evidence.” The lesson of the problem is that evidence comes in degrees and never amounts to proof. Scientific laws are the best-supported expectations we have, and they remain expectations. This is the second unit’s conclusion, and it matches the first. Logic could tell us that our premises conflict and could not tell us which to give up. Experience can tell us what has happened so far, and what will happen next is something we infer at our own risk. The sun will rise tomorrow. You are right to be sure of it, and you cannot say why without assuming it.

該得出什麼結論？不是「怎樣都行」。休謨自己引出了一個清醒的教訓：「有智慧的人……使他的信念與證據成比例。」這個問題的教訓是：證據有程度之分，而且永遠不等於證明。科學定律是我們擁有的、根據最充分的期待，而它們仍然是期待。這是第二單元的結論，而它跟第一單元的結論相呼應。邏輯能告訴我們前提互相衝突，卻不能告訴我們該放棄哪一個。經驗能告訴我們到目前為止發生了什麼；至於接下來會發生什麼，那是我們自負風險去推論的。明天太陽會升起。你對此有把握是對的；而你說不出為什麼，除非先假定它。