欧阳峰的博客

在大学读书的时候，在数理化之余也喜欢舞文弄墨，写个古诗，对个对子之类，自觉思路敏捷，沾沾自喜。有一次和中文系一位大哥谈起，他说：真正学中文的都不玩这一套，而是讲究语言平实，意境高雅。听后我很郁闷，却还是不明就里。

最近在网上认识了“师姐”，才得以一睹高人风范。“师姐”不仅为人厚道，阅历深广，而且文字功夫很深。不管是胡言调侃（师姐语）还是引经据典，在她面前总有看高手下棋的感觉：同样是这几个字，同样是这些个语法规则，结果却大不相同。什么叫“功底”？这下至少有了些感性认识。

以下是一位网友编的《shijie－师姐专辑》。

http://www.starlakeporch.net/bbs/read.php?45,23220

【以下是和网友关于弗里曼观点的讨论。弗里曼原文在http://www.ideachannel.tv/】

【我】

Concerning free market and government control,to me the theoretical issue is pretty clear. When market works, weshould leave it alone. When it fails, we need governmentintervention. Moreover, we have good understandings on themechanisms with which a market can fail. However, applying thetheory to real-world problems would require judgments, assumptions,etc., upon which we often don’t agree.

For example, Friedman was against minimum wage laws. He believedthat it will force the employers not to hire the workers who arenot valuable enough to justify the mandatory pay. Minimum wage lawadvocates, on the other hand, believe that the employers are takingunfair advantages over the workers whose options are limited. Ifthe employers are forced to pay a higher wage, it will reduce theirprofit, without any other effects. I don’t think both sidesdisagree on the basic theory or logic. It boils down to thefollowing question: is the labor market a free market withefficient competition, or is it monopolized by the employers? Inreality, I think, the truth is in between the two views. The labormarket is neither in perfect competition nor in monopsony.Therefore, the dispute on minimum wage laws will perpetuate, untilaccurate mathematical models can be established to quantitativelypredict and analyze the issue. I don’t expect, however, that suchmodel will materialize in the near future.

Other issues such as healthcare and social security involve thesame kind of controversy that can only be resolved by quantitative,instead of qualitative, arguments. Unfortunately, such quantitativeanalyses are either unavailable, or too complicated to play a rolein public debates.

Concerning social philosophy and objectivity, I don’t thinkeconomics is capable of being an objective science. It is way toocomplex in structure and too grand in scale to be analyzeddefinitively with mathematics. We must use our philosophy and othergeneral opinions as guidance. However, in understanding theunderlying basic principles and mechanisms, it is important to beobjective, without the discrimination resulted from one’s ownphilosophical predispositions. Therefore, I think for an economist,philosophical penchant or even passion is a good thing to have. Onthe other hand, as a student, one should be open-minded to allpolitical views and learn from them as much as possible.

【网友】

I entirely agree with your argument onquantitative rather than qualitative argument. Actually this is apoint I always stressed. True, in economics we don’t have a verygood predicting power and our computational ability is so poorcompared to our tasks. But still, some evidence is better than noevidence. No matter how incomplete evidence we have,”evidence-based” policy making is better than “faith-based” policymaking. I always support technocracy in terms of making policies.Public debates? I think educating the public to be comfortable inevidence and numbers should come first. Without this preparation,public debate won’t go anywhere.

Just that your minimum wage example is not very appropriate. Ithink most economists are against minimum wage, at least inprinciple. I am of course one of them. The reason that some of themsupport minimum wage is that its effect on unemployment ispractically insignificant (many empirical papers support this) oras the pareto optimal result (no minimum wage and onlyredistribution to the people who are working) is not politicallyviable, so minimum wage should be imposed as a 2nd best (PaulKrugman’s view). Here I have to pay a tribute to Mr. Friedman. Hisidea of negative income tax is a good step to make the first bestpareto optimal result politically viable. I don’t know enough aboutthe negative income tax practice in US to see whether it canovercome the minimum wage problem, so I am indifferent in thispractical matter, and in principle against minimum wage.

As the mechanism in the labor market is not very suitable for aequilibrium study (price increases hence demand false). Labormarket is far more complicated. The standard approach is africtional search model. And unemployment is due to the friction ofsearching. Hence minimum wage affect unemployment only via changingthe distribution of draws.

But I don’t agree with you that personal philosophy should play animportant role in economics research. And it doesn’t in reality, inparticularly in theory and theory-based empirical research. Most ofthe research in economics are initiated because ofabnormality–inconsistence b/w theory and observations. Andactually, no matter what your political view is, in order to have aresearch accepted by the professional community, you must hold itup to the acceptable scholastic standard–to be rigorous aswhatever the standard of rigor your profession defines. Simplybecause of this (and I think economics is very a very competitivefield and peer review standard is very high–it’s difficult to geta paper publish compared to many other field. Very few fresh mintPhD can have publication in any top 10 journal), the influence ofpersonal world view is very faint. I agree that some people can lettheir works to suit/support their political view–but to do thiswith wide acceptance, but to do this requires extreme level ofingenuity. Very few people can do that–maybe real business cyclegangs (e.g. Ed Prescott) are the only examples I can have in mind.Anyone slightly less than the smartest economists simply don’t havethat ability. Think of economists from the well-funded conservativethink tanks like AEI/Heritage/Cato, their desire of twistedresearch to support their political ends are enormous, but theyhardly got any success. They are not even taken seriously. So yousee embedding politics in economics research is so difficult.

As for the people doing empirical research, it’s true many reducedform (not theory-based) research are inspired by personalphilosophy, but once the project is started, twisting the data iseven more difficult than twisting the theory.

As for whether we “need” rather than we “have” politics/philosophyin doing economics research, my opinion is that we DON’T need it.Because economics as a “useful” subject is still in its infancy.The most important role in research is knowledge accumulation. Notpolicy making. We need to make policy simply because we are”supposed” to–and giving out policy recommendation might be theonly way we can get funding to support our goal of knowledgeaccumulation. We economists should know that and never havecomplacency of our own conclusion of policy recommendation–we knowwhat a shaking ground on which it builds. And again, it comes backto insisting ad-hoc, case-specific, expedient, “evidence-based”policy making process.

Thomas Paine: Enlightenment, Revolution, and theBirth of Modern Nations by Craig Nelson

自以为对美国的开国元勋们有点兴趣，居然以前没有听说过这个ThomasPaine，真是汗颜。这本传记写得不是很好，主要是时间顺序上有点乱，对于不熟悉那段历史的人，不是很容易follow。但是另一方面，书中对18世纪末美国和法国的历史背景还是介绍得比较详细。

看来，Thomas Paine是个“愤青”类的人物。与其他开国元勋相比，他没有显赫的教育和家庭背景，可以说是流落到新大陆的。但在1776年，他的“常识”一书洛阳纸贵，对传播独立的主张起了巨大作用。在随后的独立战争中，他又发表了“危机”一书，极大地鼓舞了士兵和全国人民的士气。

但是，独立战争胜利后，Paine却生计无着，不得不多次向国会情愿救济。最后由华盛顿和几位大老出面游说各州赞助，才给了他一个安身之处。他又花了几年研究和推销一种新型桥梁结构，并因此回到欧洲。

Paine在大革命前夜的法国受到热烈欢迎，并继而在英，法煽动推翻君主制的革命，发表了又一本畅销书：“人的权利”。因为这个政治观点，他在英 国受到迫害。法国革命成功后，他被邀请加入立法会（国会）。但是不久，他因反对过度暴力，特别是对路易十六被处死不满而与革命党决裂，一度被捕入狱，差点丢了性命。在此同时，他发表了“理性时代”一书，挑战当时的宗教观念，更是树敌无数。

后来他又一度寄希望于拿破仑，特别期望是他的军队能把英国“从专制中解放出来”。但当拿破仑自己转向专制时，Paine又一次失望了。

最后Paine回到了美国，但昔日光环不再，政客们（以前的战友）视其为票房毒药。最后他郁郁而终，只有六个人来悼念。报上的讣告说“他作了一些好事，也作了很多坏事”。

总而言之，Paine是个有天分有野心的人，涉猎甚广，但EQ不高，对得罪盟友没有丝毫犹豫，还经常与出版商等合作者闹翻。但同时他又自恃甚高，认为基于往日功劳他entitle to特别待遇。他有不少辉煌时刻，但最后还是被排除在历史主流之外。这对这样一个非凡天才来说，真是一个悲剧。

看看Paine，也许对我们关注的某种网上现象会有一些帮助。

http://www.time.com/time/magazine/article/0,9171,15...

摘要：

本文主要是两个科学家关于上帝和科学之关系的辩论。一个是RichardDawkings（以下简称达），牛津大学讲座教授，“自私的基因”的作者。他最近的新书“TheGod Delusion”是五周的纽约时报畅销书榜首。他认为现代科学终将能解释所有疑惑，在人类知识中没有上帝的地位。他特别以对进化论的宣传，而以进化心理学解释道德而闻名。另一个是FrancisCollins（以下简称柯），1993年以来国家人类基因研究所（National HumanGenome ResearchInstitute）的主任，支持人类基因排序的工作。同为遗传学家，柯却坦承是基督徒，并主张宗教信仰与科学不矛盾。他的著作“TheLanguage of God: A Scientist Presents Evidence forBelief”是今年夏天的畅销书。

1．有没有上帝？

达：这是科学必须回答的问题。答案是否定。

柯：上帝是否存在是超出科学以外的问题。

2．科学与宗教是否可以并存，互不干扰？

达：这是空洞的想法。宗教不断地侵蚀科学的地盘。比如关于奇迹（耶稣复活，玛丽亚受孕等）是不仅与科学的事实不符，更违反科学的精神。

柯：把科学与宗教分割开是不对的。我相信万物是上帝创造的。对自然的研究使我有机会观察上帝的全能。

3．进化论与创世纪是否矛盾？

达：是。进化论解释了万物的精致，不需要创造者。

柯：上帝的能力和智慧超过我们的理解。他也许故意把世界造成看来是进化而来的。

4．如何理解六个基本宇宙常数恰好是生命所需之值：

达：我们不需要上帝来解释这个。可能我们能找到一个理论，指出他们不是独立的常数，而是联系在一起的。也可能有无数个宇宙，而我们的恰好得到这些常数。

柯。如果不能找到统一的理论的话，“人择原理”实在不是一个好理论。用上帝来解释更符合思维简约的原则。

达：搬出上帝只是逃避了追求真理的责任。科学家不应该如此。

柯：科学家的确应该继续寻找多个宇宙的证据，或其他解释的途径。但是我们不应该排除上帝的可能。

达：答案也许是上帝，但是可能是各种各样的上帝。说上帝就是圣经中的那个，可能性也是很小。

5．创世纪与大爆炸理论

柯：创世纪不应被严格从字面解释。它不是科学教科书，而只是对上帝以及我们与上帝的关系的描述。所以，如果退出狭义的解释的话，创世纪与大爆炸理论是一致的。

达：柯不应该去理睬那些原旨主义者。

6．基督教中的奇迹是否与自然规律矛盾？

柯：如果你相信上帝存在，那么上帝就可以在特别重要的时刻做些违背自然规律的事。而调查这些奇迹正是科学与宗教相接处的地方。

达：奇迹的说法就是否定了科学的进步。我们今天的技术在古代都会被当作奇迹。

7．人的道德观是不是从上帝来，而且是上帝存在的证明？

柯：说道德由进化产生是站不住的：利他行为不利于基因的生存。

达：但是在古代小群人共存的情况下并非如此。

8. 是不是应该基于圣经而反对干细胞研究？

柯：很多坚定信教的人并不反对干细胞研究。信仰应该建立在理性的基础上。但是科学并不拥有所有的理性。而且信仰的真理在人身上表现出来时，也已受到人性的限制。

达：对我来说，干细胞没有神经系统，没有痛苦。说它们是人要受到保护，是道德的绝对化。

最后，我把他们的总结全文译出：

柯：我只想说，在超过自分之一世纪的科学和信仰生活中，我完全能同意达对于自然世界的所有结论。同时，我也能接受这样的可能：对于自然世界，科学也许不能提供所有的答案，尤其是如果问题不是“什么”，而是“为什么”。我想知道为什么。我在信仰的范围内找到许多答案。这个绝对不影响我作为一个科学家的严格思维。

达：我的脑筋并不像你说的那样封闭。我愿意接受各种各样的可能性，那些都是你我今天不能想象的。但是我很怀疑，未来科学发现的真理竟然是人们自古以来就有的某种宗教。在我们讨论宇宙起源和宇宙常数时，我认为我已经提供了反对超自然的“设计者”的有力论据。但是，那（设计者）的想法对我来说还是值得考虑的。虽然可以驳斥，但是它还是值得认真对待。但是我不认为奥林比亚的神或者耶稣降生值得认真考虑。对我来说那只是说教。如果上帝存在，它一定比任何现有的宗教更为宏大和不可认识。

通常人们都认为人的道德是教育的结果：法律，宗教和父母教化让我们懂得区别是非。但是有些灵长目动物学家发现，人的道德观德基础，如设身处地和回报等概念，在其他群居动物中也有表现。据此，哈佛生物学家MarcD.Hauser提出一个假设：人们生来就有道德的“基本语法”。这是通过进化过程植入大脑结构的。这个结构是人可以下意识地，迅速地做出道德抉择。

从人的行为中也可找出支持此论点的根据。我们知道，来自不同宗教，文化和传统的人，基本的道德准则是相同的，尽管对具体事（如人工流产）的态度会有不同。Hauser借用语言学的概念，认为人脑中固有一个“道德”语法系统。它支持和约束人的行为，但不产生具体的规则。

对幼儿的研究表明，从4岁起人就开始产生公平的概念。

另一个有趣的实验是所谓“列车两难”问题。一辆列车驶来，但轨道前方有5个人将被撞到。你可以搬动道岔把列车引向另一条轨道，那里它会撞到1个人。你是否应该牺牲那1个人来救这5个人？大多数人的回答是肯定的。现在假定你可以把一样东西扔到轨道上来使列车停住。但那样东西是一个人。你是否应该牺牲那1个人来救这5个人？大多数人的回答却是否定的。Hauser认为，这两种情况的得益和代价是相同的，但人们的选择不同。这是因为“预见的伤害”（前一种情况）和“故意的伤害”（后一种情况）在人的道德考虑中是有区别的。

但是，大多数人并不能清楚地表达这种概念的区别。所以，Hauser认为，这种区别不是通过教育得到的（因为它不能用语言表达），而是先天的。[注：这个逻辑有问题。不可表达的理念可以通过行为表率来传播和教育.]

这个“道德语法”的概念是Hauser研究动物交流是产生出来的，并得到语言学家NormChomsky的合作。[吴礼注：这就是大名鼎鼎的乔姆斯基。]他们提出，支持语言的大脑结构在动物中已有存在，可能是用于导航。他还把一些用于幼儿心理研究得手段用于动物。他认为，群居动物往往有能力辨认不道德的行为如欺骗或者违反群体的规矩。但是它们不像人类那样记住坏行为，估计其损害性，并给予惩罚。Hauser假定在道德基础上的合作会给一个群体带来生存优势，所以这种品质会通过进化保存下来。

这个道德语法”的论点目前还没有得到很多事实的证明。但是它可能对我们的社会学，哲学等领域有深远的影响。

科学可以解释一切吗？

【博主评：

温伯格的文章是有很多深意。但是只读他的（或其他人的）文章，和真正去做物理，体会是完全不同的。我觉得温伯格是代表了做基本粒子物理和宇宙论的一部分物理学家的哲学观。物理中其他领域的人看法也可能很不同。

就拿凝聚态物理来说吧。你可以说量子力学已经“解释”了所有现象，剩下的只是计算，而不是“基本”的发现。但是凝聚态物理学家会认为有一些定律并不比量子力学不“基本”。这是因为这些定律虽然在理论上可能可以由量子力学“推导”出来，但事实上离开了实验是根本不可能发现的。

我也板门弄斧，提一个替代的真理观。我们了解世界就像拼图一样。我们的观察（以及从观察中总结出的定律）是一块块碎片。拼在一起，知道他们的相互关系，就是科学的任务。有时候，我们可以从已经拼好的部分推论出新的一块的位置。我们就说这新的一块没有其他的“基本”。但是这和我们的拼法和次序有关，并不见得是自然本身决定的。如温伯格说的广义相对论与超弦论谁更基本的问题，可能就是一个伪问题：可能它们都是一个更大的理论的一部分。它们之间的可容性只有在这个更大的理论的框架下才能理解。

顺便提一下，惠勒（John ArchibaldWheeler）提出过另一个有趣的认识论。他说，我们所理解的科学研究过程可以由以下游戏来模拟：一群人坐在一间屋子里，约定好一个物品。另一个人走进来，要猜出那是什么。他挨个儿问这些人问题，回答是“是”或“否”。如果他的问题设计得好，就能成功地猜到。

但是，惠勒说，也许游戏其实不是如此。哪些人并没有约定物品。他们只是随机地提供回答，唯一的限制是不能与以前的回答相矛盾。对猜的人来说，他并看不出两个游戏有什么不同。他最后也能得到一个答案。但是这个答案不是预先存在的，而是取决于他所问的问题。

惠勒认为，科学研究也是一样。我们的研究活动都是与自然的相互作用，实际上同时也改变着自然。所以我们得到什么结果，取决于我们研究的方式。当然这是我很粗浅的理解，也有着明显的漏洞。有兴趣的话你自己去看惠勒的书吧。

】

【因长度限制，原文的最后部分被删掉。原文可在http://www.nybooks.com/articles/article-preview?article_id=14263找到。】

Can Science Explain Everything? Anything?

By Steven Weinberg

One evening a few years ago I was with some other faculty membersat the University of Texas, telling a group of undergraduates aboutwork in our respective disciplines. I outlined the great progresswe physicists had made in explaining what was known experimentallyabout elementary particles and fields—how when I was a student Ihad to learn a large variety of miscellaneous facts aboutparticles, forces, and symmetries; how in the decade from themid-1960s to the mid-1970s all these odds and ends were explainedin what is now called the Standard Model of elementary particles;how we learned that these miscellaneous facts about particles andforces could be deduced mathematically from a few fairly simpleprinciples; and how a great collective Aha! then went out from thecommunity of physicists.

After my remarks, a faculty colleague (a scientist, but not aparticle physicist) commented, “Well, of course, you know sciencedoes not really explain things—it just describes them.” I hadheard this remark before, but now it took me aback, because I hadthought that we had been doing a pretty good job of explaining theobserved properties of elementary particles and forces, not justdescribing them.[1]

I think that my colleague’s remark may have come from a kind ofpositivistic angst that was widespread among philosophers ofscience in the period between the world wars. Ludwig Wittgensteinfamously remarked that “at the basis of the whole modern view ofthe world lies the illusion that the so-called laws of nature arethe explanations of natural phenomena.”

It might be supposed that something is explained when we find itscause, but an influential 1913 paper by Bertrand Russell had arguedthat “the word ’cause’ is so inextricably bound up with misleadingassociations as to make its complete extrusion from thephilosophical vocabulary desirable.”[2] This left philosophers likeWittgenstein with only one candidate for a distinction betweenexplanation and description, one that is teleological, defining anexplanation as a statement of the purpose of the thingexplained.

E.M. Forster’s novel Where Angels Fear to Tread gives a goodexample of teleology making the difference between description andexplanation. Philip is trying to find out why his friend Carolinehelped to bring about a marriage between Philip’s sister and ayoung Italian man of whom Philip’s family disapproves. AfterCaroline reports all the conversations she had with Philip’ssister, Philip says, “What you have given me is a description, notan explanation.” Everyone knows what Philip means by this—inasking for an explanation, he wants to learn Caroline’s purposes.There is no purpose revealed in the laws of nature, and not knowingany other way of distinguishing description and explanation,Wittgenstein and my friend had concluded that these laws could notbe explanations. Perhaps some of those who say that sciencedescribes but does not explain mean also to compare scienceunfavorably with theology, which they imagine to explain things byreference to some sort of divine purpose, a task declined byscience.

This mode of reasoning seems to me wrong not only substantively,but als procedurally. It is not the job of philosophers or anyoneelse to dictate meanings o words different from the meanings ingeneral use. Rather than argue that scientists ar incorrect whenthey say, as they commonly do, that they are explaining things whethey do their work, philosophers who care about the meaning ofexplanation i science should try to understand what it is thatscientists are doing when they say the are explaining something. IfI had to give an a priori definition of explanation i physics Iwould say, “Explanation in physics is what physicists have donewhen the say Aha!” But a priori definitions (including this one)are not much use.

As far as I can tell, this has become well understood byphilosophers of science at least since World War II. There is alarge modern literature on the nature of explanation, byphilosophers like Peter Achinstein, Carl Hempel, Philip Kitcher,and Wesley Salmon. From what I have read in this literature, Igather that philosophers are now going about this the right way:they are trying to develop an answer to the question “What is itthat scientists do when they explain something?” by looking at whatscientists are actually doing when they say they are explainingsomething.

Scientists who do pure rather than applied research commonly tellthe public and funding agencies that their mission is theexplanation of something or other, so the task of clarify-ing thenature of explanation can be pretty important to them, as well asto philosophers. This task seems to me to be a bit easier inphysics (and chemistry) than in other sciences, becausephilosophers of science have had trouble with the question of whatis meant by an explanation of an event (note Wittgenstein’sreference to “natural phenomena”) while physicists are interestedin the explanation of regularities, of physical principles, ratherthan of individual events.

Biologists, meteorologists, historians, and so on are concernedwith the causes of individual events, such as the extinction of thedinosaurs, the blizzard of 1888, the French Revolution, etc., whilea physicist only becomes interested in an event, like the foggingof Becquerel’s photographic plates that in 1897 were left in thevicinity of a salt of uranium, when the event reveals a regularityof nature, such as the instability of the uranium atom. PhilipKitcher has tried to revive the idea that the way to explain anevent is by reference to its cause, but which of the infinitenumber of things that could affect an event should be regarded asits cause?[3]

Within the limited context of physics, I think one can give ananswer of sorts to the problem of distinguishing explanation frommere description, which captures what physicists mean when they saythat they have explained some regularity. The answer is that weexplain a physical principle when we show that it can be deducedfrom a more fundamental physical principle. Unfortunately, toparaphrase something that Mary McCarthy once said about a book byLillian Hellman, every word in this definition has a questionablemeaning, including “we” and “a.” But here I will focus on the threewords that I think present the greatest difficulties: the words”fundamental,” “deduced,” and “principle.”

The troublesome word “fundamental” can’t be left out of thisdefinition, becaus deduction itself doesn’t carry a sense ofdirection; it often works both ways. The bes example I know isprovided by the relation between the laws of Newton and the law ofKep-ler. Everyone knows that Newton discovered not only a law thatsays th force of gravity decreases with the inverse square of thedistance, but also a law o motion that tells how bodies move underthe influence of any sort of force. Somewha earlier, Kepler haddescribed three laws of planetary motion: planets move on ellipsewith the sun at the focus; the line from the sun to any planetsweeps over equal area in equal times; and the squares of theperiods (the times it takes the various planets t go around theirorbits) are proportional to the cubes of the major diameters of thplanets’ orbits

It is usual to say that Newton’s laws explain Kepler’s. Buthistorically Newton’s law of gravitation was deduced from Kepler’slaws of planetary motion. Edmund Halley, Christopher Wren, andRobert Hooke all used Kepler’s relation between the squares of theperiods and the cubes of the diameters (taking the orbits ascircles) to deduce an inverse square law of gravitation, and thenNewton extended the argument to elliptical orbits. Today, ofcourse, when you study mechanics you learn to deduce Kepler’s lawsfrom Newton’s laws, not vice versa. We have a deep sense thatNewton’s laws are more fundamental than Kepler’s laws, and it is inthat sense that Newton’s laws explain Kepler’s laws rather than theother way around. But it’s not easy to put a precise meaning to theidea that one physical principle is more fundamental thananother.

It is tempting to say that more fundamental means morecomprehensive. Perhaps the best-known attempt to capture themeaning that scientists give to explanation was that of CarlHempel. In his well-known 1948 article written with Paul Oppenheim,he remarked that “the explanation of a general regularity consistsin subsuming it under another more comprehensive regularity, undera more general law.”[4] But this doesn’t remove the difficulty. Onemight say for instance that Newton’s laws govern not only themotions of planets but also the tides on Earth, the falling offruits from trees, and so on, while Kepler’s laws deal with themore limited context of planetary motions. But that isn’t strictlytrue. Kepler’s laws, to the extent that classical mechanics appliesat all, also govern the motion of electrons around the nucleus,where gravity is irrelevant. So there is a sense in which Kepler’slaws have a generality that Newton’s laws don’t have. Yet it wouldfeel absurd to say that Kepler’s laws explain Newton’s, whileeveryone (except perhaps a philosophical purist) is comfortablewith the statement that Newton’s laws explain Kepler’s.

This example of Newton’s and Kep-ler’s laws is a bit artificial,because there is no real doubt about which is the explanation ofthe other. In other cases the question of what explains what ismore difficult, and more important. Here is an example. Whenquantum mechanics is applied to Einstein’s general theory ofrelativity one finds that the energy and momentum in agravitational field come in bundles known as gravitons, particlesthat have zero mass, like the particle of light, the photon, buthave a spin equal to two (that is, twice the spin of the photon).On the other hand, it has been shown that any particle whose massis zero and whose spin is equal to two will behave just the waythat gravitons do in general relativity, and that the exchange ofthese gravitons will produce just the gravitational effects thatare predicted by general relativity. Further, it is a generalprediction of string theory that there must exist particles of masszero and spin two. So is the existence of the graviton explained bythe general theory of relativity, or is the general theory ofrelativity explained by the existence of the graviton? We don’tknow. On the answer to this question hinges a choice of our visionof the future of physics—will it be based on space-time geometry,as in general relativity, or on some theory like string theory thatpredicts the existence of gravitons?

The idea of explanation as deduction also runs into trouble when weconside physical principles that seem to transcend the principlesfrom which they have bee deduced. This is especially true ofthermodynamics, the science of heat an temperature and entropy.After the laws of thermodynamics had been formulated i thenineteenth century, Ludwig Boltzmann succeeded in deducing theselaws fro statistical mechanics, the physics of macroscopic samplesof matter that are compose of large numbers of individualmolecules. Boltzmann’s explanation o thermodynamics in terms ofstatistical mechanics became widely accepted, eve though it wasresisted by Max Planck, Ernst Zermelo, and a few other physicistswh held on to the older view of the laws of thermodynamics asfree-standing physica principles, as fundamental as any others. Butthen the work of Jacob Bekenstein an Stephen Hawking in thetwentieth century showed that thermodynamics also applie to blackholes, and not because they are composed of many molecules, butsimpl because they have a surface from which no particle or lightray can ever emerge. S thermodynamics seems to transcend thestatistical mechanics of many-body system from which it wasoriginally deduced

Nevertheless, I would argue that there is a sense in which the lawsof thermodynamics are not as fundamental as the principles ofgeneral relativity or the Standard Model of elementary particles.It is important here to distinguish two different aspects ofthermodynamics. On one hand, thermodynamics is a formal system thatallows us to deduce interesting consequences from a few simplelaws, wherever those laws apply. The laws apply to black holes,they apply to steam boilers, and to many other systems. But theydon’t apply everywhere. Thermodynamics would have no meaning ifapplied to a single atom. To find out whether the laws ofthermodynamics apply to a particular physical system, you have toask whether the laws of thermodynamics can be deduced from what youknow about that system. Sometimes they can, sometimes they can’t.Thermodynamics itself is never the explanation of anything—youalways have to ask why thermodynamics applies to whatever systemyou are studying, and you do this by deducing the laws ofthermodynamics from whatever more fundamental principles happen tobe relevant to that system.

In this respect, I don’t see much difference between thermodynamicsand Euclidean geometry. After all, Euclidean geometry applies in anastonishing variety of contexts. If three people agree that eachone will measure the angle between the lines of sight to the othertwo, and then they get together and add up those angles, the sumwill be 180 degrees. And you will get the same 180-degree resultfor the sum of the angles of a triangle made of steel bars or ofpencil lines on a piece of paper. So it may seem that geometry ismore fundamental than optics or mechanics. But Euclidean geometryis a formal system of inference based on postulates that may or maynot apply in a given situation. As we learned from Einstein’sgeneral theory of relativity, the Euclidean system does not applyin gravitational fields, though it is a very good approximation inthe relatively weak gravitational field of the earth in which itwas developed by Euclid. When we use Euclidean geometry to explainanything in nature we are tacitly relying on general relativity toexplain why Euclidean geometry applies in the case at hand.

In talking about deduction, we run into another problem: Who is itthat is doing th deducing? We often say that something is explainedby something else without ou actually being able to deduce it. Forexample, after the development of quantu mechanics in themid-1920s, when it became possible to calculate for the first timein clear and understandable way the spectrum of the hydrogen atomand the bindin energy of hydrogen, many physicists immediatelyconcluded that all of chemistry i explained by quantum mechanicsand the principle of electrostatic attraction betwee electrons andatomic nuclei. Physicists like Paul Dirac proclaimed that now all ochemistry had become understood. But they had not yet succeeded indeducing th chemical properties of any molecules except thesimplest hydrogen molecule Physicists were sure that all thesechemical properties were consequences of the law of quantummechanics as applied to nuclei and electrons

(the rest is deleted due to length limitation)

闲谈：谈话的目的不是交流信息，而是表达双方的goodwill。这时就根本不应该较真。如果对方说“今天天气更好”，你说“看你怎么定义今天。早上一点至两点之间曾下过阵雨”，那就不好玩了。　

教育：谈话的目的是把自己的知识传给对方。那么是不是较真就要看对方的情况。既要足够较真免得对方误解，又不能太较真反而把对方弄糊涂了。很多人犯错误就在这一点，特别是教孩子时。有时追求概念绝对严格，绝对全面，反而把对方turnoff。

显摆：谈话的目的是显示自己的知识，比如在面谈时。在社交场合也有这种需要。这时是否较真，就看自己的感觉和需要了。

抬杠：谈话的目的就是看谁赢得争论，真正的是非倒不是主要的。这时不但要较真，而且还可以强辩，狡辩。其中的乐趣就在于思辩练习。

以上几种沟通模式都是可以接受的社交行为。但是如果双方处于不同的模式，就会造成问题。最明显的是一方在“抬杠”模式，而另一方在其他模式。

另外，“教育”和“显摆”之间也常会混淆。比如一方说：“这个很简单，你自己去查一下吧。”这是在显摆：我是知道答案的，但我没有兴趣帮助你知道。如果对方是在“教育”模式，确实想学习，就会有挫折感了。有一次我访问一个教授，他对我解释他的研究。我听了想用自己的话重复一遍。我开始说“Sowhat you did is just…”谁知他不等我说完就勃然大怒，差点拂袖而去。我连忙解释我用just这个词并没有贬低他的意思，才算把谈话继续下去。现在想来，其中问题就在于沟通模式不同。我是在“教育”模式，把确认自己的理解正确看得比showrespect更重要。而教授是在“显摆”模式，而认为我的反应一定是评价，而不是讨论。

而且，模式在沟通过程中还会变。经常是在教育和显摆之间。如果本来的目的是“教育”对方，而演变成了“显摆”，那就叫”carriedaway”，而对方就难免一头雾水了。

所以要有成功的沟通，一定要对当前的模式有敏感的理解。把握“较真”与“装闷”的尺度，关键也在于此。

第一遍读时，感觉全是胡说八道，根本没有的事编了那么多。

再读一遍时，被那些武功情节深深吸引了，倒忘了那是假的了。读完后深深佩服金庸的想像力。

第三遍读时，觉得里面有很多人生哲理和世故经验。特别喜欢的是《天龙八部》中的佛学和《笑傲江湖》中的孤独九剑和“正”，“邪”之间的转化。《倚天屠龙记》中的人情也令我感动。那些打打杀杀的场面反倒跳过去了。

第四遍再读，发觉其实就是报上连载的小说，写得引人入胜，还是为了吸引眼球，而不是什么微言大义。　

转了一圈，还是胡说八道。这也符合金庸风格吧。　

以下是我与网友的一个讨论

【网友A】

对于技术密集型产业，譬如软件产品的开发，人力资源的因素是第一位的。在有限资源的条件下，怎样组建一支团队很有学问。见过两种组建团队的方法，高下立判。

假设软件行业熟练工平均工资是10万，那么200万一年的投入，可以组建一支20人的队伍（实际情况当然不是这样，人头费上还有各种福利和保险的overhead,要另加50％－100％）。一种策略是招收20个平平的员工，看不到杰出人才（A++talents）与普通人才的质的差别。由于工资仅仅达到市场价格，没有市场优势，同时也没有优秀人才特有的凝聚力和向心力，人员变动频繁，使得overhead增加，工作连续性不能保障，精打细算，结果反而捉襟见肘。很多教授办公司，往往使的是这招，所谓产品，也多是政府客户的一些合同，比较好糊弄，这种方式居然有时也可以长期运行下去，但是要想真正做大占领工业市场，基本肯定没戏。

另一个策略是，抓两头，杜绝中庸。所谓两头，就是关键的技术岗位一定要找顶尖人才挂帅，关键岗位包括，核心技术引擎的架构师，产品设计师，用户界面开发师，质量检测师，等等。其实，只要产品是朝阳项目，吸引顶尖人才的花费平均需要高出中庸人才的待遇大约50%，就是说大约15万左右就可以了。在顶尖人才的领导下，不妨在每个部门招收一些new graduates，把招揽顶尖人才所耗费的钱省回来。如果顶尖人才与newgraduates的比例是1:2, 那么平均一个 new graduate的花费大约7.5万，这个价格在工作市场是具有竞争力的，候选人的基数大，可以在新手中挑选最聪明最有潜力的那种。这些新手工作机会难得，特别好学，有顶尖人才做导师引路，learning curve容易克服。对于那些三个月半年还是不能上路的（如果挑选严格的话，不能workout的新手可以控制在1／3以内，就是说，三个人有一个后来发现不能成为有实质贡献者），绝对不要犹豫，开掉了事。因为开的还是基本在见习期的新手，对于公司没有什么副作用。这是优胜劣汰的健康的人员流动。

必须承认，人的才智和创造力相差很大。从投入产出效益看，对顶尖人才多投入的50%，所获得的往往是两倍至数倍，是绝对合算的交易。一个聪明的管理者，应该善于慧眼选拔顶尖人才并建立顶尖人才吸引顶尖人才，顶尖人才组建自己队伍的良性机制。这样的管理和如此建立的团队，才是有希望的。

【我】

这要看企业发展的阶段  

在产品推出初期，性能还达不到客户要求，市场也比较小。这时候技术上的进展就非常重要，而且大多数技术进展是自下而上的，也即由技术人员启动的。这个阶段，公司的组织，重点等都是很动态的，对人员素质的要求就比较高。另一方面，由于规模比较小，主管者能够掌控每个成员的表现和水平。在这个阶段，花大钱雇高手是好办法。

在产品成熟之后，情况则相反。这时产品性能已经不是主要问题，而销售和支持，成本等成为竞争成败的主要原因。在内部，由于团队变大了，不可能继续“人盯人”的管理方法，而要借助流程，政策等死板的机制。同时，技术工作不再由发明创造驱动，而是由营销部门根据客户要求提出任务，而开发部门是肯定能完成的（技术上挑战不大，只是速度和质量的问题）。由于公司注重长远运作，需要保证个别成员的离开不会影响整个项目。这些因素都使得高手和庸人的价值差别减小。这个阶段，雇庸人是应该的。公司的取胜主要靠经营，而不是靠技术团队。

雇新毕业生还有一个考虑，就是对企业文化的影响。如果你想塑造自己的文化，雇新手是好办法，因为他们可塑性最强。如果你对自己的文化没有vision，可以雇有经验的，从融合其他公司的文化开始。

新手中“最聪明最有潜力的那种”，你要考虑公司能不能为他们提供长期的成长机会。这种人一般野心也大。开始时进步很快，不久就会要求新的机会。如果必须要离开公司才能继续发展，那你就是替他人做嫁衣裳了。

【网友 B】

不完全同意   

你的说法适用于公司打算这辈子就搞一个系列产品的特例。

而现代社会里面，一家公司要想永远立于不败之地，需要持续的创新能力。

否则，一个产品周期完成了，公司也完蛋了。

【我】

诚然。但是大公司“持续创新”谈何容易。   

我经历的公司中，能靠买小公司来“创新”成功就算很不错了。土生土长的“创新”项目往往是不了了之的。

问题是：大公司的“创新”往往是为了“成长”（growth)。只有持续成长，才能维持高的PEratio，而这才是最终目的。但是要对“成长”有显著贡献，创新的产品就必须有大的市场，而不是边缘市场。这样大市场的产品，要末是市场已经形成，竞争也已经激烈。这时进入，就是在“追赶”模式，并没有高手什么事，不如雇一帮肯加班的学生。或者是市场还不成熟，那风险就很大，不光是技术，还有市场。象MiMax，号称是集所有先进技术之大成，这么多年修改了多次还是打不开市场，连Sprint，Intel这样的大腕也一筹莫展。通常高级主管很少有眼光和自信来豪赌一把。往往是干了一阵就放弃了。

只有象往日的AT&T和今天的Microsoft这样地位的公司才有“持续创新”的本钱。但象这样的公司，实在是绝无仅有啊。

【网友 B】

看你怎么定义高手了   

比如，你进入已经成熟的市场，要追赶，其实这个时候是很需要高手的。

因为你要在成熟市场分一杯，高明的战略就是在现有产品的基础上做一点

小的改进，用differentiation战略来取得竞争优势。而这个小改进，

靠庸才是不行的，雇肯加班的学生也不行。

小日本几乎都是在别人发明的技术基础上做小改进来赢得市场的，你不能说

小日本的公司里面不需要技术高手。

【我】

这倒也是。不过这种“小的改进”通常难的不是实施，而是立项 

要知道怎样的改进对客户的价值最大。一般这些立项都是行销的人主导的。技术部门只是实施，并没有太大挑战性。靠技术发明主导相当于“拿着扳手找螺丝”，往往反而坏事。

当然也有人说工程没有大的革命，都是小改进积累而成。问题是小改进通常是“人民”，而不是“英雄”创造的。

说到日本公司，我没有亲身经验。不过听说对技术高手的价值并不是很高估，大家干活待遇差不多，晋升也主要靠资历。

【网友 C】

之所以有“不思进取”的人，很多时候是management没有把incentive搞对。  

让人觉得干不干无所谓。

【我】

那也不是绝对的。  

有些人把工作当成事业做，有些人就是混碗饭吃，主要的追求在工作之外。不是说第二种人就不敬业，但是他们不愿在职业上放更多投资。

【网友 C】

当然不是绝对的了。  

这就是为什么我用的是“很多时候“，不是”都是“。：）

在一般公司里，普通职员干好干坏，只要没坏到“捅漏子“的地步，对职员来说真是没多大关系。

【我】

对。但是在小公司就有很大关系。一个高手可直接影响公司进项。  

所以在我看来，如果自认为是高手，就不要在大公司泡着，除非想当CEO。

【网友 C】

同意。   

但是，即使在小公司，高手有了大贡献而得不到回报的情况也有。这就是为什么我提到管理上要把incentive搞对了。

【我】

对。小公司也有可能不公平。但是大公司没法公平。 

因为人太多，不可能个个都是高手。所以必须有制度，流程等，保证庸手也不犯大错误。而这些又反过来束搏了高手的发挥空间，使得高手和庸手的价值差不多了。

后来就对火鸡敬而远之。直到很多年以后，学了一个烤火鸡的配方，又经过多年试验，才比较有信心在感恩节请客了。如今，别人感恩节请俺，都要指定“带火鸡来！”

言归正传，以下是我烤火鸡的方法。

1。前期准备。感恩节总是周四。但准备工作很早就要开始了。以下假定火鸡是二十到三十磅之间。更大的有关时间都要适当加长。

周一：如果是冰冻火鸡，取出冷冻室，放在冰箱里解冻。

周二晚上：把盐和花椒在锅里干炒成细粉末，涂在火鸡表面和肚内。把火鸡放回冰箱。

周四一大早：取出火鸡，洗净表面盐末，置于室温。

另配涂液：水加很多糖，以及其他调料（酱油，姜末，酒等），看口味而定。每十分钟在火鸡表面刷一次。

2．烤制（周四当天）

十点：火鸡放入烤盘，用铝箔包住后腿和翅膀的端部，以防烤焦。火鸡上用铝箔盖住如帐篷。开始烤（400度）。

十一点：改为350度。

同时做填料备用：糯米饭（偏干一些），加入胡桃，栗子，葱，姜，和其他调味品（酱油，蚝油等。不要加太多酒，因为不容易挥发掉）。

十二点：将填料放入火鸡肚内，用铝箔塞住出口。

下午两点：除去覆盖的铝箔，火鸡翻身，刷涂液。

以后每半小时翻身，刷涂液。

四点：测试是否烤熟（用筷子插入大腿肉厚处，如没有血水流出即为熟。）如果已熟，可根据口味决定是否停止。这时皮表面会有油脂渗出。烤的时间长，皮会比较好看，而且皮下脂肪都去除，肉也不腻。但是如烤得太长，皮会比较硬。如果不想吃皮，可以多烤一会。

烤完后，取出冷却。取出填料装盘。

六点：开始切块。

3。食用：

填料：挑出葱，姜等不能食用的部分，其余装盘上桌。

红肉：切成片状，装盘上桌。

白肉：留下，打成肉馅包饺子。

骨架：剔去肉后烧汤。