失败的价值

作者：Richard Zare
转自2006年5月《科学》

我为我的讲话选择了这个题目：“失败的价值（The Virtues of Failure）”，你们可能会奇怪：怎么今天我竟希望谈论失败呢?首先，失败对我是司空见惯的。第二，我相信创新性的研究是无数次失败和极少几次成功的混合。对科学界以外的许多人或刚开始做研究的学生们来说，这一事实也许并不那么显而易见。我们在报纸上和科学杂志上重视或强调的总是所获得的成就，其中文章给人的印象是，成功的次数大大超过了失败次数。这一错误观念又常常被一些口头讲话所加强：演讲者往往把研究工作讲得让人听起来像是不费什么力气的事，那不过是一步接一步的逻辑步骤罢了。然而上述印象实在是一种误导，每个研究者都知道这一点，真正的研究乃是一部由错误组成的喜剧，其中错误的事情一件接着一件发生。不妨用丘吉尔（Winston Churchill）的话来说，研究进展其实是怀着永不衰减的热情，在一个接着另一个失败的道路上蹒跚前行。

倘若研究真正具有创新性，那么关于什么将会发生或将被发现，事先实在是没有多少可以预测的。创新性研究并非在某种表格上填空白，或者在已经很好地确立的知识边界上作些拓广而已。当然这类活动也有它自己的地位，不过我不称它为创新性的。对任何开始做研究的人来说，要学到的最重要的教训之一就是：实验是要失败的，而且失败是常事。实验研究不是做实验室的教学实验——它已被设计成每次都能成功。不管你对一个实验研究所依据的理论理解得多好，也不管你的实验计划设计得多么完善，结果常常跟你最早想象的一点电不相同，实验科学乃是对未知的一种探索．所以预先的计划充其量不过是一种猜测而已，有时候，这些猜测被证明为完全想错了，而一系列实验数据看来完全未产生任何有意义的结果。实际上，伟大的发现总会使我们大吃一惊。没有什么惊异的话，便不们认识世界的方式。

我相信这些感受是许多其他领域所共有的。我注意到，虚构小说的作家常常说：书中的角色在作案后立刻驱使故事朝着非预期的方向发展。类似地，非虚构小说的作家也常常发现：一旦当新的事件突然冒出来时。他们原来正要讲的故事马上变得不一样了。

说服开始做研究的人接受失败，承受再用另一种方法去尝试风险，这很困难，但又是必要的。一个更加成熟的研究者知道：失败是创造过程的一部分。我发现：发展一种心态是有益的，我称之为“恬然自安的精神分裂症的（contented schizophrenic）”心态。抱此心态．你愿意相信某种东西，但同时又还不相信它。你在同一时刻既热情地相信，又批判地提问题，这看起来似乎是两种互相矛盾的活动，但这种心态恰恰有助于探索自然之谜。你必须提出你自己最好的思想——关于某些东西可能是什么样子，然后立刻开始设计方法去检验这一关于自然界行为的思想，看它是否欺骗了你自己，或使你看不到正在发生的事情。学会与模棱两可的思想共处，此乃一种可以学到手、甚至也可以教会别人的东西。

就其核心来说，科学是一种原创性的活动。我们似乎只可以去反驳看来已经确立的事物，但当经过彻底努力的反驳归于失败之后，我们才会逐渐接受：某种断言可能是成立的。这是一个从假说到理论再到科学定律的过程。而当我们变得更加聪明、更加深思熟虑之后，我们才能发现：即使有些我们最珍爱的定律，也仅在它们有效的范围内成立。实验研究工作常常是在一系列失败中穿插个别的成功。一个聪明人从每一次失败中学到东西，并且把失败作为引导我们走向真理的自然界固有的方式而加以接受。当然，人的本性会因一次次失败而感到沮丧；然而，从失败中学习．学会与失败共处，对于变为一个成功的研究者是最关键的两个因素。若没有几乎总是不断的失败，则难得的成功也就不会使我们感到那层甜美。“恬然自安的精神分裂症”，接受失败，把它当作产生真正有价值之创新的唯一道路。失败确有其独特价值，激励我们不断地去揭示自然界之谜。在研究的竞技场中，失败肯定是一个几乎从不离开你的伴侣，聪明的研究者利用失败作为一种强有力的、但常常是秘密的武器，去取得成功。

科学创新的每一步进展都要付出极大的努力，而且必须以无数次失败为代价。这一点需要社会各界，首先是家庭、亲友和同事的理解与支撑。我们需要“锦上添花”，但更需要的是“雪中送炭”。爱因斯坦在1934年说：“……当获得的（新）知识已经昭然若揭时，令人喜悦的成就看起来几乎只不过是一件理所当然的事情，任何一个聪明的学生都能不太费力地掌握它。然而，那多少年在黑暗中热切的寻觅、紧张的期待，信心不断交替起伏。以致筋疲力竭，直到最后光明出现——只有亲身经历过它的人才能理解。”


扎雷（Richard Zare, 1939—）美国斯坦福大学 Marguerite Blake Wilbur 自然科学讲座教授，共发表论文700多篇，并因其研究和教学而赢得了很高的荣誉，包括美国国家科学勋章，其研究工作具有很高的深度、广度和影响力。他以激光化学领域的创新研究而闻名于世，2005年获Wolf Prize。

　　一项由美国斯坦福大学和英国爱丁堡大学的化学家共同完成的新实验最终证明了啤酒爱好者们长期以来怀疑的现象:当啤酒被倒入玻璃杯中时，啤酒泡沫有时不是上浮而是下沉。

　　“泡沫比啤酒轻，所以人们通常认为它们一定是要往上浮起的，”斯坦福大学自然科学教授Richard N. Zare说，“不过无数喝酒的人都宣称啤酒泡沫实际上会沿着玻璃杯壁向下沉。难道他
们是正确的，或者说这一现象违背了物理学法则？”

　　在1999年澳大利亚的研究人员宣布他们创建了一个计算机模型表明啤酒泡沫下沉现象在理论上是可能的之后，啤酒泡沫问题的似乎到达了终点。澳大利亚研究人员是基于一杯英国黑啤酒的泡沫所建立的模型，这种爱尔兰人酿造的大众啤酒中既包含氮气，也有二氧化碳气体。

　　但Zare和他的前斯坦福大学博士后同伴、现爱丁堡大学教授亚历山大怀疑黑啤酒模型，并决定通过分析几升啤酒液体进行实验测试。

　　“事实上，亚历山大和我不相信这个模型，并认为是人们喝多了产生的幻觉，”Zare回忆道：“我们进行了自己的实验，非常有趣但非决定性的。于是亚历山大手持一架每秒可拍摄750幅图片的摄像机，详细记录了所发生的一切。”

　　对影像进行的详细分析验证了澳大利亚研究组的发现：啤酒泡沫能够而且确实下沉到玻璃杯底部。为什么会发生这种现象呢？

　　Zare解释说：“答案事实上非常简单，它基于浮上来的必须沉下去的道理。这啤酒泡沫现象中，由于玻璃杯壁的拉力，位于玻璃杯中心的啤酒泡沫比在旁边的更容易浮上来，当它们浮上来后，使啤酒水平高度上升了，必须进行回溢，在回溢过程中，啤酒液体将会携带着泡沫——特别是小泡沫——沿着杯壁往下流动，最后停下来。这相当有趣，也非常容易证明。”

　　亚历山大说，不仅仅是黑啤酒，其它不含氮气的啤酒都会发生这种现象，因为泡沫非常小所以能够被液体推动着下沉。他还表示，只要留神注意周围的世界，就能找出事情发生的原因。

　　（编译/菩提）

　

这篇文章是前几天在图书馆时翻中文版的<科学>时发现的，觉得很好，就请别人打印出来，转来客栈。

很好的文章内容！

我很早也提到过类似的内容，不过我所说的是以痛苦作为代表，也就是失败属于痛苦中的一部分。我说过痛苦对人所起得作用：躲开逃避威胁自己的事物。或者说通过累积痛苦经验使人走向更有利于自己的方向。

这文章也只提到个人，社会也存在这种机制。不过社会不叫痛苦或者失败，叫具有负面影响的行为。包括犯罪、战争等。这些的存在使社会自觉发现自身存在缺陷，并不断改善自身以达到发展的目的。特别是战争，本身具有竞争性，尽管负面影响很大，但对社会的变革具有很彻底的效果。

但并不是说这些负面的内容越多越好，毕竟是有负面影响，过多会导致灭亡。如何将正面和负面达到一定平衡的稳定状态，同时又能以最大的效率改善社会，才是应该做的。

同样对于个人，痛苦如果过多，要么积累到无法承受，要么变得麻木不仁，这对个人都不是一件好事。关键也是要达到一种动态的平衡状态。

顶！

：：这篇文章是前几天在图书馆时翻中文版的<科学>时发现的，觉得很好，就请别人打印出来，转来客栈。
===================================
网上有电子的，为什么要打印？

XXFF兄,哪里有?有就更好:)

回楼上：国内正规期刊一般同方，万方，维普等会收入电子版，搞个帐号就能DOWN全文。该文万方中就有。

呵呵，XXFF兄，我以为还有更好的地方:(,不过，有时在网上看文章还是不如直接去看书:-)

失败是成功的转机

断浪小妹的观点与中国古代思想的“Wisdom”是相通的：烦恼即菩提，在失败的种子里蕴育着转机的希望。

我们首先应该定义一下什么叫失败的实验。
我想得到一个良好的材料。我花了查了很多资料，做了很多工作。精心设计了一个实验，可结果得到的材料差的很。这是失败的实验吗？
当然不是。想做什么就能得到什么，那是上帝。我们凡人是没这么牛的。
所谓失败的实验是指无法得到任何结论的实验。比如实验中设备坏了，数据被人改了，仪器失灵了而我们不知道。只要实验有结论，哪怕是一个很不起眼的结论，比如证实了一个早已确立的理论，找到了一个错误的配方，都是成功的实验。
我们不应该把一个项目最后做的那个实验称为唯一正确的实验。

http://www.stanford.edu/group/Zarelab/zare/toulouse/thevirtuesoffailure.htm

The Virtues of Failure

Richard N. Zare (Department of Chemistry, Stanford University)

Thank you, Professor Vigué, for your most kind and generous introduction, so ample and magnificent that only my mother would have wished that it were longer. I am truly grateful to be invited to Université Paul Sabatier, this great and ancient university, to receive an honorary degree and to be part of these most special ceremonies.

Your President has asked me to address you in French despite my protestations that public humiliation should not be part of the festivities. Alas, my command of this language is poor, indeed, embarrassingly so. My knowledge of French is based on two high-school courses taken more than forty years ago in Cleveland, Ohio. The first course was taught by Mademoiselle Barr, who was young, vivacious, and most curvaceous. I remember that every boy in the class looked forward to Mademoiselle Barr, who delighted us in wearing tight-fitting blouses and spoke with enthusiasm about the time she once visited France and discovered the meaning of joie de vivre. The next year, my second class in French was taught by a much older man who was a strict disciplinarian. My interest in French declined markedly as this course progressed.

Those of you familiar with the United States will recognize that the Midwest is what we call linguistically challenged.  By this I mean that American English is spoken there almost exclusively, and frequently not well either. My studies of French allowed me to read Alexander Dumas' "The Three Musketeers", and to learn by heart the phrase, Le chat et le rat n sont pas amis.  This last phrase has proved to be useless to me. The occasions for using it have been until now meager to nonexistent! Sad to say, since that time, my abilities to speak French have deteriorated, largely through lack of use. I beg your indulgence while I proceed. I can only promise you as the English king Henry the Eighth said to each of his many wives, Our time together will be short. 

I have chosen for the title of my talk, The Virtues of Failure.  You might be surprised that I wish to speak about failure today. First, failure is no stranger to me. Second, I believe that innovative research is a mix of many failures and few successes. This fact may not be apparent to those of us outside science or to those students of the sciences who are just beginning. We read in newspapers and in scientific journals accounts that stress the accomplishments achieved. These articles give the impression that successes vastly outnumber failures. This false idea is often reinforced by oral presentations in which the speaker makes the research enterprise sound like it were effortless, consisting of one logical step after another. But this impression is misleading, as every researcher knows. Real research is a comedy of errors in which one thing goes wrong after another. To paraphrase Winston Churchill, research progress consists of staggering from one failure to the next with undiminished enthusiasm.

If research is truly innovative, then little can be predicted ahead of time about what will happen and what will be found out. Innovative research is not an activity of filling in the blanks in some form or in extending the boundaries of well-established knowledge. Such activity also has its place, but I do not call it innovative. One of the most important lessons to be learned by anyone beginning to do research is that experiments fail, and they do that rather regularly. Experimentation is not like a laboratory exercise that has been devised to work each time. No matter how well understood the theories upon which an experiment is based or how well designed the experimental plan, the results can often be nothing like what was first imagined. Experimental science delves into the unknown, so the planning beforehand is a best guess. Sometimes, these guesses turn out to be totally ill-conceived and a series of experiments seems to yield nothing of significance. Indeed, great discoveries surprise us. Without some surprise these discoveries cannot be considered to have altered the way we understand the world.

I believe that these sentiments are shared in many other fields. I note that writers of fiction often say, after the fact, that their characters took over the story and drove the story in unexpected directions. Similarly, writers of nonfiction often find as new facts tumble out that the story they thought they were telling turns out to be different.

Getting beginning researchers to accept failure and assume the risks of trying yet another approach is difficult but essential. A more seasoned investigator knows that failures are part of the creative process. What helps I find is to develop an attitude of what I call a contented schizophrenic, an attitude in which you are willing to believe in something and yet disbelieve in it, all at same time. Simultaneously, to believe passionately and to question critically appear to be contradictory activities, but such a mindset helps in seeking to understand nature. You must put forth you best idea how something may behave and then begin immediately to devise means of testing whether this idea about nature s behavior has beguiled and blinded you to what is taking place. Living with ambiguity is something that can be learned and even taught to others.

Science at its core is a subversive activity. We seem only capable of disproving things that seemed established. Only by exhaustive efforts of failed disproof do we gradually accept an assertion as likely to be so. That is the process that leads from hypothesis to theory to scientific law. And as we become wiser and more reflective, we find that even some of our most cherished laws only have limited scopes in which they are valid. Experimental work is often a series of failures punctuated by a few successes. A wise person learns from each failure and accepts that failures are nature's way of guiding us to what really works. It is human nature to be disappointed by these failures. But, learning from failure and learning to live with failure are truly key elements in becoming a successful researcher. Without nearly constant failure the few successes we experience would not be so sweet. The contented schizophrenic accepts failure as the only way to create innovation of real value. Failure does have its special virtues in encouraging us to unlock the mysteries of nature. In the research arena, failure is sure to be a nearly constant companion. The smart researcher uses failure as a powerful but often secret tool for finding success.

Let me thank you again for conferring upon me this great honor.

非常感谢omni兄提供的英文全文:)

我们不应该把一个项目最后做的那个实验称为唯一正确的实验。

－－－－－－－－－－－－－－－－－－

回乐在其中兄：上面说得很好啊：）

中文的没那有趣的开头，英文的没老爱那段有意思的话

like小妹看得真仔细啊,中文没那段开头我不清楚是为什么,但后老爱那段话是翻译者倪光炯先生,即昌海兄的本科老师,在附录中加上去的,呵呵.

很多所谓的失败实验，证明了某些方法的不可行，怎么能叫做失败呢？骗骗不懂科研的人罢了。

在高中时看到过关于这个实验（啤酒中的泡泡往下沉）的文章，当时的标题好像是“阿基米德错了？”