在所有社會科學甚至自然科學的方法論討論中，實證主義（positivism）都早已遭遇猛烈攻擊，但經濟學似乎是個例外，不僅方法論的討論不太多，且大多數經濟學家依然習慣抱著實證主義（或者修改一下成為證偽主義）不放。
姑且先撇開對於實證主義本身的方法論批評；仔細想想，經濟理論真的都能以符合實證主義的方法加以證實或證偽嗎？
Consider the basic hypothesis of the neoclassical approach: ‘All rational people are maximisers.‘ ……there are indeed many tests we can conduct in laboratories to find out if people behave in a manner consistent with the principles of utility maximisation. The problem is that such tests are not likely to settle the debate in favour or against the neoclassical hypothesis the people are maximisers.
If people turn out to maximise utility, neoclassical economists can modify their hypothesis subtly, for example they may put forward the amended hypothesis: ‘There is something that people maximise.‘ Perhaps there is. The problem is that this hypothesis can never be refuted, nor can it be verified. If we establish that people do not maximise utility, the defender of neoclassical economics can always turn around and say: ‘But how do you know that there is not something that they maximise?‘ The answer is: we cannot know. Thus the hypothesis that people maximise cannot be proved wrong(i.e. refuted) in which case we have a theory motivating all economic testbooks which is beyond the empricists’ and positivists’ method of theory appraisal.
‧註一：上面這段文字所批評的現象看來是真實存在的，比如賽局理論中關於「人是否自私」的那一大串實驗經濟學研究，或者如Gary Becker與《蘋果橘子經濟學》那樣，把「誘因」的定義放得無限寬廣。
‧註二：有一些經濟學家認為，60歲以前不要碰方法論問題，那是創造力枯竭的傢伙幹的事情，而且容易走火入魔。這話不無道理。但在做研究的過程中，方法論問題總還是會在這裡那裡冒出來，所以有空思考一下似乎也不錯。
————————————–
到目前為止，我曾在一些場合聽到經濟學家對此問題的兩種回答。
第一種可以算是方法論的標準答案，就是：經濟學與其他科學一樣，有許多自明的「公理」，例如上面提到的「效用最大化」。這些「公理」類似於基本定義，本就不是可以證實或證偽的對象；證實與證偽，是針對依據這些公理所推衍出來的種種命題與宣稱才能夠進行。這有點類似拉卡托斯的「核心-保護帶」說。
第二種回答比較妙。該經濟學家老實承認確有此問題，不過無所謂，只要能夠自圓其說、自得其樂就行了。呵呵。
