幾乎所有胡適研究者都注意到胡適一輩子重視治學（科學）方法，並以十字訣「大膽的假設，小心的求證」著稱；較少胡適研究者指出胡適雖一輩子談治學（科學）方法，但他更重視治學精神和態度，並認為習慣之養成更重要，最後且以為官四字訣「勤、謹、和、緩」告終；更少的胡適研究者談論胡適是否很一致地主張他的治學（科學）方法。本文的重點在第三點，並以沒有胡適研究者曾研究過的──康納脫（一譯康納特，James B. Conant）的科學觀──來衡量。
證據顯示，雖然胡適視康納脫為同道，但兩人最大的不同點在於胡適曾堅持有某一種科學方法，而康納脫一向認為沒有。弔詭的是，胡適自己最終也推翻了他的十字訣，個中蘊含的意義不用說，當然是沒有獨一無二的科學方法論，而只有比康納脫等人的科學精神、態度還要寬鬆的四字訣。於此，胡適自己悄悄地進行了雙重否定──否定了一元方法論，甚至連某種方法論也沒有了。
本文的結論是，十字訣和四字訣若分開處理，可能各得其所，但若把它們混在一起，認為治學方法也是科學方法，則是以寬泛代精密，甚至混淆了。在這個層次上，康納脫不可能是胡適的同志，甚至可以是論敵。試想想，胡適的十四字訣（「拿證據來」加上「大膽的假設，小心的求證」）科學觀，與康納脫的科學定義三大原素（即玄測的思想、演繹法、經驗的實驗法）相同嗎？又胡適的十字訣以及更常識性的四字訣「勤、謹、和、緩」的好習慣，能與康納脫的科學技術和策略的原則等同嗎？若不是，則他們就不是同道了。

Every scholar interested in Hu Shih must, or should, remember the following three elementary points. First, as almost every Hu Shih scholar has noticed, throughout his life Hu Shih emphasized the importance of methodology of learning (including science), and summed it up with the famous ten-word epithet, “Daring in putting forward hypotheses; careful in searching for proofs”. Secondly, some Hu Shih scholars point out that, though he advocated methodology of learning (including science) throughout his life, Hu Shih was even more emphatic about learning (scientific) spirit and attitude, taking the cultivation of certain habits as most important, and arriving at the conclusion expressed via a four-word epithet for officials’ conduct－“Diligence, Carefulness, Calmness, and Patience”. Thirdly, a few Hu Shih scholars deal with the issue whether Hu Shih was consistent in his advocacy of methodology of learning (including science). The present article emphasizes the third point, and examines it in the light of James B. Conant’s view of science, since in his latter years Hu Shih claimed the former to be his intellectual comrade in methodology of science.
By providing a detailed analysis of Hu Shih’s famous ten-word epithet and comparing and contrasting it with Conant’s views about and on science, the claim is here made that the greatest difference between them lay in Hu’s frequent insistence on the existence of a scientific method, while Conant consistently denied it. Paradoxically, however, Hu Shih at last had to give up his famous ten-worded epithet, which implied without doubt that he no longer held any shred of his former “one and only one” scientific methodology, leaving him with only the four-word epithet, with its sense of spirit and attitude now even less strict in intent than that held by Conant and others. In a word, Hu Shih himself quietly went through a double disavowal－a disavowal of his one methodology, on the other hand, and a disavowal of any methodology, on the other. At this point, Conant can no longer be counted as Hu Shih’s intellectual comrade in methodology of science, but rather an adversary: just ponder, can we equate Hu Shih’s definition of science as represented by his fourteen-word epithet (“show me the evidence,” plus “Daring in putting forward hypotheses; careful in searching for proofs”) with Conant’s three great elements (i.e., (1)speculative thinking; (2)deductive reasoning; and (3)cut-and-try or empirical experimentation) in his definition of science? Can we still equate Hu Shih’s definition of methodology of science as represented by his ten-word epithet and four-word epithet with Conant’s principles of the “tactics and strategy of science”? If the answer to either question is in the negative, then to that extent they are certainly no longer intellectual comrades in methodology of science.