這篇文章討論在一些wh-疑問句裡可見的「都」量化現象，如『他都買了什麼？』。這個量化現象與眾所周知的「都」量化現象不同，也就是左向條件（the leftness condition）不能適用於前者。對於前者「都」量化現象，我提出了新的分析：這個「都」量化現象要參考的不是句法結構，而是主題－焦點結構，也就是說，決定量化的領域與一個句子所表達的舊信息與新信息有關聯。包含這個「都」的句子可以分成為主題和焦點，這兩個部份分別地投射到在「都」量化的三分結構（the tripartite structure）上的限制（the restrictor）與核心（the nuclear scope）。這個分析能夠說明，為什麼列舉的回答適合於「都」疑問句，為什麼疑問詞不能是表達數量的，還有為什麼疑問詞一定要帶複數的語意或者要帶量詞「些」。這個分析也能夠說明這個「都」的在句子裡面的分布，就是這個「都」一定要Ｃ－控制焦點的詞組。最後，我指出這個分析可以擴展到一些很少見但還是可以觀察到的陳述句上面，並指出這兩種「都」量化現象其實也可以在一個句子中同時出現。

This paper examines a type of dou quantification found in wh-questions such as ta dou mai le shenme? ‘What are all the things that he bought?’ This type is different from the well-known dou quantification in that the leftness condition cannot be applied to the former. I propose that the former type of quantification is subject to the topic-focus structure rather than to the syntactic structure, which means that the domain of the quantification is determined in relation to 'old' and 'new' information of a sentence. Sentences including dou can be divided into topic and focus, and each part is mapped onto the restrictor and the nuclear scope in a tripartite structure of dou quantification. This analysis accounts for the reason why a list answer is appropriate to questions with dou, why wh-words in the questions cannot be quantity expressions, and why wh-words should either have a plural interpretation or take the plural form. This analysis also explains the distribution of dou, i.e., dou should c-command a focused phrase. Finally, I point out that the analysis can extend to declaratives which are rare but still observable, and that the two types of dou quantification can arise simultaneously.