因明從古印度發展而來，後來成為佛教追求知識的重要途徑，如同西方邏輯，因明也強調推論要具有特定的形式，才能達到致知的目的。因明的發展從陳那之後有了較簡潔且完整的體系，陳那對因明的發展貢獻極大，他把之前的五支作法改良為三支作法，如以下範例所示：
宗：聲是無常。因：所作性故。喻：若是所作，見彼無常，譬如瓶等（同喻）。若是其常，見非所作，如虛空等（異喻）。
宗是立論者想要成立的論題；因陳述了宗成立的理由；喻則是在說明因為何會使宗成立。宗又可分為宗有法和宗法，宗有法是立論者的主詞所代表的事物，宗法是立論者主張宗有法會具有的性質。為了論證宗有法具有宗法，立因時非常重要，必須滿足特定的條件才是正因，因必須滿足的條件則被稱為「因三相」。就表面來看，因三相的成立會使得同喻體成為一個全稱語句（在以上的範例中，同喻體是「凡所作皆無常」），這使得許多人用亞里斯多德的三段論來理解因明論式，因而主張陳那的三支論式是演繹論證。然而，陳那在討論因三相時，似乎預設了「除宗有法」這樣的但書；不少學者主張這使得同喻體無法成為一個全稱語句，進而損害因明的演繹性。於是，對除宗有法的爭議分成兩大陣營，有一派堅持陳那使用的仍然是演繹論證，另一派則主張陳那使用的是歸納論證或類比論證。本文建議我們應從另外的角度——脈絡的觀點——來看待因明的論證，這個觀點來自於史東內克（Robert Stalnaker）對於斷說（assertion）所提出的理論，筆者會指出史東內克對斷說的要求非常契合因明立宗時的要求，而其背後關於脈絡的討論更有助於我們去理解因明論證的目的。從脈絡的觀點來看，因明論證是要從雙方能接受的知態背景下去進行討論，除宗有法是去確保對話的可能性。但這只是一開始的階段，脈絡會隨著對話的進行而產生變化，尤其當立宗者能證明其提出的因能滿足因三相時，宗有法會無可避免地被帶入脈絡之中。如史東內克所言，脈絡是隨著對話而不斷變化的，它是一個動態的互動過程。總之，從脈絡的觀點來分析因明論證的話，演繹或歸納這樣評價也許並不恰當，史東內克的所提出的合理推論會是一個更佳的評斷準則。

The relation between Western logic and Buddhist logic is still controversial. A traditional view is that we can reconstruct Buddhist logic from Aristotle’s syllogism, but some doubts arise when we go deeper into Buddhist logic. Consider the canonical argument below:
Thesis: Sounds are impermanent. Reason: Because of being produced. Examples: Similar corroboration & instance: what is produced is observed to be impermanent, like a pot. Dissimilar corroboration & instance: what is permanent is observed not to be produced, like space.
Dignāga provides the canonical argument, which is a milestone in Buddhist logic. Presumably, the reason and examples justify the thesis, but Dignāga’s exclude pakṣa makes things complicated. Pakṣa is the subject of the thesis, and exclude pakṣa means that examples (instances and corroborations) cannot contain pakṣa. If we interpret exclude pakṣa as a semantic restriction, as many contemporary Chinese Buddhist logicians do, corroborations will not be universal claims. In that case, corroborations will lose their justification to support the thesis. However, as Chi Yen Liu suggests, this paper interprets exclude pakṣa as a pragmatic constrain, and proposes an approach from context to analyze the canonical argument. This idea comes from the interesting similarity between the requirement of thesis and assertion theory. By revealing the strong connection between Buddhis logic and context, I argue that exclude pakṣa is just a strategy to avoid begging the question in the early stage to make conversation possible. As the debates proceeds, if the proponent successfully argues that the reason satisfies all the conditions that Buddhist logic asks, then pakṣa will be brought into the context. Thus, corroboration will apply to pakṣa in the debate eventually.