第三人論證與柏拉圖之理型論

The Third Man Argument and Plato's Theory of Forms

本文的主旨在於，討論如何藉由柏拉圖自身理型論的立場，來避免他在《巴曼尼德斯篇》所提出的「第三人論證」（Third Man Argument）的問題。根據Geogory Vlastos 的看法，這個論證乃是建立在兩個隱藏的假定上：「自我表述」（Self-Predication）以及「非同一性」（Non-Identity）假定。近年來，出現了許多詮釋，企圖藉由證明「自我表述」假定並非柏拉圖理論之一部分，來避免第三人論證。然而，本文試圖證明，第三人論證的錯誤並非在於「自我表述」假定，而是在於「非同一性」假定。換言之，我們可以藉由證明「非同一性」假定乃是一個錯誤的、根本不會被柏拉圖理論接受的命題，來避免第三人論證。除此以外，本文試圖指出，柏拉圖提出第三人論證，實際上並非為了要進行某種對自身理論的批判。反之，本文認為，柏拉圖在《巴曼尼德斯篇》132a1-b2以及132d1-133a7所設計的兩個第三人論證版本，應是為了要警告一般人對其理型論中有關理型與個物關係，亦即「分有」關係，可能產生的誤解。也就是說，如果我們錯誤地以一種物質論的方式解釋分有關係，那麼理型論必將陷入大性逆推（Largeness Regress）；另一方面。若我們將分有看成是一種現象個物與理型間的相似關係，但將這種相似理解成一種對稱性的關係，那麼，理型論就必將陷入「相似性逆推」（Likeness Regress）。

This article is concerned with the problem of how to avoid the Third Man Argument which Plato put forward in Parmenides. According to Gregory Vlastos, this argument is based on two tacit assumptions: the Self-Predication and the Non-Identity Assumption. In recent years there have been a number of interpretations which attempted to avoid the Third Man Argument by proving that the Self-Predication Assumption is not an acceptable part of Plato's theory. However, in this article I will show that the fallacy of the Third Man Argument does not lie in the Self-Predication Assumption, but in the Non-Identity Assumption. That is, we may avoid the Third Man Argument by proving that the Non-Identity Assumption is false.Moreover, in this article I will point out that by putting forward the Third Man Argument, Plato does not really intend to raise a criticism of his own theory. Rather, the two versions of the Third Man Argument, which are set out in Parmenides 132a1-b2 and 132d1-133a7 respectively, should be considered as his warnings against the possible misunderstandings concerning the relationship between Forms and particulars, i.e. the relationship of ”participation.”In other words, if we mistakenly interpret the conception of participation in a materialistic manner, the Theory of Forms will inevitably be caught in the ”Largeness Regress”; and if we comprehend the relationship of participation in terms of the ”likeness” between Forms and particulars, and consider the later to be a symmetrical relation, then the Theory of Forms must fall into the ”Likeness Regress.”