The main question addressed by this article is this: How should one understand the role of the sentences of the Tractatus, given Wittgenstein’s statement that they are nonsensical? I begin with a presentation of three general principles of interpretation in order to avoid answering the question in an inappropriate way. I then move on to a short presentation and commentary on a selection of readings – namely, the ineffabilist, resolute and elucidatory ones – and elaborate the answers given by advocates (...) of these to the question explored here. I agree on many points with resolute and elucidatory readings: the Tractatus presents an austere conception of nonsense, and is not a book that seeks to present ontological or semantic theories. I point out, however, that these readings cannot fully explain the nature of Tractarian elucidations. Then I discuss those parts of the Tractatus which refer to the sentences of the book itself. The main proposal of my own approach is this: Tractarian elucidations should be construed as rules of translation, in that they show how to substitute certain expressions for others. They enable us to construct a notation in which everything that is expressible in ordinary language can be said perspicuously. (shrink)

RESUMO O objetivo deste artigo é mostrar que as interpretações "tradicional" e "resoluta" não livraram o "Tractatus" da aparente autoderrota paradoxal. Argumento que essas leituras apresentam apenas uma nova roupagem ao paradoxo. A leitura "tradicional" de Hacker acaba atribuindo uma conspiração metafísica ao "Tractatus", o que é incompatível com os objetivos do livro. A leitura "resoluta" de Diamond e Conant atribui a Wittgenstein uma conspiração autoral, o que contradiz suas opiniões sobre autoria e método. Com base nas dificuldades encontradas em (...) ambos os lados do debate atual, concluo este artigo propondo vários requisitos que a correta interpretação do "Tractatus" deve preencher. (shrink)