The mainstream of the philosophy of science in the second quarter of this century—the so-called “logical empiricist” or “logical positivist” movement—assumed that theoretical language in science is parasitic upon observation language and can be eliminated from scientific discourse by disinterpretation and formalization, or by explicit definition in or reduction to observational language. But several fashionable views now place the onus on believers in an observation language to show how such a language is meaningful in the absence of a theory.

In the present paper, I propose to show why logical positivism failed to do justice to the basic empirical and logical problems of philosophy of science. I also wish to consider why the drastic reaction, typified by Thomas Kuhn and Paul Feyerabend, fails t o provide a suitable alternative, and to suggest that the radical approaches of recent writers such as Mary Hesse and Dudley Shapere hold out a genuine promise of dealing effectively with the central tasks that face the philosopher of science today.

Pour le philosophe intéressé aux structures et aux fondements du savoir théorétique, à la constitution d'une « méta-théorétique «, θεωρíα., qui, mieux que les « Wissenschaftslehre » fichtéenne ou husserlienne et par-delà les débris de la métaphysique, veut dans une intention nouvelle faire la synthèse du « théorétique », la logique mathématique se révèle un objet privilégié.

La distinction de deux Wittgenstein est-elle mythique, ainsi que le voudrait R. J. Bernstein? Mythe créé involontairement par un penseur aussi peu mythomane que possible, B. Russell, quand, dans sa préface au Tractatus, il assigna à l'auteur l'intention d'établir les « conditions d'un langage logiquement parfait ». Prenant le contre-pied d'un Wittgenstein-I lancé à la poursuite d'un langage idéal, Wittgenstein-II aurait, dans Investigations philosophiques, opéré un retour au langage naturel et au sens commun.

Il est sans doute possible d'exagérer la distance qui sépare le Tractatus des Investigations philosophiques (dorénavant dénotés par T et IP). Mais on ne saurait la nier. Quelques rétractations expresses enregistrées dans le second ouvrage nous en empêcheraient (par exemple IP 23, 46, 97, 114). Le problème est plutôt de déterminer l'ampleur et la nature de la transformation, de la mesurer.

If, as Wittgenstein remarks, ‘the meaning of the word “length” is learnt among other things by learning what it is to determine length’ (Philosophical Investigations, p. 225), must we conclude that part of what one knows in knowing the meaning of this word is that length is determined in certain ways, and hence that we should explain the meaning partly by explaining how we determine length, and that the meaning consists partly in certain ways of determining length?

It is not easy to construct a clear argument to that effect, and yet we might naturally so conclude, perhaps on the grounds that generally what we learn in learning X and what we know in knowing X are the same. We learn the names of the professors by learning that that is Professor Jones, that is Professor Smith, etc., and that is also what we know in knowing the names of the professors.

Aristotle's explanation of the “active intellect” in De Anima III, 5 constitutes a problem for us simply because we have to take this philosopher so seriously. If he were a writer given to poetic lapses or mythical adornments to his work we could consider dismissing the whole chapter as unessential. However, we know that Aristotle does not write unessential chapters, and that he is invariably engaged in an attempt to explain his subject fully and systematically, neither adding to it nor leaving anything out.

The De Anima, for example, is an attempt to describe the characteristics of living things, in an ascending order from the basic functions of vegetables to the most abstract thinking by human beings. The activities of nutrition, sensation, common sense, imagination, and movement are presented in Book II and in the early part of Book III.

Before undertaking an examination of the issues current in social criticism, it is necessary to settle upon the method which ought to be common to the several social sciences. For if the adoption of a particular method results in unconscious acceptance of features of the society under critical examination, then clearly any such putative critique is in fact futile. In this paper I examine some of Marcuse's largely negative comments regarding the method currently employed in the social sciences. I am not concerned to arrive at a complete view as to the appropriate method, but rather to expose some questions that need to be answered if any such view is to be formulated.