Brunschvicg’s distinction between the systematic doctrine and the idea of criticism remains helpful for understanding the sources of the enigmatic uses of Kant in post-Bergsonian French thought.1 In retrospect, it provides the model for an attitude of interpretive discrimination that has been pervasive, one that abstracts what is actually valid and needed as criticism from the dead weight that is the systematic doctrine. While the viability of this means of access has generated extraordinarily inventive uses of Kant within the space of the idea of philosophical criticism, by severing ties to the second order details and complexities of his system, it also helped to conceal the risks and the conceptual interest of these very uses of Kant. This is particularly true for Foucault and others who worked on contemporary problems as Kantian problems, on the basis of study of Kant’s text at its most systematic, historical, at times even philological order of consideration; however, the effects of this work are not always immediately visible in their writings. It is precisely the effectiveness of Brunschwig’s distinction between the idea of criticism and the doctrine of the system in Kant that invites its own transformation, one that it programs before the fact. The limitations of the idea-doctrine reading of Kant suggest a transposition of this conceptual couplet to a field receptive to the dynamic features of the sources and the form of criticism. The present analysis outlines such a field. It presents a general grid for understanding Kant that is organized not by the idea of criticism-doctrine of the system opposition, but by a conception criticism as a practice described in terms of its degree of abstraction from the historical, philological, and systematic details of the text. Rather than two types of standpoints in relation to Kant with different objects, methods, and implications, this way of reading Kant defines

a single standpoint that can be occupied at different levels of abstraction. The appropriate level of consideration can only be fixed in relation to the object of criticism, in terms of more or less determinate objectives. On this grid, Kant and Foucault practice criticism as a way of thinking, and the ideal space of analysis of the relation between them is the point of abstraction at which criticism functions as a single shared practice, in each case described in relation to different problems and texts. This practice involves the exercise of a standpoint from which Kant’s texts can be understood in a historical and conceptual space regulated by a conception of its object, its method, its unity, and its implications at a second order. The following analysis targets that ideal point, the optimal level of abstraction for the description of the practice of Kantian criticism, and has no pretension to exegetical originality or philological proof: while constructed from textual material in Kant, its main points are at a remove from the current problems of Kant scholarship, and although philological and historical problems are involved in the analysis, they are not its primary object. When considered from a unified standpoint, the Kantian problems treated represent a distinctive and recognizable practice. The textual exercise aims to reorganize a series of problems in Kant in order to accentuate his affinities not only with Foucault, but with a series of figures in the history of thought whose relation to him are obscured by peculiarities of the way he is studied, and that may appear arbitrary, contingent in relation to what is in fact accessible for study in Kant’s texts. As it might have amused Foucault to think of the strategic need in the present context, through textual analysis of the unity of criticism as a practice, one can relieve a certain number of instruments of thought like troops from a fruitless battle without issue in the heights of Kant-studies, freeing them for use on more fructuous terrains.