Instead of engaging in a psychological interpretation of Augustine's adolescence, Hadot's reading allows us to sec that we arc in the presence of a theological discussion of the nature of sin, and that Augustine's lengthy recounting of his theft is not autobiographically motivated, but is necessary in order for us to see the way in which sin is a perverse imitation of divine reality.
Moreover, by placing the Confessions within the Christian exegetical tradition, Hadot is able to show that the last three books of the Confessions, in which Augustine seems to abandon autobiography to devote himself to exegesis, far from being foreign to the rest of the work, do not ultimately have
"a different object from the account that is narrated in the biographical part." 81 Hadot demonstrates that Augustine very often brings together the two states of his soul - obscurity, then light - with the two states of the earth at the beginning of the account of Genesis. In its first state the earth was invisibilis and incomposita, and in its second state it received the illumination of the Fiat lux.82 In Book II, Augustine presents his adolescence as a state of obscurity and bubbling fluidity, and Hadot has shown that in this description one can recognize "the vocabulary employed in Book XIII of the Crmfessions to describe the chaos of Gen 1 , 2."83 Furthermore, in Book XIII the images of darkness and fluctuation serve precisely to describe "the state of the soul
Introduction
1 7
still 'formless,' before its conversion to God." M Thus Hadot can claim that
"the idea of the passage of the creature from a formless to a state of
formation and of conversion dominates the whole work." 85
Book XIII the
biblical account of creation becomes the description of the phases and stages of the salvation of humanity.86 Putting together Augustine's autobiographical and exegetical descriptions, Hadot can demonstrate the inner unity of the work, the fact that for Augustine "Genesis is . . . the account produced by the Holy Spirit of the conversion of the soul, as the Confessions is the account that he himself produces of his own conversion." 117 Hadot therefore warns us that we must interpret this text in light of the literary genre to which it belongs, the tradition of exegesis of Ambrose and Origen, and that we will commit a misunderstanding if we believe we have discovered the self "already" in the Crmfessions.Hll We find in Hadot's own interpretation of Augustine the initial outline of a kind of historical psychology, one that discovers in the Confessions the beginnings of the modern self. However, this is followed by a more detailed attention to the mechanisms of literary composition and to the theological genre of the Confessions, an attention that both prevents the apparent autobiography from becoming the philosophical center of the work and permits us to see the unity between the first ten books and the last three.
There is, of course, a self to be found in the Confessions, but "it must not be understood as the incommunicable singularity of the man Augustine, but, on the contrary, as universal humanity of which the events of the life of Augustine are only the symbols. " 89
Hadot's insistence on not separating conceptual structure from literary structure also played a significant role in his interpretation of Wittgenstein's work. As far as I have been able to determine, Hadot presented the first detailed discussions in French of Wittgenstein's books, reviewing everything from the Tractalus Logico-Philosophicus to the Philosophical Investigations and Remarks on the Foundations of Mathematics.'m In his 1959 discussion of the later Wittgenstein, Hadot argues, quite remarkably, that the goal of Philosophical Investigations requires a certain literary genre, that one cannot dissociate the form of the Investigations from Wittgenstein's conception of philosophy.
It is a therapeutics that is offered to us. Philosophy is an illness of language . . . The true philosophy will therefore consist in curing itself of philosophy, in making every philosophical problem completely and definitively disappear . . . Wittgenstei n continues [from the Tractatus to the Investigations] . . . to devote himself to the same mission: to bring a radical ;•nd definitive peace to metaphysical worry. Such a purpose imposes a certain literary genre: the work cannot be the exposition of a system , a doctrine, a philosophy in the traditional sense . . . [Philos11pl1in1/ hn•rstiRt1l1tms l wi11hcs to Rel liulc by little on our spirit, like a
1 8
Introduclion
cure, like a medical treatment. The work therefore does not have a systematic structure, strictly speaking [pas de plan, a proprement parler].91
At the time Hadot was writing about Wittgenstein, and even today, so many philosophers ignored the way Philosophical Jnvesligalions is written that it is astonishing, at first sight, to see an historian of ancient philosophy clearly understanding the import of this aspect of Wittgenstein's work. But Hadot has long emphasized that ancient philosophy presented itself as a therapeutics and that this goal profoundly affected the philosophical writing of antiquity.92
As early as 1 960 Hadot wrote that in ancient philosophy "more than theses, one teaches ways, methods, spiritual exercises," that "dogmas" have only a secondary aspect.93 No doubt it was precisely Hadot's understanding of the history of ancient philosophy that made it possible for him to see central, but still neglected, characteristics of Wittgenstein's work.
In "Jeux de langage et philosophic," Hadot was to employ Wittgenstein's notion of a language game in an historical perspective that, as he recognized, went well beyond anything with which Wittgenstein was preoccupied. Hadot argued that we must "break with the idea that philosophical language functions in a uniform way" and that ''it is impossible to give a meaning to the positions of philosophers without situating them in their language game." 'H Aware of the different philosophical language games of antiquity, Hadot could well insist that an ancient formula be placed in the concrete context of its determinate language game, that its meaning could change as a function of a change in language game.95 Thus Hadot could draw the general historiographical conclusion that we must "consider as very different language games those literary genres, so profoundly diverse, represented by the dialogue, the exhortation or protreptic, the hymn or prayer . . . the manual, the dogmatic treatise, the meditation . " 96 And we must also distinguish between the attitudes represented by dialectic, rhetorical argumentation, logical reasoning, and didactic exposition, since we will often be able to establish that "the very fact of sitt1ating oneself in one of these traditions predetermines the very content of the doctrine that is expressed in this language game." 97 By overcoming the temptation to see philosophical language as always functioning in the same way, Hadot could take account of the conceptual and literary specificity of different philosophical attitudes.