The primary historical problems to be examined in this book are listed as 'discontinuity, rupture, threshold, limit, series, and transformation.' First, however, Foucault must perform the 'negative work' of dismantling various received forms of continuity as they appear in historical work. The primary example he gives is the notion of 'tradition,' with its attendant themes of sameness, permanence, and origin (which is often linked with the idea of innovative 'genius'). Related ideas up for critique are those of influence (the linking of ideas over time), development or evolution, and 'spirit' (as it serves to associate a period with a collective consciousness, as in 'spirit of the times'). In order to begin, we must replace history as it is structured by these 'ready-made syntheses' with a mere 'population of dispersed events.' Also to be thrown out are the extant categories by which historical materials and events are divided up; this includes categories like literary genres, as well as broader divisions like those between politics, philosophy, and literature. Foucault points out that such categories are themselves historically contingent 'facts of discourse.'
Two of the most important categories to dismantle are those of the book and of the oeuvre. The book is a false unity because its boundaries are unstable and permeable. Is the unity of the book the same, for example, in the case of an anthology, a volume of a history of France, a transcript of a trial, or a novel? Do two books by two authors have the same relationship to each other as two books in a single cycle by the same author? What about the relationship between Joyce's Ulysses and Homer's Odyssey (on which Joyce's novel is structured)? 'The frontiers of a book are never clear-cut,' Foucault writes. Every book 'is caught up in a system of references to other books, other texts, other sentences: it is a node within a network its unity is variable and relative.' Again, the idea of the book as a free-standing work is itself an effect of the field of discourse of which it is a part.
The uvre (the totality of texts by a given author) is subject to even greater instability and complexity. The name of the author is a sign attached to each of the texts, but it signifies in different ways if the text is, for example, published under a pseudonym, extant only in unfinished form, or merely a notebook. Should a translation of someone else's text be considered a part of an author's uvre? What about a questionnaire he or she filled out? What about works the author abandoned or disowned? In the end, the idea of an uvre depends on the imagination of a certain 'expressive function,' a process that is highly variable.
Two final, 'linked, but opposite' forms of continuity must be rejected. The first is a tendency always to avoid 'the irruption of a real event' by implying or asserting a vague, fundamental, 'secret origin' that precedes it, 'an ever- receding point that is never itself present in any history.' The second is a tendency to take actual statements and 'manifest documents' to be expressions of a deeper, silent 'already-said' that makes statements possible. In contrast to origin, we must look for irruption; in contrast to silent movements of collective thought or spirit, we must look to actual statements, 'as and when [they] occur.'
These forms of continuity are not simply to be thrown out as such, but to be interrogated as effects within 'the totality of all effective statements in their dispersion as events.' This field of statement-events is the field of Foucault's investigation. This project is not like that of linguistics, which is concerned with finite statements only as instances of general, 'infinite' rules of language. Nor is it aligned with the history of thought, which seeks out generalized 'discursive totalities.' Rather, Foucault seeks to 'grasp the statement in the exact specificity of its occurrence,' to account for the reasons why a given statement had to be that precise statement and no other.
Even if we end up re-discovering the various forms of continuity via this new project, we will have made at least three important steps. First, we will have advanced the understanding of what a statement is, showing how it is linked to writing and speech, to its own repetitions and transformations in future statements, and to a wide range of other statements that precede or follow it, even while focusing our attention on the unique, irruptive specificity of the statements themselves. Second, we will have removed links between statements from linguistics and from conjectural histories of thought, avoiding restrictive links between the statement and the speculations about the psychology of the author. Finally, this new project will leave us free to discover new forms of continuity, this time via a set of 'controlled decisions' rather than a blind acceptance of 'secret' wholenesses.