May 23, 2018:
"Epistemological Break" is a concept which the philosopher and historian of science Gaston Bachelard produced to think demonstrable discontinuities in the histories of sciences, for example the abrupt transition from Newtonian to Einstinian physics, in opposition to ideologies of continuous scientific progress such as positivism. Because Bachelard lacked the concept of the Problematic, his descriptions of these discontinuities are less precise than Althusser's, and, he tended to focus on one particular type of discontinuity in which the successor theory subsumes its predecessor, for example the way Euclidian geometry became a subset of later non-Euclidian geometries. Althusser transformed the concept into something which is both more general and more specific. More general because he encompasses theories which are destroyed and thrown away by the break, for example, the way the concept of "phlogiston" is eliminated by the concept of "oxygen" and the periodic table; more specific because he insists that the break is between the ideological pre-history of a science and its constitution as a science per se. Here's an example using Freud. When Freud surveys the literature in the first chapter of The Interpretation of Dreams, he demonstrates that the multiple theorizations preceding him are not only inconclusive but mutually incompatible: somatic disturbances, fragmentary memories of the previous day, gastric accidents, etc. He shows that these theorizations all treat dreams as phenomena of consciousness, and that this doesn't work, leading to what Althusser would describe as a closed theoretical space incapable of producing new knowledges. Freud's epistemological break is his production of a new concept, The Unconscious, which becomes the object of the new science he inaugurates.
The break which Althusser especially wants to demonstrate is between Marx's early philosophical works and the mature science of Capital. His purpose is to show that the philosophical categories of the 1844 Manuscripts aren't Marxist at all, but are rigorously and systematically pre-Marxist, specifically Feuerbachian plus an "injection" of Hegel's idealist dialectic necessary to make them appear historical. He makes the sensible argument that the status of these manuscripts is obvious given that Marx abandoned them as failures, locked them in a trunk, and never looked at them again. But he importantly devotes a series of increasingly detailed analytical texts to fully explicating the exact relationships of the problematics operative within them and the sequence of radically discontinuous steps Marx took in leaving them behind.
Althusser's thinking about Marx's break evolved considerably over the history of his work. In this first period, he saw it as a single event, which he situated precisely in 1845 in the "Theses on Feuerbach" and The German Ideology. The "Theses" say goodbye to Feuerbach with his ideological problematic of The Essence of Man, Alienation, Species Being and the rest; The German Ideology inaugurates the Marxist scientific problematic of Mode of Production, Productive Forces, Relations of Production, and so on; in Marx from that moment on there's neither Feuerbach, nor Hegel. Over time, Althusser's views evolved self-critically. By 1966-7 he saw the break not as an event but as a process, and he produced meticulously detailed readings demonstrating that while Marx inaugurated the break in 1845, Hegelian and Feuerbachian elements remain in Marx's thought to the end, including a contradictory and tense interrelationship in Capital. By 1978 he was even more critical, arguing that the expository structure of Capital itself is idealist, and that it contains Feuerbachian categories hiding behind other names. For the later Althusser, this self-antagonistic impurity is inevitable in any text: no text is ever pristine, no thought ever free from the problematics inherited from its own past. The break never ends, and, crucially, it's possible to backslide from post-break scientific positions to pre-break ideological ones which are always operative and never extinct. This is not the view he defends in this first period. Here, the break is sudden, clean, and done.