Summary of Part 2
The second Part of The Archaeology of Knowledge, “The Discursive Regularities,” begins to develop Foucault’s own historical method of “archaeology.” Remember that, in contrast to previous historical methods, Foucault wants to exploit the discontinuities of the historical record. Thus, the first chapter of this Part, “The Unities of Discourse,” is an attack on all the previous ways in which the mass of historical documents has been organized to create categories that seem to offer coherence. For instance, when people talk about “tradition” or “influence,” they are connecting something in the present to something in the past, making a continuity. Sometimes, Foucault says, these connections may be defensible. But we shouldn’t take them for granted. Rather than assuming continuities across time, we should prove them. Similarly, Foucault rejects categories that divide documents into different disciplines like science, literature, philosophy, and fiction. Often, these divisions are retroactively imposed on historical documents when these categories did not exist at the time of their writing. Once again, we must not allow history to begin from a position of taking these artificial divisions for granted.
In addition to these disciplinary and temporal continuities, Foucault also rejects an even simpler continuity we might take for granted: what he calls the “book” and “oeuvre.” Just because a bunch of statements have been bundled together in a book does not mean they go together. And just because a single author wrote a bunch of books, forming his or her oeuvre, does not mean all those books go together either. Statements within books cite and reference and motivate statements in other books, too, and authors change over time so we cannot say the Foucault in 1950 is Foucault in 1960. Thus, these forms of organizing discourse, of providing continuing between different statements, are also artificial and should not be taken for granted. Instead, Foucault advocates for a “pure description of discursive events.” This means exploring the statements within and across different books and authors in order to see what categories emerge from within the discourse, instead of being imposed by historians and their own categories of continuity.
If we do not rely on material books, biographical authors, disciplinary formations, and cultural “traditions” to organize archives of historical documents, then how else do we organize history? This is the question Foucault explores in the second chapter of this Part, “Discursive Formations.” Here, Foucault rejects a few alternatives. We might make categories according to all the documents that talk about a single object, concept, or theme (for instance, everything that has been said about trees or morality or religion); or we might look at all the statements that are written in a similar style. Instead, Foucault argues we should look for what he calls “systems of dispersion.” This is his term for how statements are not cohered around something in common, but instead create differences that run in all sorts of directions. Statements are connected in terms of “regularity,” when a certain kind of statement keeps showing up in a particular time and place. Then the question becomes why this statement keeps showing up, why it has become what Foucault calls a relatively stable “discursive formation.” The proper historical task in turn is to spell out the “rules of formation,” or the conditions under which a statement has achieved regularity.
Chapter 3, “The Formation of Objects” demonstrates how we might do so by way of a history of psychopathology, which Foucault also examined in his History of Madness. Psychopathology deals with very many and sometimes contradictory “objects,” or topics of discourse, from what it has called hallucinations to minor behavioral disorders and sexual aberrations to criminality and intellectual disorders. How is it that all these objects started to be related to one another, forming this new discourse of psychopathology? When and where did these objects emerge? Foucault offers three avenues for research. First, look at “surfaces of emergence,” or in what contexts these objects are first talked about. For instance, perhaps some objects were first talked about in relation to the familial context or a religious context, and so they emerge within a domestic or religious discourse. Second, look at “authorities of delimitation,” or the institutions that are authorized to talk about these objects. This might be medical professionals, for instance. Third, look at “grids of specification,” or how different objects are related to one another or put into organizational systems.
While doing this kind of analysis, however, Foucault warns against a couple of traps. One is thinking that the objects already existed and people just had to start talking about them. Rather, talking about something often brings it into existence. For instance, before people called madness “madness,” and thus made it a psychological problem, they might have been aware of people possessed by the devil, or morally abhorrent individuals; but they did not know of any people who were "mad" in the modern sense. Thus, in an important way, "madness" did not exist at that time. It is when certain phenomena are named as objects that their nature and meaning change. This means an object cannot be defined all by itself. You can’t look internally at madness and understand what it really is. Rather, you have to look externally to see the discourse in which “madness” appears and gives it meaning. This means we are studying not the internal make-up of objects, but their relation to one another and how meaning is produced from the outside, as a whole. This, finally, is also what distinguishes Foucault’s method from a philological study of the meanings and etymologies of words. Foucault doesn’t just want to understand the definition of a word. Because he is rejecting the distinction between names for things and things in themselves, to understand the words implies actually understanding the thing. So Foucault claims that he is studying how objects come into being and what rules are in place in a particular society that govern when and how they do so.
Chapter 4, “The Formation of Enunciative Modalities,” lays out a method for learning precisely this: the law operating behind statements. The question has become not what is inside a statement, but under what conditions it is enunciated at all. Foucault invites us to examine several aspects of a statement: who is speaking or allowed to speak it, in what institutions or settings is it proper to be spoken, and what is the speaker’s relation to the object he or she is speaking about? For instance, with madness, we might notice that late 19th century statements about its nature, causes, and cures are made almost exclusively by doctors in clinical settings who are observing what they call “madness” in others rather than experiencing it themselves. This clarifies, too, that statements about madness do not belong to any one individual or speaker. We aren’t studying what motivates a particular person to say a particular thing. It is by noticing patterns—or regularities—of statements that we understand the larger context of an object’s appearance, or the invisible rules that govern when and where and by whom a statement can be said and understood. This is more sociological than psychological.
In Chapter 5, “The Formation of Concepts,” Foucault broadens his scope to see the relation between different objects, or how concepts influence one another in a discursive “field.” This refers to how some concepts require other concepts to make sense and so one follows another; as well as how some concepts coexist at the same time, in which case they may either be independent or contaminate one another. At the end of this analysis, we may have a very diverse field of concepts, but we have shown how a set of rules has governed how the concepts formed. Again, these rules have to do with when and where it makes sense for something to be said, not rules of language like grammar. Sometimes, different groups of concepts will develop into a larger theme or theory (as when hallucinations, criminality, sexual aberration, etc. all came under the umbrella of psychopathology), which Foucault calls a “discursive strategy.” Then, additional tasks emerge. First, we must study “points of diffraction,” when and where related concepts become incompatible in the same statement. Second, we must consider the “economy of discursive constellation,” which means the rules and conditions under which some statements were chosen over others when they were found incompatible. Third, we should consider “a field of non-discursive practices,” or all the things that did not get said at a certain time and consider why such speech was restricted. Again, this will have to deal with who has authority to speak and in what contexts speech about a concept makes sense.
In the final Chapter of this Part, Foucault provides a useful summary of what he has proposed so far:
We set out with an observation: with the unity of a discourse like that of clinical medicine, or political economy, or Natural History, we are dealing with a dispersion of elements. This dispersion itself – with its gaps, its discontinuities, its entanglements, its incompatibilities, its replacements, and its substitutions – can be described in its uniqueness if one is able to determine the specific rules in accordance with which its objects, statements, concepts, and theoretical options have been formed: if there really is a unity, it does not lie in the visible, horizontal coherence of the elements formed; it resides, well anterior to their formation, in the system that makes possible and governs that formation. (80)
To recap: first, Foucault has recast discourse as about dispersion instead of continuity. Then, he has found ways of looking at the rules that dictate what is said and where and when, which explains diversity rather than unity in the discourse. To the extent that statements nonetheless go together, this unity is not based on surface appearances, like a common theme or a common author, but on being motivated by the same underlying set of rules.
Analysis of Part 2
This Part lays out in greater detail Foucault’s assault on unities or continuity in discourse, which he seems to think are always an illusion. In addition to the historiographical tasks spelled out, Foucault is proposing a few psychological tasks as well. For it is as if he is asking historians to become comfortable with the uncomfortable. Historians have to let go of the comfort that the illusion of sovereign subjects provides. And historians have to become comfortable with associations between statements being drawn in unpredictable ways.
There may be unspoken political motivations for Foucault’s abandonment of particular unities, especially the unity of the author or singular creator. The Archaeology of Knowledge was published in 1969, a year after massive political mobilization and student revolutions and uprising throughout the world, but especially in France, where Foucault was teaching. He seems sympathetic to the revolutionary sprit, and this requires solidarity with classes and groups of people rather than identification with single leaders. People want democracy, not monarchy. And similarly, Foucault implicitly suggests that grouping discourse according to particular authors would be a kind of nostalgia for a monarch or dictator. Instead, we should see how statements are circulated in a social system in ways that leave behind such “sovereignty.”
At the same time, Foucault is implicitly suggesting the great difficulty of revolution. For rarely does a discourse change overnight, and Foucault does not believe in radical or sudden ruptures. His interest in “regularity,” or the slow congealment of discourse over time, suggests a different time horizon. Moreover, his interest in “authorities of delimitation” suggests that new concepts emerge often within a field of power, like the field of professionals or political elites, more than in the field that targets or disrupts power. Part of the effect of Foucault’s approach is to demonstrate how much cannot be said at any given time, because of the rules that govern discourse.
Foucault’s comment that archaeology is not etymology is an important one. It suggests he is not interested simply in the history of language, or how different words are formed at certain times or derive from other words. This would be a kind of impersonal history, too, but it does not tell us much about social transformation. Rather, what is interesting for Foucault is how different words get grouped together or relate to one another over time. A concept is not just a word, but a bringing together of different words into a new object of study. Because concepts shape how we see the world, in a real sense words are the objects that we see. This is an important point for Foucault's argument.
Overall, one of the things that makes this Part effective is Foucault’s toggling between “negative” and “positive” aspects of his account. The negative aspects refer to his critique of different unities. For instance, he shows why it doesn’t make sense to restrict statements to authors or books because of how they flow beyond any one container, and how people change over time. But instead of just arguing that what everyone else is doing is wrong, he also lays out a methodical process for describing the relations among statements in a new way. He thus presents a problem, but also presents an answer. This pattern will continue in Part 4, where Foucault is particularly intent on contrasting the historian of ideas (who he thinks is wrong) with the archaeologist (who does what the historian of ideas cannot).
