What, in short, we wish to do is to dispense with 'things'. To 'depresentify' them. To conjure up their rich, heavy, immediate plenitude, which we usually regard as the primitive law of a discourse that has become divorced from it through error, oblivion,illusion, ignorance, or the inertia of beliefs and traditions, or even the perhaps unconscious desire not to see and not to speak. To substitute for the enigmatic treasure of 'things' anterior to discourse, the regular formation of objects that emerge only in discourse. -- Michel Foucault (II, 3)
Ordinary experience is full of things lying around in plain view. These things (forks, trees, the moon) are so real and so present to us that to question the "objectivity" of our experience of them would be very odd. It would, likewise, be a bit odd to assert such objectivity. Such questions and assertions, however, begin to make sense when we want to understand the constitution of "theoretical" objects, or objects of science. These are the objects that characterise the specific discursive formations that Foucault is interested in. Since his inquiries address not "appearances" or "phenomena" or "experience" directly, however, but statements about them, let us remember that the relevant mystery to be illuminated is not (as in Kuhn) how experience is "seen as" objectively given, but how it is "talked about". Objects of discourse are what discourse is about.
Here it is important to keep in mind that within a scientific discourse, objects may appear just as real and as present as those of ordinary experience, and they may therefore be talked about in straightforward terms, as though they were "just lying around" waiting to be described by science. This presence is something Foucault is very keen to have us question in order to understand what produces it. Such understanding, in fact, is exactly what it means to understand what an object is.
As always, the important thing is learning how to describe something. In describing objects of discourse, we do not describe the things that science takes for given directly. Psychopathology, for example, starts with the presence of manifest "motor disturbances, hallucinations, and speech disorders," later including also, "minor behaviourial disorders, sexual aberrations and disturbances, the phenomena of suggestion and hypnosis, lesions of the central nervous systems, deficiencies of intellectual or motor adaptation, criminality." (II, 3) But these are not what need to be described. They are to be "dispensed with"; and instead Foucault suggests noticing three "rules of their formation".
First, he would have us describe the surfaces upon which these things emerge--things like a "speech disorder". What needs to be described here are other things that belong to contexts like families, groups, workplaces and churches, and, in other periods, contexts like art, sex, and crime. These are surfaces on which the objects of discourse, the things they are about, can turn up, and they therefore specify the things among which the objects of a particular science are themselves lying about. So if the family offers a surface for psychopathological objects this means that, say, a hallucination will turn up among ordinary things like bedtime stories, or Christmas dinners, or weddings. Or if crime is to offer a surface, then the "homocidal maniac" will turn up alongside the the murderer and pickpocket. We might say that the "surface of emergence" is the background against which the object of a specific discourse (the object "proper to a discursive formation") distinguishes itself, i.e., where it "stands out".
Second, Foucault would have us describe the authorities that are authorized by society to talk about these objects. Here it is important to keep in mind that all such authorities can talk also about the objects of other discourses. So, for example, the objects of psychopathology (behaviours indicative of various kinds of madness) can be delimited by medical professionals, as well as legal professionals and religious authorities or even, as Foucault notes, literary critics, each of which can talk also about other things (things other than instances of madness). Not only can they talk about other things: when they talk about the phenomena that interests psycholopathology, they are not talking about it as madness (except in a colloquial sense). It is, however, in the context of statements made by these broader bases of authority that the objects proper to a particular discourse (like psychopathology) are understood. That is, it is only because the objects of a particular science are discussed beyond the confines of that science, that these objects can be manifest, or present, or real to the scientists who talk about them.
Third, we should describe what he calls "grids of specification", which offer a set of coordinates that allow us to locate different psychopathological objects relative to each other. Foucault identifies grids like "the soul", which is divided into faculties, any of which may be defective, "the body", which is divided into organs that may, each and severally, be identified as defective, and "biography", i.e., "life histories", which may identify traumatic episodes that, again, are considered defective developments in the life of an individual. A particular mental disorder may be specified against one or more of these grids, but without such a grid there is nothing objective to say about it.
Having layed out this descriptive task, Foucault urges caution when drawing conclusions. First, he says, the presence of these surfaces, authorities and grids does not itself produce objects: "it is not the families . . . that decide who is mad", that is, madness (in the sense available to psychopathology) is not talked about in families (or in law courts or general hospitals). "It would be quite wrong to see discourse as a place where previously established objects are layed out one after another like words on a page." (II, 3) This, especially, if "mad" objects were supposed to be lying around there as well. It is only among these things (familiar things and criminal things; medical things and legal things; spiritual things and bodily things), as they are ordinarliy understood, that madness turns up. And it is not among any set of these things alone that this happens. All three "rules of formation" (surfaces, authorities, grids) must operate together in order to produce the things of scientific discourse (objects).
Thus, Foucault says, the description of the objects of discourse is an "imposing" task, involving a "complex group of relations". It does not lead us to a description of the internal constitution of objects as they are given, but a description of their "field of exteriority". The sense in which madness is not inside people (as the "presence" of madness to the psychopathologist suggests) but always outside other groups of relations: the family and common crime; somatic illness and religious revelation; mental faculties and bodily organs. He then draws attention to what he calls "primary" and "secondary" relations that are not the same thing as "discursive" relations. Take two of the contexts that are not object domains of psychopathological discourse, but part of the rule of their formation: families and criminality. As Foucault points out, there are all kinds of social relations between the institutions that conceive of themselves as civil (the middle-class family) and those that define crime (the legislatures and courts). These are "primary relations" that are not always straighforwardly correlated with the production of "mad" objects (like hallucinations). Moreover, not even what psychiatrists might say about the relation between non-psychopathological things (christmas presents, say, and shoplifting) can be straighforwardly applied in describing the emergence of their objects. But they must, of course, be kept in mind when trying to capture the distinctively discursive relations that are the theme of archaeological description.
Lastly, it must be kept in mind that in looking for objects, and trying to describe the regularities of their formation in a particular discourse, we are not trying to identify the thing (e.g., madness) that a word (e.g., "madness") refers to. We are not trying to trace the connection between the sign and what it "really" signifies. Instead, we are trying to trace the relation between a sign and all the other signs that compose the discourse in which the thing is significant, i.e., we are trying to understand, not the conditions under which a sign is attached to a thing, but the conditions under which a thing becomes a sign. Objects of discourse emerge among things lying around in plain view, ordinary things. These things are "ordinary" in the sense that the scientist (here the psychiatrist) must approach them as a layperson in order to become an expert about his objects. Objects, then, are composed of the signs that ordinary things are capable of being.
