In the foreword of Les Mots et les Choses, Foucault speaks about three areas - the fundamental codes of a culture, the reflection of thought about those codes and the nude experience "of the order ":

The fundamental codes of a culture - those that govern its language, its perceptive schemas, its interchanges, its techniques, its values, the hierarchy of its practices - fix, from the start, for each man, the empiric orders with which he will have to work and in which he must be found. In the other extremity of thought, scientific theories or philosophers' interpretations explain why there is in general an order, to what general law this order obeys, what principle can in general justify it, for what reason it was this order that had been established and no other. But between those two such distant areas, it reigns a domain that, in spite of having an intermediate role above all, is not less fundamental: it is more confuse, more obscure and, certainly, less easy to analyze. [...] Thus, between the glance already codified and the reflexive knowledge, there is a middle area that delivers the order in its proper being: it is there that it appears, according to cultures and according to periods [...]. In a way that, this middle "area", as long as it manifests the manners of being of the order, can present itself as the most fundamental [...]. So, in every culture, among the use of what someone could call the order codes and the reflections on the order, there is the nude experience of the order and of its manners of being (1966: 11-12) (1).

The fundamental codes of a culture are rules that fix for the men born in a given community with a specific culture their language, their perceptive schemas, their systems of economic interchanges, their techniques, etc. The reflection of thought on those codes is constituted of theories, interpretations of those codes that try to justify them before other possible orders. Between the codes and the thought reflection, there is the experience "of the nude order", that cannot be reduced neither to the explicit codes that try to fix that experience nor to the reflection that is done on those codes. We have, like this, the following design:

fundamental codes of a culture
nude experience of the order
reflection of thought on those codes

The nude experience of the order was defined here just in a negative way, in opposition to the other two areas of the design. In order to say more explicitly what this experience is, it is important to consider some explanations that Foucault advances in L'Archéologie du savoir , regarding his methodology (more exactly, the methodology of his first three books: Folie et déraison, Naissance de la clinique , and Les Motes et les Choses ): 1) this intermediate area, of the "nude experience of the order", which could be called pre-discursive, is not dissociated from speeches, texts, enunciations, spoken or written, not only words but also graphs, lists, schemas; 2) it is the area of dispersion of those things.

Enunciation ( énoncés ) has an extremely important meaning here. In L'Archéologie du savoir , Foucault defines enunciation as "a function that crosses a domain of structures and of possible units and that makes [those units and structures] appear, with concrete contents, in time and space" (1969: 115). That formulation, apparently enigmatic, should be understood in the following way: a enunciation is a type of possible arrangement that, observed from logical or linguistic systems, can be said to constitute things as propositions, sentences, etc. But the enunciation is not aside, nor above, nor below sentences or propositions, it "is always inside units of that kind, or even inside sequences of signs that don't obey the laws of those units" (1969: 145). In this way, while it is a function the enunciation has four characteristics (2). These four characteristics, as defined by Foucault, reveal a dynamics of the enunciations that defy any logical or linguistic analysis that would try to put a straightjacket on any kind of discourse. On the other hand, they also defy, above all the second characteristic, explanations that would try to equate the dynamics of the enunciations to the intentions of given social subjects.

The analysis of the intermediate area, of the nude experience of the order, draws the multiple unfoldings of the enunciations, their (often discontinuous) irruptions that other reflections and interpretations so many times masked under the form of explanatory units. It is an analysis that reveals that area as an "historical a priori ", as Foucault already said in Les Mots et les Choses :

In the present study, it is that experience [of the nude order] that we intend to analyze. [...] Such analysis, it's clear, is not in the field of the history of ideas or of the history of sciences: it is rather a study that makes an effort for finding out what makes possible some knowledge and theories; in accordance with what order space the knowledge was constituted, on the ground of what a priori historical, and in the element of what positivity ideas could emerge, sciences could be constituted, experiences could be reflected in philosophies, rationalities could be formed, maybe in order to be disarticulated and vanish very soon (1965: 13).

(1) Our translation (as the other ones to be found in this paper).
(2) A) The enunciations, if they don't have referents, they have referentials: "a enunciation doesn't have before itself [...] a correlate (or an absence of a correlate), in the same way a proposition has (or hasn't) a referent or a proper name designates an individual (or nobody). It is rather linked to a 'referential' that is not constituted of 'things', 'facts', 'realities' or of 'beings', but of possibility laws, of rules for the existence of the objects that are there nominated, designated or described, [and for the existence] of the relationships one find affirmed or denied. The referential of a enunciation [...] defines the possibilities of emergence and of delimitation of what gives to the sentence its sense, to the proposition its truth value" (1969: 120-21). B) An enunciation maintains a relationship with a subject that cannot be reduced to the relationship the enunciation has with the person that articulated it, that uttered it: "If a proposition, a sentence, a group of signs can be considered 'enunciations', it's not because there was, one day, somebody to utter them or to deposit, somewhere, its temporary trace; but only to the extent that a subject's position can be assigned. Describing a formulation as an enunciation doesn't consist in analyzing the relationships among the autho0r and the things he said (or he wanted to say, or he said unintentionally); but in determining which is the position that an individual can and should occupy if he is to be his subject" (1969: 126). C) "An enunciation always has margins that are populated by other enunciations" (1969: 128) (what means that it always has relationships with other enunciations, which vary, for instance, if it is taken as a literary, or a scientific enunciation, or as a prophecy, etc. D) An enunciation always has a material existence: as repeated, rewritten, transcribed, it can stay the same but can also have its status altered. A sentence said by a writer in the real life or put by that same writer in a mouth of a character of one of his books won't be necessarily the same enunciation. The utterance that 'the earth is round', proffered before or after Copernicus, can be said to express the same proposition (if we maintain it has the same truth conditions), but it doesn't need to be considered the same enunciation.
*Este trabalho foi apresentado no simpósio "Husserl and the historical a priori of the sciences", o qual ocorreu no Instituto Max Planck, em Berlim, de 1 a 3 de julho de 2004.