Foucault has strayed quite far from the basic element on which his methodology must operate: the statement. The term 'discourse,' with its somewhat shifting meaning, seems to have swallowed up this basic element, which Foucault will now seek to define 'at its root.'
The statement is not defined by its propositional content, since two identical propositions can have different enunciative characteristics depending on their location within separate discourses. Neither is the statement identical with the sentence; not only lists but also mathematical equations, botanical tables, and even graphs must be regarded as statements. Neither, again, is the statement the same as the 'speech act,' an entity formulated recently by English grammarians. The speech act comes close in that it is not the propositional content of speech, nor the physical act, nor the intention of the speaker, nor the consequences of the speech, but rather the speech as itself an act (of promise, of prayer, of condemnation, etc.). But speech acts often entail multiple statements in certain orders or relations; thus, the statement is not reducible to the speech act.
All these borrowed forms place too many limits on the definition of the statement, which appears as little more than 'irrelevant raw material' for these theories. Perhaps the statement is nothing more than any series of signs (or even a single sign). Language itself would then become a statement. But language only exists in the abstract, as an extract from the set of all statements. How about material signs as such, like a handful of printer's characters or the keys of a typewriter? They are not the same as a series of characters written down at random, which is a statement (namely, in this case, a statement about statements). If Foucault hits the first five keys on the French keyboard (A, Z, E, R, T), he has made a statement; but these keys themselves are not a statement.
Thus, the statement is neither simply a language nor simply a material sign. The statement is a necessity if we are to say anything about propositional content, grammar, or speech act, but it is not limited to any of these things. In short, the statement has no definitive structure, no ordering principle; it is rather that which enables such structures to be read. It is has no criteria of unity, because 'it is not in itself a unit, but a function that cuts across a domain of structures and possible unities.' It is this function that will be discussed next.
Foucault makes a strong case for a previously unarticulated aspect of the sign. The first half of the twentieth century saw the emergence of a whole range of theories about language and signs, from Saussurian linguistics (which posited that signs gain their meaning only through their difference from other signs) to the attempts by philosophers of the Anglo-American analytic school to see all language in terms of its logically intelligible propositional structures. Saussurian theory infuses Foucault's project. Foucault's statements appear to echo the nature of the Saussurian sign. The analytic theory about propositional content plays a less problematic role here: the theory is important for Foucault to reject with regard to the statement, but its irrelevance is clear and easy to demonstrate.
Foucault is on shakier ground in his exclusion of the statement from the theory of speech acts, which he performs through the claim that speech acts sometimes include more than one statement. In fact, the two notions share a great deal in their focus on speech as a function rather than just a content. The real difference is not so much about which theory posits units that include the units of the other (the argument Foucault uses here), nor even strictly about the dependence of the definition of the unit on the function of what is said (both theories would agree to this dependence). It is more a difference in the range of what the statement is supposed to cover: Foucault wants a definition of the statement that does not depend on any theoretical superstructure, any musing about 'context' or intentionality, or any limitations beyond the most extreme.
Thus, speech act theory would accept the five characters 'AZERT' as a speech act only after determining the context in which it was used. Foucault can call it a statement wherever he finds it enunciated, even before he knows what its relational properties are. Thus, the range of the statement is enormous, excluding only abstract series of signs (like language itself) and material signs (like ink on a page) that have not actually been enunciated. Thus, the statement as such is gaining some definition here, but not in the sense of a structure or size or principle of individuation that holds everywhere and at all times (such aspects will depend on the enunciative context). We might call the realm of statements as such by a name Foucault uses elsewhere with regard to the field of discourse as such: 'anonymous.'
The study guide doesn't make things easier to understand.
I'm really having a hard time understanding what "discursive" means... This isn't helping me at all.