>Where does Derrida say that subjects are traces? The point, to summarize at
>the risk of being really reductive, of Derrida's bit about the trace is that
>"it"--the trace or mark; he also calls it "differance," "dissemination,"
>"pharmakon," "glas,""parergon," "+R" among others--is the *condition* of
>communicability, of information, of writing conventionally understood. Put
>differently: it is the phenomenological condition of *the intention to say
>something.* Derrida never says that your intentions (or anyone's intentions)
>aren't clear to you or anyone else, only that those intentions don't
>circumscribe the meanings that can be made of your statements.
>
>Put differently: Derrida says that, instead of "deformations" of your
>intentions being marginal or accidental--deformations such as "lies" or
>"deception," "taking someone's words out of context"--the possibility of
>lying, deception, or taking someone's words out of context is "built-in" to
>the structure of utterances. That is, they are not "accidents" against which
>there is some "normative" mode of enunciation. They are as much a part, tho
>not necessarily more a part, of "communication" as "good intentions,"
>"clarity," "non-irony," "saying exactly what you mean," "meaning exactly
>what you say" and so on. All constructions of meaning (not just lies,
>deception, and taking someone's words out of context) are conventional,
>historical, and depend on the agreement, the AGENCY of people ("speaking
>subjects," in the lingo) involved
Your interpretation of Derrida sounds fine, but to say "X is conventional" and to say "X is historical" do not mean the same thing; I should like to discuss the difference in some detail, because the last sentence quoted above ends up implying the identity between "conventional" and "historical." When philosophers such as Rawls and Nozick discuss justice, their discussion involves the use of heuristics of the kind developed by Social Contract theorists of yore: treating principles of justice -- be they actual or counterfactual -- as if they were merely "conventions," or "patterns of human action," as Rawls puts it, which are pure "contingencies" with no necessary relation to the mode of production (not even "in the final analysis"):
***** The natural distribution is neither just nor unjust; nor is it unjust that men are born into society at some particular position. These are simply natural facts. What is just and unjust is the way that institutions deal with these facts. Aristocratic and caste societies are unjust because they make these contingencies the ascriptive basis for belonging to more or less enclosed and privileged social classes. The basic structure of these societies incorporates the arbitrariness found in nature. But there is no necessity for men to resign themselves to these contingencies. The social system is not an unchangeable order beyond human control but a pattern of human action. In justice as fairness men agree to share one another's fate. In designing institutions they undertake to avail themselves of the accidents of nature and social circumstance only when doing so is for the common benefit. (Rawls, _A Theory of Justice_) *****
For Rawls (as well as Habermas), it is possible to "design," in a voluntaristic fashion, the institutions & principles of justice (in the case of Rawls, the "difference principle" elaborated through "a hypothetical initial arrangement in which all the social primary goods are equally distributed") that would be "acceptable both to the more advantaged and to the less advantaged individual" (in the case of Habermas, the ideal speech situation as a counterfactual ideal of procedural justice). The word "convention," in philosophical discourse, therefore carries the dualist worldview that posits, on one hand, "individuals" or "people" (imagined as a collection of individuals with perhaps Weber-like "social positions," more or less "advantaged") and, on the other hand, "conventions," _both_ abstracted from social relations, so "conventions" may become matters of consumer choice. Individuals, considered from a historical materialist point of view, do not exist in separation from the ensembles of social relations; in fact, individuals are their ensembles of social relations.
Now, what of Derrida and other late modern philosophers? Do they regard meanings as "conventional" or "historical"? In so far as they do not consider the necessary relation between language and social relations (however "relatively autonomous" the former may be from the latter), they end up implying that language is "conventional," as opposed to "historical." Derrida's philosophical bent, in particular, seems to me to be transcendental, in that his primary interest lies in finding the _same_ conditions of linguistic (im)possibility _throughout_ the history of what he thinks of as "Western metaphysics." Despite its name and multiple avatars, _differance_ is "itself a powerful principle of unity," as Peter Dews puts it in _Logics of Disintegration_. Further, since Derrida insists on the logical priority of _differance_, he ends up returning, again and again, to the "idea of the first." As Adorno says, "The first of the philosophy of origins must become more and more abstract; the abstracter it becomes, the less it explains..." (qtd. in _Logics..._). Dews notes: "When Derrida speaks of the 'historico-transcendental scene of writing', he continues -- like Husserl and Heidegger before him -- to erase the contingency of the historical process, a contingency which can only be approached through empirical investigation...". So, if Derrida does regard "constructions of meaning" as "conventional," for him "conventions" must be at once purely contingent (= with no necessary relation to social relations) _and_ eternal (which is the meaning of "always already"). In other words, eternalized, abstract contingency (which can only exist in transcendental speculation) negates _both_ historical necessity and contingency. In this sense, _differance_ is a Platonic idea.
Yoshie