We are not free, autonomous centers of moral and intellectual activity; nor are we talking apes, whose behavior can be explained by its adaptability to evolutionary pressures, which have caused certain traits to be selected for. We are beings bound to the center: everything that we say, think or do is homage to the center. If there is a logic to history it is a very uneven one: we are all working to bring all of the centers, from the mundane references we make daily, to the authorities we obey and commands we convey to others, to the divine beings we worship, to the originary event itself, into alignment. The logic is uneven because it follows the path of the linguistic presence that is constantly threatened and constantly retrieved. The first imperative was an inappropriate ostensive: someone named an object without realizing it wasn’t there, and another member of the community retrieved it, so as to “make good” on the sign. A new form is thereby created. We can assume all cultural creation proceeds this way, a little bit sideways as one is sent astray by a mistake, and then forward as a new iterable form is produced. The ostensive sign is transformed by the introduction of the imperative, both projectively and retroactively: the ostensive sign, from then on, not only creates linguistic presence in the actual presence of some object, but is the source of new imperatives; meanwhile, the interdiction on the originary scene on appropriation of the central object can now be seen as having an “imperatival” quality to it. This also involves some abstraction from the originary object and scene: the central locus, which subsists beyond the consumption of the object, is somewhat less bound to the form of that object. It now receives imperatives along with issuing them, and is therefore increasingly defined by this dialogue with the community, what I have been calling an “imperative exchange.”
The question, as always, is what is the center telling us to do? Answering that question was once the shaman’s job, then the prophet’s, but the academics and intellectuals who have succeeded them are far less credible (there always seemed to be more false prophets than true, anyway). A very interesting essay by Marshall Sahlins, “The Original Political Society” (in the new book he co-edited with David Graeber, On Kings) challenges the notion of the egalitarian primitive community. Such communities were (and are), in fact, extremely hierarchical—it’s just that the hierarchy includes lots of non-human actors. Ancestors, animals, various gods and spirits (“metapersons”) all play roles in governing the community, and they do so in an extremely domineering and terrifying way. Sahlins sees himself as disproving Durkheim’s claim that the supernatural figures that populate the collective imagination of primitive peoples are “merely” projections of and means of consolidating the actual social relations within the group. He makes the point that the metapersonal governing “elites” in these communities have far more complex structures than the communities themselves, and include lots of positions and rules that have no equivalents within the community itself. I’m not sure this really proves what Sahlins claims (there’s no reason an interpretation or projection of social relations couldn’t be more complex than a straightforward description of them would be), but it provides a very good example of what we might consider the axiom of the sufficient center: human societies will discover/invent the centers they need in order to generate the command structure necessary to control resentment. If two young men are about to fight, I might stop them by saying “your late, sainted, mother would be ashamed of you!”—insofar as I succeed in conjuring up the late mother of one of the combatants I have invoked a center capable of issuing imperatives. For a whole society, I would need lots of mothers, fathers, other relatives, forces of nature, animals from whom the group claims to have descended, and so on, to impose the necessary measure of shame.
The sainted mother, then, who is “ventriloquized” by the peacemaker, represents a center (she serves to redirect attention, leading to the deferral of violence); and she does so by in turn referencing previous centers, which can also be invoked, if the need grows greater. (She was a stalwart at Church, she had sainted grandparents herself.) So, what the center says to us, what it commands, depends upon what kind of violence needs to be deferred in the present, and what kind of inherited resources are available (do we have a God who said, “Blessed are the Peacemakers”?). But discovering what the center wants is easier in an emergency situation (even if compliance requires more courage); finding ways to channel it for the sake of serious thinking, in the midst of complex situations that nevertheless require decision after decision to be made on schedule, is something else. It becomes an epistemological question. An epistemological question must frame an imperative, leading to an ostensive: do this in order to see that. The relation between language and metalanguage, which I have been working with in recent posts, can be put to epistemological use. I have been focusing on writing as a metalanguage, the first fully explicit metalanguage, as in representing speech writing must inquire into what, exactly, constitutes speech. But, just as David Olson points out that the features of language revealed by metalanguage were already part of language, the possibility of metalanguage must also have been internal to language, which means that something like the language/metalanguage distinction goes back to the origin of language. The repetition of the first sign issued was a commentary on that sign, selecting and singling out certain features as opposed to others. This is simply a reminder that language is self-reflexive and recursive, which is part of its paradoxical nature. The language/metalanguage distinction and relation is one particularly important instance of the differential repetition constitutive of language.
Only in the fully developed metalanguage that Olson identifies with the assessment of prose (which he, in turn finds most characteristically exemplified in “classic prose,” which seeks to reproduce the scene which the writer purports to witness), though, could we grasp the epistemological implications of metalinguistic awareness. Metalinguistic terms, for Olson, tell us nothing about the practices involved in producing the text; I’ve been treating that observation as implying that canonical metalinguistic terms are unable to generate operable imperatives. Perhaps the most widely used metalinguistic term for referring to writing is “clarity.” Writing (and, then, speaking) can be clear or unclear. This makes perfect sense on the terms of classic prose, which aims at providing an unobstructed view of a scene, reduced to its essentials. The addition of details which don’t help make the scene one upon which everyone would presumably see the same thing reduces the clarity of the writing, speaking or thinking; likewise with the absence of details required to make the scene a “complete” one. If we read enough prose, we know what we think is clear or unclear—no matter how certain we are of our own judgment, though, telling someone to “be clearer” is perfectly useless. He thought he was being clear. Now, if you tell someone that he has to give us more information about a particular figure in his narrative because otherwise we can’t know why he acts as he does at a certain point, that could be turned into an imperative, although still a limited one: what, exactly, needs to be known? You could tell him, but then you usurp the position of writer. In other words, the imperative that follows from even a more specific observation is still something like “read my mind.” You’re simply telling someone to assimilate you as a model, but without already being you, how is that supposed to work? This kind of empty or meaningless metalanguage silences the center, putting established models and their guardians in its place.
What I suggested instead is the imperative to distinguish metalanguage from language in one’s discourse. You can always do this, and you can always do more of it because you keep generating the distinction. And it’s always productive, because you can always bring out more implicit features of your thinking, which is certainly one of the primary activities “thinking” refers to. Now, metalanguage in Olson’s sense is anything that tells you, implicitly or explicitly, how to imagine or judge a representation of events whose meaning would have been unequivocal had you been on the scene reproduced by the prose. All of those words that are essentially more specialized versions of “think,” “say” or “know” (“consider,” “examine,” “explain,” “imply,” “conclude,” “extrapolate,” and dozens if not hundreds of others) are metalinguistic. In the more expanded version I’m proposing, I’d like to push the line all the way back, to the semantic primes themselves: think, say, know, want, and so on. The metaphysical view of these terms, derived from the metalinguistic inquiry into the declarative sentence, is that there is something in us, an activity, that, once we identify and study it, we can call “thinking,” knowing,” or “wanting.” In other words, we just apply the name to something already there. There’s something there, of course, insofar as we repeat and vary phrases and sentences to ourselves, and that activity issues in things that we say, but to assume that we have a word, “think,” and that therefore there must be something inside us that “thinking” refers to, and so we can isolate, define, study and improve that activity is to create a false problem—a problem, moreover, created by the metaphysical assumption that the declarative is the primary linguistic form. “Think” is a word that is used in a lot of different ways, many of which have nothing to do with any reference to a mental activity occurring somewhere “in” us. If I ask someone, “what do you think of that?,” I’m not asking him to articulate the cognitive processes at work in his mind when he is confronted with “that.”
So, if “think” is, not so much metalanguage, as possessed of a metalinguistic dimension, in what, exactly, does that consist? First of all, it consists in the limits of using “think” as an imperative. You can tell someone to think in general, or to think about something; you can’t command someone to think specific thoughts, though. If he thought the thoughts you told him to think he wouldn’t really be thinking them. So, when we say someone is thinking we are saying we would expect him to say something different than anyone else would. When I ask someone what he thinks of something, I’m asking him to say something only he would say about it. So, when I speak about thinking, I’m really thinking about speaking. And I’m not thinking about knowing, because if we all know something, we would all say the same thing. But there is something happening when someone is asked to say something only he could say that we don’t have words for, even though it’s undoubtedly taking place in language. So, “think” is metalinguistic insofar as it refers to and assesses that process (if I ask someone what they think about something and they say exactly the same thing as someone else I’ll be disappointed—I might tell him he’s not really thinking) but without providing an operable imperative for advancing it. The way, then, of displaying the distinction or boundary between language and metalanguage in thinking, then, is to refer to some of that silent playing and working with received chunks of language that made it possible to say what only you would say, and to mention the word “think” as part and also not part of that process. Asking someone to think is asking them to be a center that iterates the originary center, and one can ask oneself to think and thereby let the center speak through you.
The discourse of the center must work its way through one, some, or all of us—someone has to speak for the center. Anyone can ask the center for something, and as soon as you do, you intuit that the center must require something from you in exchange. Help me pass this exam, God, and I promise I’ll never go out drinking the day before an exam again. If you convert this relationship into declarative form, there’s a problem: “God helps unprepared exam takers pass on the condition that they prepare themselves in the future.” The problem is, that if we present this as a statement that is true every time we say it, the preparation would never have to take place, and so it can’t be true. The declarative includes both real presents, the exam-taker’s future preparation and God’s bountiful act now in the same linguistic (which is to say, “portable”) present, in which case they have to coincide “logically,” and they can’t be made to do so without paradoxical remainder. The statement is logical in general but illogical for any individual case because the subsequent preparation could be endlessly deferred—that is, the statement would generate imperatives (don’t prepare) that cancel the ostensive base of the imperative (God helps). The imperative of declarative construction is to bring ostensives and imperatives into alignment: imperatives are issued by some ostensive, and lead us back to one. The declarative constructs a scene upon which we observe some entity in a relation to some center (approaching it, retreating, violating it, protecting it from another…), and scenic rules apply, so two incommensurable relations to the center cannot be envisioned simultaneously. (Of course, a sentence like “God helps those who don’t help themselves” could be uttered as an absurdist or caustic riposte to the commonplace, but it then makes sense as a comment on that commonplace, which is to say it shifts our attention to a different possible scene, one peopled by those overly familiar with “God helps those who help themselves.”)
It’s possible to construct a declarative version of the imperative exchange that squelches the dialogue with the center, by either taking God out of it or reducing God to a metalinguistic verification of “sincerity” or synonym for “reality”—something like “if I prepare for exams I will pass them,” or “God wants me to prepare for exams,” or “if I prepare, God will make sure I pass”—that’s what we’d end with if the exam-taker’s promise is kept and no further imperative exchanges on this topic are necessary. If this individual prepares for and succeeds on all future exams, and exam-like tasks, he may very forget the imperative exchange that made it all possible, and in that case he is no longer listening to the center. But there will be more “tests” and various conditions of “unreadiness” for them, and unless I lock myself into the pretense that I am perpetually prepared I will have to keep asking the center to advance me some capacity that will supplement my unreadiness. Passing one test makes the next one more consequential, its contours less discernable, and the reward less calculable. This process proceeds through repeated apparently failed imperative exchanges, and a series of revisions of declarative versions making a “present” out of the terms of the imperative exchange. If I keep returning to and abstracting from the center, I head towards the creation of requests to the center that are simultaneously their own answers. We will still hear the center imperfectly after all this, even if we all help each other extract the most enduring declarative forms by serving as centers for one another. But we can keep refining our power to hearken and heed.