Tokenizing Deferrality

This post will be a synthesis and reframing of many of my previous discussions of language, prompted by the observation that I have never directly applied deferral to the analysis of language itself, in the form of analyzing and interpreting language in terms of modes, degrees and layers of deferral—something which, in the context of originary debt and singularized succession, seems like a good way to tie together some as yet unconnected threads. I was prompted to do so, in part, by reading Raphael Gross’s Carl Schmitt and the Jews: The “Jewish Question,” the Holocaust and German Legal Theory (a very good book, by the way), where I noticed that the Christian concept of the “Katechon,” or the restrainer of the AntiChrist, was a very near synonym to “deferral.” This makes it possible to further extend center study into new fields of inquiry, in this case not only Schmitt, but Peter Thiel’s studies of apocalypse, which makes regular use of this theological framing of politics. (While I have been familiar with the Katechon for some time, I think what jarred me into taking it more seriously was a point in Gross’s discussion where he uses the word “deferral” in describing Schmitt’s understanding of the term, and something clicked.) I’m not going to think theologically, but now I have a way of translating into anthropological idioms this theological framing which, it seems, Thiel has found no better “secular” alternative to.

It’s strange that with all of the framing of language I’ve incorporated over the years—from David Olson’s metalanguage of literacy conjoined with Mark Turner and Francis Noel-Thomas’s “classic prose,” the originary grammar I derived from Gans’s The Origin of Language, the rhetorical notion of commonplaces updated through construction grammar, my recent interest in philology as a model for humanistic inquiry and, most importantly, Anna Wierzbicka’s Natural Semantic Primes, it never occurred to me to theorize language directly as something like tokenized deferral. Deferral is scenic creation, and language records the scenes upon which it has been operative while creating scenes within scenes upon which we can say “this is the same” in a new way. All of the approaches to language mentioned above are intrinsically scenic and the utter ascenicity of Chomsky’s generative grammar is the reason I could never get interested in it. If deferral is meaning, and the constitutive outside of language, then any utterance can be translated into a particular mode and degree of deferral on a particular scene within the stack of scenes (and the translation itself can be so translated, and so on). To explain what brings me here, though, I need to introduce another approach to language, one which my day job as a composition instructor provided me with access to: the theory of academic discourse advanced by Laura Aull (and her sometime collaborator Zak Lancaster). Aull points out that the formulaic nature of academic discourse (and, of course, which discourse is not formulaic?) is a result of the assumption of its users that they are participants in an ongoing conversation in which no one will ever have the last word: hence the systematic hedging and the use of concede/counter moves that concede as much as possible to (“steelmanning”) the position one then goes on to counter. Now, Aull, for the purposes of doing large scale corpus studies, has compiled a list of expressions used for these purposes, but as one reads texts with these considerations in mind it becomes easy to see that there are many ways of hedging and conceding/countering in less formulaic ways and ways so uncommon that they wouldn’t rise to the threshold of being on lists used to conduct corpus searches. I’ve proposed a class in AI inquiry in which students would generate new terms for an expanded and potentially infinite taxonomy of hedges and concede/counters (it will never be approved but it was valuable to think it through) but, then, being somewhat familiar with the originary hypothesis, it wasn’t much of a leap to conclude that all language is doing this and that this is pretty much what language is.

My “amendment” to Gans’s hypothesizing of the creation of the declarative through the “negative ostensive” or “operator of negation” involved the declarative articulating the complementarity of the imperatives flowing into the declarative, one directed toward “reality” and the other toward others on the scene: in other words, the ambiguity of the operator of negation, wherein the imperator turned interrogator might be told to cease demanding the object and the object itself has been ordered off the scene (since in pre-declarative language there’s no way for something to simply not be there), leads me to posit this bidirectional imperative (or differently “vectorized” imperatives). So, in our declaratives we are calling upon our interlocutors to refrain from advancing upon their object by “channeling” imperatives from “reality” affirming that command: you must cease and convert your striving because the scene has been so ordered—that is the declarative, operationalized. I can now say that this involves a kind of “loan” or “credit” from the center, which must be backing a scene, which becomes a stack of scenes—that’s the relevant “reality.” It is your obligation or debt to the scene that enables you to “understand” a sentence, which is to say, to activate the questions and ostensive-imperative “units” that would confirm your position within the nomos or scene and thereby provide you with access to part of or a version of what you have demanded. So, to speak in sentences is to draw upon and issue lines of credit. The debt we owe to the center is a product of deferral, and the lines of credit issued are therefore tokenized deferrals. So, it’s not that the extreme deferentiality of academic language is a model for language but that it models language as deferral which can then be applied to all modes of language use, including direct, even brutal utterances, which simply enact the complementarity of the operator of negation and invocation of an unyielding reality in a particular way. A general giving a direct order to march is enacting deferral by establishing order and purpose amongst his troops and is maintaining the imperial institutions that keep peace within that order; the scribes who will record the poetic renditions of the battle over the centuries are participating in the same practice of deferral. And, of course, this holds true for people doing “bad” and ‘evil” things with the difference being that what is ultimately judged to be bad or evil has undone or subverted more deferral than it enacted.

We can, then, now address directly the question of how to measure and tokenize utterances as modes, degrees and increments of deferral. Deferral works on different levels; even more, deferral is the creation of those levels—deferral is scene creation. When you say “excuse me” when just avoiding bumping into someone in a public place you defer what is probably a minimal risk of mild conflict in that situation while also continuing earlier instances of “excuse me” in what is essentially an institution of deferral. How you say “excuse me” also, then, operates on both levels (and other levels at which this form of politeness forms part of a network of politeness in general) and the two levels can even be at cross purposes: a rushed “excuse me” (in a very close encounter) might come across as aggressive and therefore serve the immediate purpose of deferral while slightly weakening the sincerity level “excuse me” needs to operate more generally. We could take this in a more theoretical direction by examining linguistic and rhetorical forms like “irony,” which works insofar as the interlocutors or reader and writer can both inhabit two scenes at the same time, or, really, three scenes—the scene upon which the utterance would have been taken literally, the scene upon which we observe the hypothetical inhabitants of that literalistic scene, and the scene upon which we straddle both scenes and can move back an forth between them. These are the increments of deferral generated by new scenes upon which we can say “this is the same,” creating a new path connecting ostensive, imperative, interrogative and declarative. This is the scene that is tokenized, insofar as it becomes a predicate for other more far-flung scenes which we could leave to our successors to inhabit and activate.

We can further integrate deferrality with granting credit—any act of deferral involves a kind of reciprocal credit granting, insofar as time and space is granted before the possibility of confrontation comes back onto the agenda in the form of the decision to either enforce or forgive the debt. The creation of new scenes that are operational within the stack of scenes provide the most durable form of credit—a failure to live up to the sign will be registered and recorded. Language is then a pool of credit, with idioms sealing scenes within scenes that would bear upon pedagogical futures insofar as their iterations generate new scenes cutting across the stack, which is to say, new modes of cooperation across time and space. Once the center is occupied by a person, deferrality will always be asymmetrical to some degree—even if each side is crediting the other, degrees of credit granted and means and time frames of enforcement and leisure to forgive will vary. Credit is never simply a matter of the relation between creditor and debtor, since the one extending credit is always himself deriving credit from broader credit granting orders and is betting upon some degree of predictability of succession (that, at least, occupants of the center will be sufficiently “similar” to each other in relevant respects). Here the question becomes whether one wishes to short the broader credit system or buy a call option far into the future. Either way one is arbitraging at each point along the way, expanding and exercising communications systems or data exchanges so as to credit the most secure streams. What this will mean in language is a continual extension of discourse, or even the same sentence, enhancing self-referentiality, becoming increasingly unintelligible for those shorting the system and increasingly meaningful for those picking up ever more distant call options. When some information or communications or media expert says something like “what we are all doing now is..” you should see an attempt to short some credit line and find a way to issue or receive credit for doing the exact opposite. Still, sometimes systems need to be shorted, but doing so should always be accompanied by a call option that others beside you will someday be able to cash out but will probably to continue to defer, maintaining it as credit and collateral. Never forget Peirce’s admonition that in the long run every insurance company will go bankrupt, but the meaning of that is we will then be left with the insurance company that was always backing the insurance company, the human power of deferral itself, falling back upon primitives, bereft of our vocabulary drawn from the world of investment.

Wierzbicka’s Natural Semantic Metalanguage has a privileged place in treating language as modes and increments of deferral. The NSM allows for, on one end the most minute and precise analysis of differences between idioms and even within the same idiom, insofar as every utterance is arbitraging the differences between the way a particular idiom plays out to different listeners or readers while, on the other hand, tracing all of the differing idioms back to a single one, a kind of originary declarative idiom. I’ve pointed out many times that there’s no need to assume that the primes are the first words created—on the contrary, the first words would have been names-of-god, while the primes have no reference to god or the sacred (Wierzbicka explicates “God” as “a person, not like us”)—but, rather, that they represent a kind of residue of all languages as they explicate themselves and create pathways between ostensives and declaratives. We can, for example, imagine that a word like “look” came before a word like “see” in a particular language because, for example, it is better suited to directing attention in an emergency situation—but “see” would become necessary as soon as the ostensive is issued inappropriately, as a way of showing that it has failed: “I don’t see it.” This would retroactively reveal that one must see before one can look, or that looking presupposes seeing. With the use of AI, we could explicate entire languages and reveal them to be vast articulations of the primes, articulations that can be revised in real time. We are always already operating on the language (I am old enough to have a vague sense of how the American English spoken today is already quite different from the one I spoke as a child, and have noticed some of these shifts along the way, albeit haphazardly—when did people start saying “it’s all good”?) but now we could do so more deliberately, precisely by selecting and generating idioms with more deferral credit.

And we could inscribe deferrality into idioms through the primes, much like designing or revising genes in the laboratory, in this case by working out a kind of programming language that keeps expanding references from the interpersonal to transactions between individuals and institutions, institutions and data, data and the center. “I want you to do X” gets translated as “if you do X someone else can do Y,” which becomes “someone else can do Y when this person knows what I sometimes do and what you sometimes do,” which finally becomes “I want to do things in a kind of way because if I do things in this way people can see that what I do is like what other people do in some ways. If other people can see this some people can think about what I do. The people who think about what I do can say: ‘here is a good way to do this thing that you do.’” These are just very preliminary prototypes, of course—teams would have to work on this and on training these proto-idioms against the linguistic data of humanity. David Olson’s study of the metalanguage of literacy helps with the transitions here because, as a theory of media, it enables us to see the layering of language (from see to look to observe to detect, etc.) as a layering of scenes and references from one scene to others.

Those located at critical junctures in the stack of scenes could, of course, leverage the imperative chains available to them and speak much more directly and bluntly—preserving and enhancing the imperative order by using it regularly and consistently is also a mode of deferral: uncertainty in the chain of command is a significant source of rivalry, crisis and violence. All other utterances themselves have reference, at however many removes, from that imperative order, coming ultimately from the occupant of the center or, we might say, the “primearch.” The closer you are to an actionable chain of imperatives the more direct, explicit utterances are enactments of higher modes of deferral; conversely, the further one is from such a chain, the more demands, much less commands, become ridiculous, to be displaced in favor of utterances that might circulate as likely increments of deferral. But by now I’m at the point where this line of thinking needs to be instantiated in a company.

Tokenizing deferrality would coincide with continual inquiry in mimologics, which is to say the minute mimetic expressions that eventually manifest as potentially crisis forming rivalries but which, given imminent increases in scenic simulation, can be identified early on in rivalry and crisis formation; at the same time, the more minutely we can inquire into emergent mimetic logics, the more we might be simultaneously creating new ones, this time situated within the disciplines and the juridical sphere. More tacit processes of imitation, differentiation and resentment are made more explicit but since the various paths from emergent practices to those consequences that institutions would need to take notice of will always be probabilistic, disagreements requiring deferral get transferred to the construction of protocols governing institutions. The minutiae of differentiations in possible protocol constructions create a more generative set of rivalries, though, since they can be settled through the programming and accrediting of likely futures, which are the very effects of deferral that are to be tokenized. It should be possible not only to respond to but to anticipate rivalries by issuing a linguistic token, or a sample as token that would indicate the resolution of some demand in a juridical ruling that would create a new differentiation or notch within the nomos in the form of a scene articulating an imperative from the center. What counts as such a scene is a new way of saying “this is the same”—all our language comes down to that, and a new way of saying “this is the same” is a tokenizable act of deferral. All of our technology, our pedagogical platforms stacking up scenes, are nothing more than the creation of new ways of saying “this is the same” at different scales and different types of assembly because being able to continually say ‘this is the same” in new ways means offering guarantees or credit or insurance for succession because it confirms a commitment to maintaining presence against all “evil,” or attempts to short the present.