Traditional logic, a central pillar of metaphysics, involves turning a subject-predicate relation into a definition, and then using that definition to “certify” another subject-predicate relation. “Old people are bad drivers”; “that man is old”; “that man is a bad driver.” A particular subject-predicate relation, along with the definitions of the words involved, is assumed to be stable, which makes it possible for logic to take on a machine-like form of operation and ultimately because the basis of new kinds of machines. This mechanism is transparently a result of literate culture’s hypostatization of the declarative sentence, which produces both grammar and definitions. Aside from the fact that words change their meanings, can have multiple meanings and, indeed, may have less “meanings” than “uses,” any definition relies upon metaphysical or anthropological assumptions that can’t be “proven” within the system itself. But it’s very helpful for a mode of thinking to have a logic, less to adjudicate disputes within the system then for pedagogical purposes—a logic helps produce shared problem-solving devices and habits upon which more advanced forms of inquiry can be built.
I think that Anna Wierzbicka’s natural semantic metalanguage (her “primes”) can provide us with the basis of a generative, “anthropomorphic” logic. Her NSM provides us with a set of words with a stable meaning, but their meaning is not fixed through arbitrary definitions produced through a particular metalanguage, but through the existence of words in every language with these same meanings. This places these words beyond definition—any words you could use to define “think,” for example, would in turn need to be defined in other words, and so on, and you will ultimately be brought back to the word “think” itself. Now, the word “think” can be used in lots of different ways, so we can question the unity and stability of the prime words as well, but a good place to begin developing the primes into a logic is to note that the prime words limit each other. So, in sentences like the following—“I think I might come”; I’ll have to think about it before I decide”; “you may think so, but wait and see”—the word is being used in fairly different senses: first to indicate indecision, second, to refer to a process of cogitation, and, third, to contrast assumption or expectation with reality. But one thing is constant across all three uses: someone “thinks” when one doesn’t “know.” Similarly, however many ways we could use the word “do,” what they will all have in common is that insofar as you’re “doing” something, something is not “happening to you.”
It’s important to point out that there’s no reason to assume that the prime words, any, much less all, of them, were the first words in any language. It’s best to think of them as the enduring residue of declarative language—these are the words that we couldn’t make sentences without. Part of the project of transforming the primes into a logic will involve hypothesizing “paths” through the ostensive, imperative and interrogative to the declarative on the part of the primes, but that will involve looking at the primes as teleologically oriented towards becoming the declarative “infrastructure.” The primes are the minimal language needed to talk in and about a world in which imperatives can be refused or disappointed without increasing the likelihood of inconclusive and destructive conflict. If we resist the habit of seeing words like “think,” “know,” “want,” “can” and so on as representing “inner states,” “capabilities,” “potentials,” and so on, we can see that they all allow for the “codification” of various forms of hesitation: “I want” replaces some form of “give me”; “I can” introduces some space between what one has been commanded to do and the actual doing, and so on.
Wierzbicka’s purpose in developing the primes is to develop a logic of translation—first, she demonstrates the untranslatability of the “key words” in any language, and then she introduces the primes as a means of translating them. She both proves the Sapir-Whorf thesis and transcends it. A generative logic would be more a logic of translation than of “correction.” Instead of taking one claim and validating or disqualifying it, we want to be able to translate discourses into other discourses. We then get a logic that both reduces a discourse to its minimal elements and expands it into other discourses. At a certain point I will introduce construction grammar into the equation—construction grammar is the linguistic theory that contends that meaning resides not in individual words but in formulaic constructions. This theory of language agrees best with both Michael Tomasello’s demonstration in Constructing a Language that children learn language through the absorption of “chunks” of language learned in daily interactions and with studies of oral culture that show the basis of oral poetry in fixed formulas and commonplaces. Wierzbicka herself may not see things exactly that way, but we will be able to make her NSM consistent with construction grammar. Once we do, we will be able to construct a logic that is based on translation operations carried out on familiar constructions.
Let’s take a look at a couple of prime words in relation to non-prime words that are very close in meaning. (Of course, the results of this exercise will be different in different languages.) First of all, “see,” which is a prime, and “look,” which isn’t. We can right away see a hierarchy between the words: you can see without looking, but you can’t look without seeing. Seeing is built into looking; looking is a particular way of seeing. You look in order to see something, while you see whatever is in front of you (even involuntarily)—looking adds a layer of intention onto seeing, which is intentional only in the most minimal sense of seeing something. You ask someone if they see something, or what they see, while you ask someone what they’re looking at. If you ask someone whether they see some particular thing you have in mind, and they say they don’t, you will tell (command) them to “look there.” Once they look, you ask if they see it now—“seeing” is the ostensive confirmation of the command to look.
We can do the same kind of exercise with primes like “touch,” “feel,” “want,” “think,” “say” and “know,” but none of them seem to have such an obvious “complement” as see/look—for example, the relation between “want” and “need” seems to me less complementary, as does the relation between “say” and “speak,” or “tell”—and I’m not at all sure what other words might be “closest” to “touch” or “think,” especially if we want to stick to a pre-literate vocabulary. So, we might want to have more of a method before approaching those—it will probably turn out that there are several different kinds of relationships, involving not only semantic differences, but ostensive-imperative relations, first vs. third person reporting and so on. But “hear” has a relationship to “listen” that seems to me perfectly analogous to see/look—“listen” adds exactly the same layer of intentionality to “hear” as “look” does to “see,” and the interrogative—imperative-ostensive loop also seems to me identical—you might need to listen more closely just like you’d need to look more closely. It’s certainly no coincidence that these are the two senses through which we take in “meaning”—but, of course, we have to assume that this analogy is not identical across all languages (otherwise, “look” and “listen” would also be primes).
If we continue on with these two, then, we could trace a path from see/look through all the other words used to indicate taking something in visually—“observe,” “notice,” “view,” “identify,” “spot,” “distinguish,” and so on—or aurally (a quick look at an on-line dictionary reveals that there are far fewer of these). So, if someone “makes a distinction,” he sees something—seeing something would be the ostensive “verification” at the end of whatever trail from seeing gets us to “distinguish.” We always come back to the primes—to start spanning out a bit, if someone “speaks” or “tells” something, that person must have saidsomething—you can always ask what, exactly, they said—which is a demand that a quoted statement be provided. If someone “comprehends,” theyknowsomething; if someone “reflects” or “contemplates,” they thinksomething. If you distinguish, you see that two things are not the same (all primes). If you identify, you see one thing that is not the same as anything else. If you observe, you see something happening (or not happening). We can use the other primes to add in these layers of intentionality: you wantto see if something will happen, or if something is not like other things, or if one thing is not the same as one other thing; and once you have seen, you knowthat something happened, that things are not the same, and so on. Each layer of intentionality is a layer of deferral, and being able to say that maybewe canknow or see allows us to add more layers. And we can construct some kind of ostensive-imperative-interrogative pathway in any of these cases, which would in turn open the inquiry to questions of institutions, or where we do these things. In this way, we can develop ways of detecting the equivalent of what traditional logics call “fallacies”: if some statement can’t be brought back to “this person said,” “this person saw,” “some person could see if…,” then it has no path back to the ostensive and is ultimately devoid of meaning.
We can take any sentence and break it down into the primes to as granular a level as necessary. So, for example, “the armed robber killed the victim who resisted.” We can start with a formulaic sentence: “someone did something bad to someone else.” There are a lot of bad things people can do to each other, so we’d need to approximate further. “This someone wanted something that the other had. The other did not want this someone to have it.” Along the way you’d have to lay out the moral objection to armed robbery and murder simply by translating them into the primes—why is it “bad” to do something to another because you want something the other has? We’d work our way through “do,” “affect,” “change,” “hurt” and so on—there are good ways of affecting and changing people and bad ways. The bad ways might be when the person affected can’t do some things that person did before. But, of course, we can imagine cases in which it would be good to ensure someone can’t do at least some of things he did before—so we need to get more precise. It would be making it so that person can’t do things which are good, or that we know are good, or that all people think are good—with each of these claims calling for scrutiny in turn.
A generative logic of translation, predicated upon a fluency in the primes, would be enormously helpful in, to refer to a famous paper of Charles Sanders Peirce, “make our ideas clear.” And we could do so in a way that never loses touch with a basic human being in the world, or ethics and morality. Everything we do or say is either “good” or “bad”—or, at least, that question will always be pertinent. We can interrupt even the most abstruse chain of reasoning, filled with hypotheses, speculations, assumptions, conditionalities and so on, at any point, and ask questions like, “if you say this, what other things can you say?” “What can’t you say?” “What can you do if you think this?” “If you say this can you say that what others will do because they heard it will be good?” Shouldn’t anyone be able to answer such questions? A statement worth working with, and re-translating in turn into other spaces, would be one that can be completely dissolved into something like things that we do because we want to see that something is the same as before, because we could then say it is good—or some other articulation of the primes. It’s a kind of laboratory built into language, allowing for both the testing of hypotheses and the invention of new discursive devices.
The primes could lead us to more adventurous and paradoxical logics. I suggested above that insofar as you are doingsomething, something is not happeningto you. I meant this very literally—describing what you are doing as you do something excludes consideration of whatever might also be happening to you—which might, of course, be represented later. But maybe the mutual exclusion is the equivalent of Euclidean geometry, where we simply assume the existence of points, lines, right angles and so on. Maybe in a more non-Euclidean prime logic we explore ways in doing things is a way of having things happen to you and having things happen to you is a way of doing them. Maybe saying things is a way of hearing things and there are similarly transactional relations between seeing and thinking, doing and wanting, and so on. We could then bring this more pataphysical prime logic to bear on the layers of intentions we uncover in the disciplines. The disciplines are built so as to foreclose such possibilities, but leave themselves open to them in all kinds of ways. Imagine a pedagogical enterprise that prepares people to conduct such clarification operations.
Maybe this should be more formalized. It may be better to produce sample translations to serve as models. At any rate there’s plenty of work to do. But the end point should be to combine the traditional functions of logic (determining the clarity, consistency and truth of statements) and rhetoric (invention, responsiveness to conditions) so that anyone who acquires fluency in prime logic can intervene effectively anywhere, with a non-arbitrary base of assumptions.