If Montezuma had collected all the pictograms drawn by his messengers, filmed their gestures, tape-recorded their words, and then locked up all this testimony in a chest before having the messengers put to death and then committing suicide, what remained in that chest would be the content of the expression maçatl for the Aztecs. It would then be up to the archaeologist who found that chest to interpret those interpretants in his turn, and only through the interpretation of that content would the archaeologist be able, afterward, to conjecture what the Aztecs’ CT of a horse had been.
A CT does not necessarily spring from a perceptual experience; it can be transmitted culturally (in the form of an NC) and lead to the success of a future perceptual experience. Such was the NC of maçatl that the messengers communicated to Montezuma by means of images, gestures, sounds, and words. On the basis of these interpretations Montezuma must have tried to get an «idea» of horses. This «idea» is the nucleus of the CT that he temporarily constructed on the basis of the NC received in the form of interpretations.14
The way in which NCs are expressed also helps to loosen the knot regarding the question as to whether we have mental images or not.15 An NC is expressed sometimes in words, sometimes with gestures, sometimes through images or diagrams. Basically, the drawing of Marr’s 3-D model, insofar as it is public, is an element of the NC that interprets a procedural modality of our CT. In our brain, what corresponds to that presumed image? Neuronal activation, let’s say. Now, even if the pattern of that activation did not correspond to what we intuitively call an image, those cerebral phenomena would represent the cause or the equivalent of our ability both to conceive and to interpret our type of the horse. We postulate a CT as a disposition to produce an NC, and we treat an NC as proof that there is a CT around somewhere.
3.3.2.1 Instructions for Identification
The NC of the term also supplies criteria or instructions for the identification of one of the tokens of the type (or rather, as they say, for the identification of the referent).161 use «identification» instead of «recognition» because I should like to reserve the latter term for cognitive phenomena strictly dependent on a previous perceptual experience, and the former term for the capacity to identify perceptually something about which we still have no experience. I identified an alligator, the first time I saw one on the banks of the Mississippi, on the basis of the instructions that had been supplied to me previously through words and images. That is, the NC of the word alligator had been communicated to me.
By supplying instructions with which to identify a token of the type, the NC orients one toward the formation of a tentative CT. If the messengers supplied Montezuma with good interpretations, his tentative CT would have been so rich and precise as to permit immediate identification, with few readjustments on the basis of direct perception. But sometimes the instructions supplied by the NC are insufficient. The messengers might have insisted to such a point on the analogy with deer that Montezuma was led to construct a tentative CT so imperfect, it rendered easy identification of horses impossible at the first encounter, so that he confused horses with the oxen in the train of the soldiers. 17
3.3.2.2 Instructions for Retrieval
There is another possibility: that the messengers did not successfully express the properties of the horse to Montezuma. They might have restricted themselves to telling him that some strange and terrible animals had appeared at a point on the coast, and that if he went to that place, he would see white men who wore iron trappings and moved by sitting with legs astride something; and this something was what the messengers were referring to. In this way they would have supplied Montezuma with instructions not for the identification but for the retrieval of the object.
The cases I am about to cite concern the CTs of individuals, of which I shall have more to say in 3.7.6, but in any event they serve to distinguish identification from retrieval. First case: Every evening I bump into a man in the bar; I recognize him every time, but I don’t know his name, and if I were to correlate an NC to the generic name man, it would simply be the description «the man I see every evening in the bar.» One day I see this man robbing the bank across the road. On my being questioned by the police, by means of verbal interpretations I help the specialized artist make a fairly accurate sketch of him.
I have supplied instructions for the identification of this person, and the police can elaborate a CT of him (albeit a vague one—with the result that there is a risk of their erroneously identifying someone else). Second case: Every evening I recognize a man in the bar, even though I have never observed him closely, but one day I hear him saying over the telephone that his name is George Brown and that he lives at number 15 London Road. One day this man argues with the bartender, whom he kills by smashing a bottle over his head, then he flees. The police question me as a witness, but I am quite unable to give the artist instructions for a sketch (at most I can say that the man in question is tall, with ordinary features and an unpleasant look), but I can supply his name and address. On the basis of my private CT, I cannot supply instructions for identification; but on the basis of the NC that I associate with the name George Brown (a being of the male sex who lives at 15 London Road), I am able to provide the police with instructions for his retrieval.
3.3.3 Molar Content (MC)
When, having seen horses in the flesh and having talked with the Spaniards, Montezuma acquired other information about horses, he could have reached the point where he knew what a Spaniard knew about them (though not as much as a zoologist knows today). In this case he would have had what is called a complex knowledge of them. Note that I am talking not about an «encyclopedic» knowledge, in the sense of a difference between Dictionary and Encyclopedia (to which I shall return in 4.1), but about «broadened knowledge,» which includes notions that are not indispensable for perceptual recognition (e.g., that horses are reared in such and such a way or that they are mammals).
With regard to this broadened competence I shall talk of Molar Content (MC). The format of Montezuma’s MC might be different from that of his first messengers or his priests, and it would be in continuous expansion. We cannot be sure how it evolved—we need only think of the fact that in our times the MC of horse includes the information that this animal flourishes in the American continent (something that certainly could not have been said in Montezuma’s day). I would not identify the MC with knowledge exclusively expressible in propositional form, because it could include images of horses of various breeds or of different ages.
A zoologist has an MC of horse, and so does a jockey, even though the two areas of competence are not coextensive. It is on the level of MC that Putnam’s division of linguistic labor occurs, something I would prefer to define as a division of cultural labor. On the level of MC there ought to be generalized consensus, albeit with some fraying and gray areas (see 3.5.2). And since it is this area of consensus that constitutes the nucleus of the present discourse, I would tend to avoid considering the MC, which can assume different forms depending on the subject and represents portions of sectorial competence. Let us say that the sum of the MCs coincides with the Encyclopedia as a regulative idea and a semiotic postulate, as is said in Eco 1984, 5.2.
3.3.4. NC, MC, and Concepts
On reading the first version of these pages, someone asked me the difference among NC, MC, and concept. I would not know how to answer the question before I resolved two cases: (i) What is the difference between the cognitive type of the platypus constructed by its first discoverer and the concept of platypus that he obviously could not have had previously, not even in the case of an overpopulated Platonic universe? (ii) What is the difference between the first Aztecs’ concept of the horse and a zoologist’s concept of the same animal?
For the first question, it strikes me as evident that, right from the Kantian idea of a schema for empirical concepts, it was clear that, if a concept existed, it ought to have been mediated by the schema. But if we introduce the schema, then there is no need for concept—and let proof of that be the possibility of constructing schemata for concepts we do not have, like that of the platypus. Therefore the idea of concept becomes a perplexing residue.
For the second question, if by «concept» we mean a mental conception, as etymologists would have it, there are two answers: either the concept governs perceptual recognition, and in that case it is the same as the CT and is expressed not by the classic definition but by the NC; or it is a rigorous and scientific definition of the object, and in that case it is the same as a particular sectorial MC.
It seems
