CTs may be in the mind, in the brain, in the liver, in the pineal gland (were the pineal gland not already occupied, these days, by melatonin); they could even belong to an impersonal warehouse, packed into some universal active intellect, whence a miserly divinity takes them and doles them out to me, out of occasionalism, every time I need them (and the cognitive scientists who spend their lives questioning subjects that cannot tell a glass from a plate will have to decide why some of those subjects’ cerebral areas are no longer in tune with the divine wavelength).
But we must start from the principle that, if there are felicitous acts of reference, it is because, both in recognizing a second time something perceived previously and in deciding that object A and object B can satisfy the requisite of being a glass, a horse, or a building—or that two forms are both definable as right-angled triangles—we relate tokens to a type (whether it be a psychic phenomenon, a physically existing prototype, or one of those Third World entities that philosophy has always tried to account for, from Plato to Frege, from Peirce to Popper).
Postulating CTs does not even oblige us to decide beforehand whether they assume, wholly or in part, the configuration of a mental image or are simply computable and processable in terms of discrete symbols. It is well known how this debate between iconophiles and iconophobes is today a central issue for cognitive psychologists. We might restrict ourselves to a summary of the Kosslyn-Pylyshyn polemic:12 on the one hand, forms of mental representation of the iconic type seem indispensable if we are to explain a whole series of cognitive processes with respect to which propositional explanations emerge as insufficient, and this hypothesis would also seem confirmed by computer simulations; on the other hand, mental imagination would seem to be a simple epiphenomenon, explicable as an elaboration of information accessible only in digital terms. From this standpoint, therefore, mental images are not incorporated into our hardware but are only secondary output.
Now, it could be said that love does not exist on a neural level and that falling in love is an epiphenomenon based fundamentally on complex physiological interactions, which might one day be expressible through an algorithm. This does not stop the epiphenomenon «falling in love» from being central to our personal and social life, to art and literature, to morals, and often even to politics. The result is that a semiotics of the passions does not wonder about what happens in our hardware when we feel hate or fear, anger or love (even though something worth investigating certainly does happen), but about how it happens that we recognize them, express them, and interpret them—so that we understand perfectly well what it means when Orlando is described as furioso rather than in-namorato.
Semiosic experience tells us that we have the impression that we retain mental images (even if a mind does not exist), and above all that we interpret many terms publicly and intersubjectively through visual representations. And so the iconic component of knowledge must be postulated as well as the existence of the CT, if we are to account for what common sense proposes to us. Images are as much systems of instructions as they are verbal devices, and if I have to tell someone how to get to Main Street, I can supply him with lengthy verbal indications regarding the streets he must take as well as show him a map (which is not an image of Main Street but a diagrammatic procedure that makes it possible to find it). Which of the two procedures is the better depends on the capacities and disposition of my interlocutor. 13
My refusal to stick my nose into the black box might be interpreted as a confession that philosophy (and in this specific case general semiotics as a philosophy) is an «inferior» form of knowledge with respect to science. But this is not the case. We can postulate the CTs in the black box precisely because we can have an intersubjective check on what constitutes their output. We have the instruments with which to talk about this output—and this is perhaps the contribution that semiotics can make to the cognitive sciences, that is to say, to the semiotic aspect of the cognitive processes.
3.3.2 From CT Toward Nuclear Content (NC)
Whereas at first the Aztecs might have felt that their CT was private, as soon as they all began indicating the same animals by pronouncing the name maçatl, they must have realized that, on the contrary, the CT had established an area of consensus. At first the area of consensus could be postulated only to explain the fact that they understood one another by using the same word. But bit by bit they must have proceeded to collective interpretations of what they understood by that word. They associated a «content» with the expression maçatl.
These interpretations were as similar as we can imagine to a definition, but we certainly cannot think that our Aztecs had said to one another that by maçatl they meant, as an encyclopedia might put it, «a browsing herbivorous mammal of the family Equidae, order Perissodactyla, with a highly developed middle toe of the foot enveloped by a protective nail (hoof).»
At first this agreement must have taken place as a disordered exchange of experiences (some would have pointed out that the animal had hair along its neck, some that the hair in question swirled in the wind when the animal was galloping, while someone must have been the first to note that the trappings were extraneous to the creature’s body, and so on). In other words, the Aztecs gradually interpreted the features of their CT, in order to homologate it as much as possible. While their CT (or CTs) could have been private, these interpretations were public: if they had written them down, or had used pictograms, or if someone had tape-recorded what they said to one another, we would have a verifiable series of interpretants.
As a matter of fact we have them, in the sense that indigenous testimony is still extant, and while we do not know exactly what went through the heads of the first Aztecs when they first saw horses, it is only because we have reason to suspect that the testimony is decidedly tardy, interpretations of the interpretations that the conquistadors had made of the Aztecs’ first behavioral responses. But if these interpretants were available integrally, as is the case with the scientists who saw a platypus for the first time, the Aztecs would not only make clear what their CT was but also circumscribe the meaning they assigned to the expression maçatl.
We shall call this set of interpretants the Nuclear Content (NC).
I prefer to speak of Nuclear Content rather than Meaning, because by time-honored tradition one tends to associate meaning with a mental experience. In certain languages the confusion is greater than in others, and we need think only of the English word meaning, which can stand for «that which exists in the mind» but also for an intention, for what is recognized as being, for what is denoted or understood, for sense, signification, and so on. Nor should we forget that meaning can also appear as a form of the verb to mean, which is variously defined as to have in mind, to intend, to stand for, and only in a few cases does it come to denote a socially recorded synonymy (the example given in Webster is «the word ja in German means yes»). The same variations of sense are found also in the German verb meinen. As far as Italian is concerned, even though the term significato is more often understood as «a concept expressed by a sign,» the pair significato and significare can be used for the expression of thoughts or sentiments, for the emotional effect caused by an expression, for the importance or value that something assumes for us, and so on.
On the other hand, the term content—in Hjelmslev’s sense, as the correlate of an expression—is less compromised and may be used, as I shall use it, in a public sense and not a mental one. Once this has been made clear, when the requirements of argument of some current theory may encourage me to do so, I shall use the word meaning, but only as a synonym of content.
In certain cases CT and NC can practically coincide, when the CT wholly determines the interpretants expressed by the NC and the NC makes it possible to conceive an adequate CT. Nevertheless I wish to make clear yet again that the CT is private, while the NC is public. We are not talking about one and the same phenomenon (which some might call generically «the Aztecs’ competence regarding horses»): on the one hand, we are talking about a phenomenon of perceptual semiosis (CT) and, on the other, about a phenomenon of communicative consensus (NC).
The CT—which cannot be seen and cannot be touched—may be postulated only on the basis of the phenomena of recognition, identification, and felicitous reference; the NC represents the way in which we try intersubjectively to make clear which features go to make up a CT. The NC, which we recognize in the form of interpretants, can be seen and touched—and this is not just a metaphor, given that the interpretants of the term horse include a great
