Violi (1997, 7.2) distinguishes between essential and typical properties: it is essential that the cat is an animal; it is typical for it to meow. The second property can be deleted, but the first cannot. But in that case we are back at the old difference between dictionary and encyclopedic properties. On the contrary, Violi (7.3.1.3) holds that even functional properties (closely connected to the CT by virtue of an affordance typical of the object) are indelible, and so it is hard to say that something is a box while denying that it can contain objects (if it could not, it would be a fa^e box).6
Let us examine the following sentences:
(1) Mice are not MAMMALS. This is what in Eco 1976 (3.1.2) I defined as a semiotic judgment, that is, an assertion that confirms or challenges encyclopedic or dictionary conventions existing within the bounds of a given language or, better still, an assertion on the current taxonomic paradigm. Within the bounds of the paradigm this assertion is certainly false, but many people would be able to recognize and name a mouse without knowing that (1) is false. It could be that (1) should be understood as “I assert on the basis of new factual evidence regarding their reproductive process that mice can no longer be classed as MAMMALS.” AS will be seen in 4.5, assertions of this kind were in circulation for eighty years regarding the platypus. The proof of their truth was in the first instance up to the researchers who were carrying out an empirical examination of the animal’s physiology and anatomy. But naturally it would be enough to change the taxonomic criterion to assert that the platypus is not a MAMMAL. In any case the assertion (1) does not refer either to the CT or the NC of the mouse; it is not a part of that area of common competence of which we spoke in 3.5.2. If anything, it is part of the MC: let the zoologists decide what is more or less worthy of deletion as far as they are concerned.
(2) Mice do not have tails. If the assertion were understood as being supported by a universal quantifier referring to all the mice in existence, it would be sufficient to provide at least one mouse with a tail to falsify it. However, in everyday life, I think it unlikely that anyone would make such an assertion, which would presuppose that the speaker had made a previous inspection of all mice (billions of them) one by one. This utterance should simply be transcribed as “The property of having a tail is not part of the CT of the mouse or of the NC of mouse” We have seen that NCs are publicly verifiable, and I would say that it would be easy to challenge assertion (2). Those things we call mice (usually) have a tail; the stereotype of the mouse has one, as does its prototype, if the prototype exists somewhere or other. It seems improbable that someone would say (2), but it was possible, as we shall see, for someone to say that the female platypus has no mammae (it was a case of a CT-in-process). Now, is the property of having a tail cancelable or not? I think the question is badly put: when a CT is interpreted, all the properties have the same value at first, also because we still have to know to what extent the type is really wholly shared by all speakers. The acid test is when a token is recognized. Which brings us to the next example.
(3) This is a mouse, but it has no tail. It is possible to find a dead mouse and recognize it despite its mutilation. Our CT of the mouse also envisages the characteristic tail, and yet this is a cancelable property.
(4) This is a mouse, but it is not an animal. Here we must refer back to what has already been said: the attribution of animality has nothing to do with ascription to a category; we are dealing with a perceptual primitive, a precategorial experience. If this is not an animal, it cannot be a mouse (it must be the usual robot mouse that is chased through many pages of the philosophy of language by a robot cat). The property of being an animal is indelible.
(5) This is a mouse, but it has the sinuously cylindrical shape, tapered at the extremities, of an eel. Granted that someone, without having to take a philosophy of language exam, were so foolish as to utter (5) seriously, we would be unlikely to agree. An almost oval shape slightly tapering toward the nose is part of the indispensable (and indelible) conditions for recognizing a mouse. The importance of this gestalt is such that we can be flexible about the tail, and flexible even with regard to the presence of the paws. The gestalt of the mouse, once it is perceived, allows us to deduce the paws and the tail (if mouse, then tail).7 The presence on the ground of four little paws or a tail, on the other hand, allows us to infer the mouse that is not there only by abduction. In this case, we behave like the paleontologist who takes a jawbone and reconstructs a cranium, but precisely because he is referring to a CT, albeit a hypothetical one, of that prehistoric being. This amounts to saying that, if a 3-D model is part of a CT, it plays a role so important for recognition and identification, that it cannot be deleted.
(6) This is a mouse, but it is eighty meters long and weighs eight hundred kilograms. No one can exclude that, after some tinkering with the genetic code, this assertion might one day be utterable. But in such a case I would say we would be talking about the appearance of a new species (we shall call them mice2 as opposed to normal mice1). It should suffice to think of the different tone with which the assertion There is a mouse in the kitchen would be uttered depending on whether the reference was to a mouse1 or a mouse2. This means that the CT of the mouse also includes standard dimensions that, no matter how negotiable, may not go beyond a certain threshold. Let us recall the question put by Searle (1979), i.e., why is it that when we go into a restaurant and ask for a hamburger, we do not expect the waiter to serve a hamburger a mile long enveloped in a plastic cube? It is curious that not long after the formulation of this example, an American restaurant chain prepared a manual for its chefs containing specifications regarding the size, weight, cooking times, and condiments required for a standard hamburger; this was not in reply to Searle but because it was economically and industrially important to make public the standard concept of hamburger.
Naturally this manual was the elaboration not only of a CT but of an MC of the term hamburger: however, it established the nuclear conditions for recognizing something as a hamburger, if not with regard to specifications of weight and cooking times at least with regard to its dimensions and approximate consistency. This is therefore why a property such as standard size seems, if not indelible, at least hard to delete. Of course, saying that there is an eight-hundred-kilogram mouse is less of a problem than saying there is a mouse that is not an animal, but while in order to justify the first statement, it has to be admitted that we are dealing with a fake mouse, to justify the second, it is necessary at the very least to postulate a wholly improbable world, and so this mouse, if it is not fake, must be at least fictional or fictitious.
(7) This is an elephant, but it has no trunks Here we have to distinguish between the proposition This is an elephant, but it no longer has a trunk (similar to the case of the mouse without a tail) and the assertion that a given animal is an elephant that nonetheless has no trunk but a snout made in some other way (like the snout of a kangaroo or the beak of an albatross, let’s say). I think that each of us would react by maintaining that in such a case we are dealing no longer with an elephant but with some other animal. We can imagine a breed of mouse without a tail, but the idea of a breed of elephant without a trunk is unconvincing. This case is in fact similar to (5).
The trunk is part of the characteristic gestalt of the elephant (more so than the tusks; don’t ask me why, but try to draw an elephant with its trunk and no tusks, and other people will usually recognize it; but if you draw a beast with tusks and the round snout of the porpoise, no one will say you have drawn an elephant). At best we can say that the presence of the trunk is not sufficient for the recognition of an elephant, because the trunk could also belong to a mammoth, but there is no doubt that its absence eliminates the elephant. It is an indelible
