A curious situation, but not that dissimilar to the one in which I had put Galileo and Tycho Brahe, intent on looking at the same sun, acknowledging they were seeing the same thing and nevertheless obliged, in terms of the MC they attributed to the term sun, to define it in different ways within the framework of a different system of assumptions.
But although they recognize the brick house as the same one as before, do the citizens really perceive it in the same way? In New York today, dwarfed by the skyscrapers on Fifth Avenue, the neo-Gothic churches with their tall spires that once seemed towering affairs now strike us as minuscule, almost miniaturized. In the same way, how will this handsome and majestic house, so imposing when it was built, appear now, set against the skyscrapers of the new center that have sprung up around it? And that’s how, on the one hand, an object does not change and is always perceived as such, and how, on the other hand, by virtue of the town plan system of which it is a part, is seen differently.22
The principle is also taken up again in Quine (1995: 43 ff.): although they are dependent on perceptual stimuli, observation sentences “change and develop with the growth of scientific knowledge.” The parameter of an observation sentence is given, as well as experience, by the “pertinent linguistic community.” It is “public
Figure 4.2 (Vanville 1997)
pressure” that obliges the subject to correct the observation sentence Look at the fish when faced with a whale.
Let us try to reformulate the question in the terms used by Putnam (1987: 33): “There are ‘external facts,’ and we can say what they are. What we cannot say—because it makes no sense—is what the facts are independent of all conceptual choices.” Let there be three spatiotemporal points x1, X2, and x3: how many “objects” are there? In a world a la Carnap there would be three objects (x1, x2, and X3); in a world according to the Polish logicians, there would be seven objects (x1, X2, X3, x1 + X2, x1 + X3, X2 + X3, xi + X2 + X3). The number of identifiable objects changes according to the conceptual framework. Yet (and I stress this) we recognize as an initial stimulus three spatiotemporal points, and in the absence of that agreement on the initial stimulus the debate on identifiable objects could not even get started. Not only that, the two universes would not be comparable.
That two systems are structurally incommensurable does not mean that their two structures cannot be compared, and the two maps of Vanville we have been playing with until now demonstrate this.
We are able to understand the two systems, and we are able to understand what it means when both cities contain the same brick house. On this basis we can certainly understand that the instructions (i)-(3) that held good for Vanville 1951 no longer hold good for Vanville 1997. Yet by taking the first map and checking the meaning of the expression Tully Road, we are able to establish from the second map that that content now corresponds to two different urban entities, nameable as Tully Road and Cicero Road
.
This allows us to say that, if we found a treasure map from Vanville 1951, which said that starting from the corner of Elm and Giorgione, before turning southwest along Tully Road, three meters before the corner of Midtown Square, on the right we would find a buried chest of Spanish doubloons. In Vanville 1997, this sentence would translate as “Starting from the corner of East Elm and Giorgione, before turning southwest along Cicero, three meters before the area occupied by Lake Barbarelli, on the right, we will find a chest of Spanish doubloons.” The interesting aspect of the business is that, by negotiating the criteria of reference and the criteria of translation between two systems considered incommensurable, we might really find those doubloons.
One of the more diverting problems to be found in old (and sometimes new) Italian translations of hard-boiled American crime fiction is that the detective often gets into a cab and says “Portami nella città bassa” (literally, “Take me to the lower city”). Sometimes he asks to be taken to the “upper city.” The Italian reader immediately thinks that all American cities are like Bergamo (which is divided into upper Bergamo and lower Bergamo) or like Turin, Florence, Budapest, or Tbilisi, with one part of town on the plain and, across the river, another part on the hills. Obviously it is not like that. In the English text the detective asks to be taken downtown (or uptown).
But let’s put ourselves in the translator’s shoes, remembering that often he or she has never set foot in the United States. How must these terms be translated? If the translator asked a native for an explanation, the native would tell him that “uptown” and “downtown” are concepts that change from city to city: sometimes they mean the business center, sometimes the red-light district and therefore the oldest part of the city, sometimes the area along the river, according to how the city developed (in New York these concepts are occasionally absolute—and so Wall Street is certainly downtown—and occasionally relative, so if you want to go to the Village from Central Park you tell the driver to go downtown, while if you wanted to go there from Wall Street, you would tell him to go uptown).
Solutions? There is no rule, but the translator would need to know in which city the story is set. Then he would have to look at the map (and consult a good guidebook), understand what the detective is going to do (visit a gambling house, a five-star hotel, a seedy dive, or to find a ship), and then have him tell the cab driver each time to take him to the center, to the business quarter, to the old town, to the port, or wherever the devil it is he wants to go. The referent for downtown is to be negotiated, to the extent in which the meaning is negotiable, according to the city (to the system).
The chance of an observation sentence’s being true is also a matter for negotiation. But this does not alter the fact that the observation sentence is based on perceptual evidence, on the fact that that brick house was built after all, and that in some way it is perceived even by a dog who knows nothing about the Vanville town plan. You can avoid noting its presence, but you cannot deny that it is there. However, the moment its presence is noted, it must be named and defined, and this cannot be done except within the context of the city as a system.
4.7 Contract and Meaning
By now I think it is clear that all this presupposes a contractual notion both of CTs and of NCs and MCs. I dealt elsewhere (Eco 1993) with the various attempts made over the centuries to construct (or rediscover) a Perfect Language. Most of these attempts were based on the assumption that it is possible to identify a series of primitive notions, common to the whole species, and arrange them in an elementary grammar, so as to construct a metalanguage in which the notions and propositions expressed in any natural language are entirely translatable, under all circumstances, and in a way devoid of that ambiguity proper to our mother tongues. Why, given that I have mentioned semiosic primitives and CTs connected with perceptual experience, might it not be possible to construct such a perfect language on this basis, which today might even assume the form of a mentalese that explains the way in which the human mind works and the way in which a silicon-based mind might humanly work?
Because, I think, it’s one thing to proceed in the course of our experience by elaborating CTs and NCs, but it is another thing to say that these entities of ours are really universal and metahistoric in their format. You cannot construct a Perfect Language, because it would exclude that moment of negotiation that makes our languages efficacious.
Everybody finds himself more or less agreed regarding to the recognition of a rat, but not only is the zoologist’s competence different from mine, the zoologist must also continually check to see whether his NC has the same format as mine. Is the fact that rats are carriers of disease a part of the NC of rat? It depends on the culture, the circumstances, and naturally the age. In the seventeenth century, people had still not associated rats with plagues, but they do today, and when pestilence strikes now, anyone, before perceiving the rat as a quadruped, would perceive it as a threat.
The CT and NC are always negotiable; they are sort of “chewing-gum” notions that assume configurations that vary according to circumstances and cultures. Things are there, with their invasive presence; I don’t think there is a culture that can induce one to perceive dogs as bipeds or feathered creatures, and this is a very strong bond indeed. But apart from that, meanings scatter, dissociate themselves, and reorganize. Even the so-called “dispositional” properties give us serious reasons for doubting whether the proposition Sugar is soluble (in whatever language it is uttered) is the same when it is expressed in Latin America (with reference to brown cane sugar) or in Europe (with reference to white beet sugar). This “solubility” requires different times.
The same negotiability, as has been demonstrated by the story of the
