As for isolated utterances, these highly improbable affairs (pronounced only in language laboratories) allow for no negotiation; only autistic subjects exchange fragments of their extremely private idiolect, asserting that men can be unmarried without being bachelors, that elephants can or cannot have trunks. But to negotiate with my patient readers that it is possible to say truthfully that Ayers Rock is a mountain and at the same time that Ayers Rock is not a MOUNTAIN, I had need of lengthy argumentation in textual form, and I could not rely on my interlocutor’s goodwill, or on her charity—to be hoped for—toward me.
This brings me handily to some thoughts on the cartoonist Peyo’s little blue gnomes, known in English as the Smurfs, which were originally called Les Schtroumpfs.27 The characteristic of the Schtroumpf language is that in it, as often as possible, proper and common nouns, verbs, and adverbs are replaced by conjugations and declinations of the word schtroumpf.
For example, in one of the stories a Schtroumpf decides to rise to power and launches a political campaign. His speech sounds like this:
Demain, vous schtroumpferez aux urnes pour schtroumpfer celui qui sera votre schtroumpf! Et à qui allez-vous schtroumpfer votre voix? A un quelconque Schtroumpf qui ne schtroumpfe pas plus loin que le bout de son schtroumpf? Non! Il vous faut un Schtroumpf fort sur qui vous puissiez schtroumpfer! Et je suis ce Schtroumpf! Certains—que je ne schtroumpferez pas ici—schtroumpferont que je ne schtroumpfe que les honneurs! Ce n’est pas schtroumpf!…C’est votre schtroumpf à tous que je veux et je me schtroumpferai jusqu’à la schtroumpf s’il faut pour que la schtroumpf règne dans nos schtroumpfs! Et ce que je schtroumpfe, je schtroumpferai, voilà ma devise! C’est pourquoi tous ensemble, la schtroumpf dans la schtroumpf, vous voterez pour moi! Vive le pays Schtroumpf!
The Schtroumpf language seems to lack all the requisites of a working language. It is a language devoid of synonyms and full of homonyms, more than a normal language could handle. But not only do the Schtroumpfs understand one another perfectly well, the reader does too, and that is what counts.
This would seem to work in favor of Davidson’s position. The Schtroumpfs do not speak in a void (they do not utter sentences outside any given situation) but in the context of a comic strip, and therefore in a multimedial context, where we not only read (or hear) what they are saying but also see what they are doing. But this is the situation in which we usually interpret other people’s words—and it is because we speak within a situation that we are able to apply the deictics, such as this or that. Therefore it might be said that, on hearing that burst of homonyms in a given situation, we attribute the speaker with the same beliefs we would cleave to in the same situation, and by the principle of charity we lend him those terms he has not uttered but which he could or should have uttered.
Or we might say (as Wittgenstein might have said) that in the Schtroumpf language the real meaning of the term is its use (obviously I am referring not so much to the Schtroumpfus Schtroumpfico-Schtroumpficus as to the Schtroumpfische Unter-schtroumpfungen).
But here two objections arise. The first is that we “lend” or attribute to the speaker the terms he has not uttered precisely because these terms (with their conventional meaning) preexist in our lexicon. If readers understand my joke about the Schtroumpf and Wittgenstein, it is because they have already heard the original titles mentioned. We can negotiate a contract only because a predefined semiotic (intertextual) system already exists, in which the various expressions have a content.
In the second place, the electioneering speech quoted earlier does not refer to the perceivable situation (i.e., to what the picture shows). It refers to the “political speech” scenario and its rhetoric. It refers to a large quantity of utterances we have heard in analogous situations and therefore to the universe of intertextuality. An expression like Un quelconque Schtroumpf qui ne schtroumpfe pas plus loin que le bout de son schtroumpf is understandable, because we know the stock phrase He cannot see any farther than his own nose. An utterance like Je me schtroumpferai jusqu’à la schtroumpf can be decoded, because we have heard on an infinity of occasions I will fight to the death, and we have heard it within the context of the rhetoric of deliberative speech. La Schtroumpf dans la schtroumpf is understandable, because we have heard hand in hand thousands of times.
This means that the Schtroumpf language responds to the rules of a linguistics of the text, where the sense depends on the identification of the textual topic. It is true that (see Eco 1979) every text is a lazy machine that requires active interpretative cooperation on the part of its receiver, and this laziness seems to invite us to make texts in Schtroumpf. Our collaboration is possible, because we appeal to the universe of intertextuality, and we can understand Schtroumpf, because all speakers use the term schtroumpf and its derivatives always and solely in those contexts in which a phrase of that kind has already been pronounced.
The Schtroumpf language is a parasitic language, because, although nouns, verbs, and adjectives are replaced with the all-purpose homonym, it would not be understood if it were not backed up by the syntax (and the various lexical contributions) of the base language (be it the original French or its translations). Now, in one of the stories we meet the enemy of the Schtroumpf, the wicked wizard Gargamel. He speaks the same French on which Schtroumpf is based, but normally. Gargamel casts a spell to change himself into a Schtroumpf and goes to the village of his little enemies.
But he has to restrict himself to sidling along walls without replying to those who ask him why (we are told) he does not know the Schtroumpf language. How is this possible, if we have seen that the base language is the same as his, and he could interpret what the Schtroumpf say to him if only he applied the principle of charity? The fundamental rule of Schtroumpf is: Replace every term of the common language with schtroumpf as often as you can without excessive ambiguity. But Gargamel’s problem is clearly that he finds all contexts ambiguous, or incomprehensible, for the simple reason that he has no intertextual information.
Let us suppose that an English speaker of average culture hears a Schtroumpf poet reciting I schtroumpfed lonely as a schtroumpf. He would certainly grasp the reference to Wordsworth. Obviously he would also grasp the reference to Shakespeare in To schtroumpf or not to schtroumpf. But he might well be stuck on hearing Schtroumpf is the schtroumpfest schtroumpf, because he may never have read T. S. Eliot before and may not know that some months are crueler than others. He would find himself in the same situation as Gargamel.28
Every application of the principle of charity to what someone is about to say is based on a modicum of lexical information but, above all, on a vast amount of information about what has already been said.
Chapter Five
Notes on Referring as Contract
After having spoken of meaning as contract, one is tempted to see if the notion of contract/negotiation might not also apply to the phenomenon of referring, and if so, to what extent.
It is no accident that the paragraphs of this essay are not numbered: this is precisely to exclude even the slightest suspicion that my discourse harbors any systematic ambitions. The question of referring, in all its ramifications, is one that would put the fear of God into even the strongest among us. Here I have restricted myself to a series of problematic observations, which throw light on some reasons why it is convenient to think that operations involving referring have a contractual nature—or at least a strong contractual component.
In Eco (1976: 163) I accepted Strawson’s (1950) proposal, whereby mentioning or referring is not something an expression does but something that someone can use an expression to do.
Strawson went on to say, “[To] give the meaning of an expression … is to give, general directions for its use to refer to or mention particular objects and persons” and, “[To] give the meaning of a sentence is to give general directions for its use in making true or false assertions.” I still think that this is a satisfactory arrangement and that referring is a linguistic act. Which does not alter the fact that it is very tricky to say what kind of linguistic act it may be and what are its conditions of felicitousness.
Between the meaning of an expression, which also provides instructions for the identification or retrieval of the referent, and the meaning of the sentence, which also ought to regard the expression’s truth value, what remains empty is precisely the space for referring.
Can we Refer to all Cats?
First of all, so that
