Bertuccelli Papi (1993: 197) gives the example of these two sentences: (i) Alice left yesterday and Sylvia three days ago and (ii) Alice left yesterday and Sylvia two days before. If the two sentences are uttered on Saturday, in both cases Alice must have left on Friday and Sylvia on Wednesday. But in (i) the expression ago refers to the day of the utterance (Saturday) while in (ii) the adverb before is anchored to the point of temporal reference contained in the sentence itself (yesterday). If we were to replace before with ago in (ii), Sylvias departure date would be shifted to Thursday. The au thoress suggests that ago is therefore «intrinsically deictic,» while before changes its value according to the temporal point of reference it stands in relation to. In any case it is clear that the use of the two expressions to designate a precise day depends on highly complex rules of textual linguistics, and I do not see why this set of rules cannot be understood as the content of the respective expressions—if by NC we do not mean a simple definition but also, or sometimes only, a complex set of instructions for identifying the referent.27
It has been said that «/ denotes the one who utters the sentence» is an insufficient instruction for identifying the referent, given that the referent changes according to the context and the circumstances and therefore does not represent the content of the pronoun I. Yet again we are confusing instructions for the identification of the referent with a way of fixing the reference. The instruction for identifying the referent of I is as generic as that for identifying the referent of interlocutor (a term that identifies different persons according to the situation of linguistic exchange), of assassin (given that Caesar’s assassin and Kennedy’s assassin refer to two different people) or even of cat (given that the instructions for identifying cats are certainly not sufficient to fix the reference of the cat I gave to Louis yesterday). To give instructions for identifying, in a variety of circumstances, the possible referent of a generic term is not the same as deciding, by pragmatic negotiation, how to fix the referent when referring to individuals.
Putnam (1981, II) admits that a pronoun such as I has no extension but a function of extension that determines extension according to context. I would agree to consider this function of extension as part of the NC of the pronoun, and we might admit that it is a matter of an instruction for identifying the referent in an act of reference. Putnam also says he would rather not identify this function of extension (which would be an intension in Carnap’s sense) along with meaning. But here (and I refer the reader to the discussion in 3.3.2, on the difficulties that can sometimes be caused by the term «meaning») he simply means to say, on the one hand, that this rule is an abstract function and, on the other, that it does not include all that we understand by the meaning of an expression, in the sense that both cube and regular polyhedron with six square faces—says Putnam—have the same intension and extension in all possible worlds but retain a difference of meaning.
In point of fact, the NC of a pronoun includes an instruction for identifying the referent (as an ability to apply a function of extension in a concrete fashion), and yet there is more to the content than that. I shall give a series of examples, which apart from anything else ought to provide grist for my contractual mill.
Let us suppose that someone says, I’m sorry, we can’t come tonight. If the content of we were wholly identified with an instruction for identifying the referent, we would be up against a tricky problem, because it would oblige us to identify a community of authors of the utterance, whereas we can identify only a single individual. But we also possess a pragmatic rule whereby someone may speak on behalf of the group for which he is, let’s say, the spokesman. And thus we go to search in the dialogical context to see whether a group had been designated previously. Finding that the speaker was invited to dinner along with his family, we know that the plural pronoun refers to the members of that family.
But there are also semantic-pragmatic rules. For example, the rule of the royal «we.» In such cases we know that a single person has the constitutional right to use the first person plural instead of the first person singular of the personal pronoun. But even when we know this, other contractual elements come into play. If a monarch today says, We feel tired, we know right away that he is using the royal «we» as a matter of etiquette, and therefore the we refers to him individually and the sentence is intended to express an inner state of his. If, on the other hand, the same monarch says, We confer upon you the Order of the Golden Fleece or Today we have declared war on Ruritania, he is expressing something that was not the general will until that moment but becomes it as soon as the sentence is uttered. Therefore in some way that we refers (willy-nilly) also to the subjects listening to him. According to the context, the receivers fix the reference of the pronoun in different ways.
Now let us suppose that a scientist writes, We cannot reasonably admit that the hole in the ozone layer has a decisive influence on the world’s climate. To whom does that we refer? Not to the members of her family, not to the subjects she does not have. However, for the meaning of we, an ideal dictionary ought to provide the contextual selection «can be understood as the authoritative plural, thanks to which a single speaker presents himself or herself as an interpreter of the scientific community, of sound reasoning or common sense.» At this point we can identify the referent in various ways: (i) there is a first reading that I would define as «rhetorical charity,» by which we recognize linguistic use as a stylistic mannerism, and we refer the we to the writer (we translate we as I, as if the writer was expressing herself in another language); (ii) there is a «fiduciary» reading, whereby we refer the pronoun to the scientific community (what the writer is saying is Gospel); (iii) there is a reading «of persuasion,» whereby we feel involved and think that in fact we, the readers, are obliged to be subjects who think that way about things.
There is finally a reading in terms of textual semiotics (not available to just any receiver), which leads us to reflect on what the writer—in using the authoritative plural—wanted to have us believe about her: not only has she made an explicit statement regarding a physical phenomenon, she has also implicitly presented herself as a subject entitled to speak in our name too, or in the name of a superior cognitive authority. I admit that this reading should have nothing to do with the phenomenon of referring: we are still referring to the author of the written text, even though we now see her in a different psychological light.
Yet it cannot be denied that a bias with regard to the writer (she wants to convince us by claiming an authority to which she has no right) can determine the way we referentially interpret that we. We can decide that she did not intend to use a stylistic mannerism with which to say I, that she in fact wanted us to understand that she intended to refer to the scientific community. This decision would involve an alethic judgment regarding the proposition she has expressed. Supposing we are convinced that the hole in the ozone layer does in fact influence the climate of the planet, and that every reliable scientist today has stated this. If this person meant I, then she said something false about a physical fact; if she meant we, she said something false about the opinions already expressed by the scientific community—or perhaps she wanted to deceive us on both counts.
Whatever the reading, the sense of the sentence changes as well as the lexical content of that we, which therefore does not boil down to the instruction for identifying the referent. Without a first tentative application of the instruction, it would not be possible to decide whether it was necessary to interpret the pronoun as an authoritative plural; but without the knowledge of that aspect of the content, it would still not be possible to apply the instruction, in any of the senses considered above.
Chapter Six
Iconism and Hypoicon
It may be that the moon, and likewise the rest of the universe, does not exist; or the Moon may be an image projected by some Berke-leyan divinity. But, even so, it would still count for something for us, and for the dogs who bay at it by night (Berkeley’s god has a thought for them too). We therefore possess a cognitive type of the Moon, and it must be a very complex one. We recognize it in the sky whether it appears full or only as a sickle, whether it looks red or as yellow
