All a culture does, then, is to select the data for its own memory. It may not do what Stalin did when he erased from historical photographs the faces of the comrades he had sent to their deaths, or what Orwell’s Big Brother did when he corrected the news in The Times every morning. But when we read that some English secondary schools have proposed abolishing the teaching of the Crusades so as not to offend the sensibilities of their Muslim students, it becomes apparent that culture is a continual process of rewriting and selecting information.
1.9.5. Cancelation, Cross-reference, Latency
Still, there is a difference between the Plinian and medieval encyclopedias and the structures of a modern Median Encyclopedia. The first premonitions of this change can already be seen in the encyclopedias of the sixteenth and seventeenth centuries. Let us recapitulate: an Arbor Porphyriana aspired to provide a definitive image of the Great Chain of Being; and that image (had it been exhaustive and had it included all of the beings in the universe—a hypothesis that was formally impossible, as we have seen) would have been definitive in the sense that all its nodes appeared to be primitives. When one knows that a man is Animate, one knows intuitively all one needs to know, and there is no need for any science to define what Animate is to distinguish it from what is Inanimate (even though medieval science often does so).
Similarly, when Pliny’s encyclopedia or those of Rabanus Maurus or Honorius of Autun explain to us the “nature of things” or the “image of the world,” they assume that they have told us all we need to know, to such a degree that we need a well-trained art of memory to remember it.53
The form of the modern encyclopedia, on the other hand, is that of naturalistic classification in which, if we say that a horse is an Ungulate, this taxonomical node is understood as a link (exactly in the hypertextual sense of the term) that refers us to a repository of specialized knowledge—and it is there that the properties of the ungulates will be specified (see, in this connection, Eco 1997, 3 and 4).
In this sense it has been said that in speaking of the modern encyclopedia, more than of forgetting, it is appropriate to speak of the “latency” of knowledge (Cevolini 2006: 99). It is not as if the information in excess (the object of Specialized Encyclopedias—and even the information in excess vis-à-vis a Specialized Encyclopedia, such as, for example, the history of astronomical theories proven to have been erroneous) is actually forgotten. It is, so to speak, “frozen,” and all the expert has to do is to take it out of the freezer and put it in the microwave to make it available once again, at least as much as is needed to understand a given context. This latency is represented by the model of the library or the archive (or even the museum)—containers always available even though no one may currently be using them, and even if they haven’t been used for centuries (see Esposito 2001, ch. 4 especially paragraph 4.4).
If we now return to paragraphs 1.3.5 and 1.3.6 we will see how both Wilkins and Leibniz anticipated these techniques of latency that constitute the form that modern cultures came up with to get around the Vertigo of the Labyrinth.
1.9.6. The Maximal Encyclopedia and Virtuality
In this sense every encyclopedia refers back to ever vaster portions of knowledge, through a series of cross-references that has been defined as virtual. In the background is the truly virtual encyclopedia, the Maximal Encyclopedia. The Maximal Encyclopedia is virtual in nature, not only because we never know where it stops; the fact is that it contains potentially even what it in fact (today) no longer contains.
We remarked that the Median Encyclopedia does not record the names of all those who fought in the battle of Waterloo. But what would happen if a scholar wanted to reconstruct that list today? Let’s say he has access to archives that have remained unexplored until now, or that he acquires a document similar to the catalogue of the Thousand, the volunteers who sailed from Quarto to Sicily with Garibaldi in 1860 (now readily available even on Wikipedia). That scholar would be exploiting forgotten and repressed portions of the Median Encyclopedia that are still part and parcel of the Maximal Encyclopedia.
We know that in his Poetics Aristotle cites tragedies of which no record survives. What encyclopedia do these works belong to? For the present only the fact that Aristotle cited the mere title of these works forms part of the Median Encyclopedia (or at least of a Specialized Encyclopedia). If one day (as was the case with the Gnostic texts of Nag Hammadi) some of these plays were to be discovered buried in a jar in the desert, they would have already been part of the Maximal Encyclopedia, even if no one up till then could have claimed so, while from that time on they would be part of one or more Specialized Encyclopedias. But what would happen if on the other hand they were never found and our knowledge of them continued to be limited to an acquaintance with their titles?
For the very fact that there are good reasons to believe they once existed, we would continue to think that they might form part of the Maximal Encyclopedia, even though for the moment they belong to it only in a virtual and optative fashion–or else that they are part of it but only in the possible world in which they have been discovered, or that they were part of the Median Encyclopedia of Aristotle’s day.
The Maximal Encyclopedia, then, despite the fact that its name we have been giving it suggests that, to quote Anselm, it is something quo nihil majus cogitari possit (“than which something greater cannot be thought”), is in fact an accordion-like structure, and one day it could expand beyond anything we dream of today. Which offers no small encouragement to future research.
1.9.7. The Text as Producer of Forgetfulness
At this point, we understand how, every time we construct a local “ontology” in order to disambiguate a proposition in a given context (as we observed in paragraph 1.7), we are performing ad hoc the same operation that a culture performs in constructing its own Median Encyclopedia. We prune, we narcotize, we eliminate some notions, retaining only those we consider pertinent.
How do we go about identifying—in our efforts to pinpoint the appropriate context—the notions to prune? We consider the context as if it was a text, and we behave exactly as we behave when we are trying to understand a text. A text (in addition to being a tool for inventing and remembering) is also a tool for forgetting, or at least for rendering something latent.54
Classical mnemonics could not be used for forgetting because a mnemonic technique is a mutilated semiotics. A semiotics in the Hjelmslevian sense is a system that—in addition to a lexicon—also contains rules for syntactic combination, and allows us to develop discourses, or, in other words, texts. A mnemonic technique on the other hand was more like a simple dictionary or a repertory of significant units that cannot be combined among themselves. A mnemonic technique did not facilitate the articulation of mnemotechnical discourses.
But if a mnemonic technique, insofar as it is a semiotics, cannot be used to forget, a semiotics that is not a mnemonic technique can produce forgetfulness or cancellation at the level of the textual processes themselves.
If in a semiotics the correlation is not based on simple automatic equivalence (a = b), but on a principle of inferentiality, however elementary (if a, then b), the meaning of an expression is a potentially huge package of instructions for interpreting the expression in different contexts and drawing from it, as Peirce would have it, all the most remote inferential consequences, in other words, all its interpretants. On these bases we ought then to know in theory every possible interpretant of an expression, whereas in practice we know (or remember) only the portion that is activated by a given context. Interpreting the expression in context means magnifying certain interpretants and narcotizing others, and narcotizing them means removing them temporarily from our competence, if only for the duration of the current interpretation (cf. Eco 1979, 1984).
If the interpretation of a sign, as Peirce maintained, always makes us learn “something more,” this something more (in a given context) is always learned by giving up something less, that is, by excluding all the other interpretations that could have been given of the same expression in another context.
If, as a matter of principle (and on the strength of the ideal global encyclopedia), knowing how many miles Paris is from Bombay is part of the meaning of the name Paris, when we are reading Les Misérables we learn many things about Paris, but we are expected to forget the distance (and to act as if we had forgotten it—if we already knew it) between Paris and Bombay.
There are many cases in which, in the course of the interaction between a reader and a text, instances of forgetfulness occur, encouraged in some way by the text itself. If, as I recalled in my Role of the Reader (1984), a text is a strategy that aims at stimulating a series of
