From the Tree to the Labyrinth: Historical Studies on the Sign and Interpretation

by Umberto Eco

  In section 1.3.3 we alluded to Alsted’s notion according to which, with respect to the encyclopedia, individuals are like so many containers, each capable of holding a content commensurate with his or her receptive capacity, and none capable of containing in themselves the whole sum of knowledge. In any kind of communicative interaction, it is clearly necessary to presuppose and infer the format of the individual encyclopedia of the persons speaking to us, otherwise we would attribute to them intentions (and knowledge) that they do not have. This is basically why we so frequently bring into play the principle of charity. But as a rule (except when we are dealing with anomalous interlocutors such as a child, a foreigner from a remote and unknown culture, or a mentally challenged individual) out of considerations of economy we have recourse to what we consider to be a Median Encyclopedia. Though its extent is difficult to measure, a Median Encyclopedia is identified with the contents of a given culture.

  So, just as knowledge of the works of Plato (except for the Timaeus) were not part of the Median Encyclopedia of medieval culture, in which notions about Plato came from Neo-Platonic sources, in the same way part of our current Western encyclopedia is the idea that the Ptolemaic system was believed to be true in the past whereas today it is considered erroneous (without its having been forgotten). To sum up, we might say that the Median Encyclopedia is represented by an encyclopedia in the publisher’s sense of the word—and probably a mid-sized one-volume encyclopedia and not the thirty-two volumes of the 2010 Britannica.39

  The fact that it is a Median Encyclopedia does not mean that all of its contents are shared by all members of a given culture, but rather that it is shareable (we will examine in more detail later, in section 1.9.5, the concept of “latency” of information). Even an educated person may have forgotten (or never have known) the date of Napoleon’s death, but that person knows that the information is accessible and usually knows where to find it. This is why it is said that a cultivated person is not someone who knows the dates of the beginning and the end of the Seven Years War, but someone who can come up with them in a couple of minutes.

  The Median Encyclopedia cannot be identified with an extensive library containing thousands, even millions of volumes, because such a library, though it may not be commensurate with the Maximal Encyclopedia, nevertheless encompasses the Median Encyclopedias of exotic cultures, of past civilizations, and ideally all of the Specialized Encyclopedias, past and present. That library is instead merely an attempt to approximate the Maximal Encyclopedia, fatally incomplete because the Maximal Encyclopedia does not contain only those ideas that have been committed to the written word.

  The fact that, not just the Maximal Encyclopedia, but even the “parody” of it represented by a normal library should provoke the vertigo of a knowledge so exaggeratedly extensive that nobody could ever capture or contain it in their own individual memories, leads us to the problem of memory and forgetfulness; the problem, in other words, of the Vertigo of the Labyrinth.

  1.9.2. The Vertigo of the Labyrinth and the Ars Oblivionalis

  Since classical antiquity, the problem of the need to forget appears contemporaneously with the development of mnemonic techniques by which to commit to memory the maximum possible amount of information (especially in the centuries in which information was not as readily obtainable and transportable as it has since become, with the invention first of printing and subsequently of electronic devices). In De oratore (II, 74), for example, Cicero cites the case of Themistocles, who was gifted with an extraordinary memory. When someone offered to teach him an ars memorandi, Themistocles replied that his interlocutor would be doing him a greater service if he taught him how to forget what he wished to forget than if he taught him how to remember (“gratius sibi illum esse facturum, si se oblivisci quae vellet quam si meminisse docuisset”), inasmuch as he would prefer to be able to forget something he did not wish to remember than to remember everything that he had once heard or seen (“cum quidem ei fuerit optabilius oblivisci posse potius quod meminisse nollet quam quod semel audisset vidissetve meminisse”).

  Themistocles’s concern, incidentally, anticipates (and perhaps inspires) the anxiety of Borges’s Funes the Memorious, who remembered each and every detail of his experiences and perceptions, to such an obsessive and unbearable degree, down to the mere rustling of a leaf heard decades earlier, that he was practically unable to think.

  The problem of the excess of memory explains why one of the terrors of the practitioners of mnemonics was that of remembering so much as to confound their ideas and forget practically everything as a result. It seems, in fact, that at a certain point in his life Giulio Camillo had to excuse himself for his state of confusion and for the gaps in his memory, citing as an explanation his protracted and frantic application to his theaters of memory. On the other hand, in his polemic against mnemotechnics, Cornelius Agrippa (De vanitate scientiarum) claimed that the mind is rendered obtuse by the memorative art’s resort to “monstrous” images and, being so overburdened, is led to madness. Hence, subterraneously parallel to the fortunes of the ars memoriae, the reappearance from time to time of the phantasm of an ars oblivionalis (see Eco 1987a and Weinrich 1997).

  In the twentieth lesson (“Lettione XX”) of the Plutosofia of Filippo Gesualdo (1592), the “methods for oblivion” are reviewed.40 Gesualdo excludes mythical solutions such as drinking the waters of Lethe. He is aware that Johannes Spangerbergius in his Libellus artificiosae memoriae (1570)41 had already reminded his readers that people forget from corruption, that is, from forgetfulness of past species, by diminution (old age and sickness), and from ablation of their cerebral organs. Likewise, it is obvious that we can forget by repression and suppression, drunkenness and drugs, but all these are cases of natural occurrences that must be studied and are studied elsewhere.42

  Gesualdo, however, is intent upon developing an art of forgetting that employs the same techniques as an art of remembering,43 and he advises:

  First, having recited, and wishing to consign the images to oblivion; either in the daytime with closed eyes or in the dark and quiet of the night, you should go wandering with your mind through all the imagined places, evoking an obscure nocturnal gloom that hides all of them, and proceeding in this fashion, and going back a number of times with the mind and not seeing any images, every figure will soon disappear.

  Second, go peering into all the places with the mind, back and forth, and contemplate them empty and bare, as they were formed for the first time without any images in them, and this operation should be performed several times.

  Third, if the persons in the places are persistent, let them be seen by the mind from all angles many times, and let them be contemplated in the same way they were first set up, with bowed heads and dangling arms, without any additional images.

  Fourth, just as the painter pastes over and whitewashes his paintings to cancel them, so we too can cancel the images with colors painted over them. And these colors are white, green or black; imagining over the places white curtains, green sheets or black cloths; and going over the places a number of times with these veils of colors. And one can also imagine the places stuffed with straw, hay, firewood, merchandise, etc.

  Fifth, I consider it an excellent rule to put in place new figures; because just as one nail drives out another, so forming new images and putting them in the places already imagined, cancels the first images from our memory. It is true that it is necessary to imprint them with great care and mental effort and to repeat them frequently in the dark and quiet of the night, so that the intense and vivid idea of the second ones drives out the first Ideas.

  Sixth, imagine a great storm of winds, hail, dust, ruined buildings and places and temples, a flood that leaves everything in a state of confusion. And when this noxious thought has continued for a while and been repeated several times, finally go walking among the places with your mind, imagining the weather bright and calm and peaceful, and seeing the places empty and bare as they were formed for
the first time.

  Seventh, after the longest period of time possible, imagine an Enemy, terrible and fearsome (the more cruel and bestial and belligerent the better) who, with a troop of armed companions, enters and passes impetuously among the places and with scourges, cudgels and other weapons drives out the likenesses, assaults the people, shatters the images, puts to flight through doors and windows all of the animals and animate persons who were in the places,

  Until, after the tumult has passed and the ruin, seeing the places with a mind recovered from its terror, they will be seen bare and vacant as before. And if this destruction were to occur at the hands of hostile armies, such as the Turks and Pagans, it would be even more effective, because that terror confounds everything and sets everything upside down.44

  We do not know whether anyone ever put the artifices Gesualdo recommended into practice, but we are entitled to suspect that all these stratagems made it easier to remember than to forget whatever it was the practitioner wanted to forget, and to remember it with even greater intensity—as occurs when lovers try to blot out the image of the person who has abandoned them.45

  The physiological and psychological reasons why an Ars oblivionalis is impossible depend on the contiguity/similarity dialectic on which the classical mnemotechnical systems are founded. If object x has been imagined to be in contact with object y, or if object x is in some way homologous with object y, every time object x is evoked so is object y. But if this is how the Artes memoriae work, it is hard to see how one can imagine an object x that, when evoked, somehow acts on our cerebral cortex by canceling y. Jakobson (1956) described the structures of the different types of aphasia and how they manifest themselves, but he did not say how they can be produced artificially. It is no accident, however, if Jakobson, in order to explain the internal mechanics at least of a neurophysiological problem like aphasia, if not its causes, should appeal to the paradigmatic and syntagmatic axes, two categories from the linguistic and semiotic domain. This suggests that we could view the arts of memory (of forgetting as well as recalling) in semiotic terms.

  1.9.3. Mnemotechnics as Semiotics

  There can be no doubt that any mnemotechnical strategy belongs to the field of semiotics, if one accepts a definition of the sign as something that stands in the eyes of someone in place of something else in some respect or capacity. To associate in some way a y with an x means to use the one as the signifier and expression of the other. To make a knot in one’s handkerchief is certainly a semiosic strategy, as was the trail of white pebbles or beans that the character in the fairytale dropped making it possible for the children to find their way back out of the wood. These are two different kinds of strategy, because the knot in the handkerchief is an arbitrary sign for whatever one decides to associate with it, whereas the trail of pebbles institutes a vectorial homology between the sequence of pebbles and the path to be followed and stands for that specific path and not for any other possible path—but all this tells us is that different mnemotechical strategies call into play different semiosic procedures.

  The earliest Greco-Roman mnemotechnics present themselves as a sequence of empirical solutions based upon associations inspired by rhetorical criteria—relying, in other words, as Aristotle suggested, “on something similar or contrary or closely connected.”46 We have the hint of a system when these same classical mnemotechnics propose the organized institution of places, such as a memory palace or a city, even though the organic structure of the loci is often used to accommodate random series of res memorandae or things to be remembered. But the more elaborate systems certainly present themselves as a semiotics in the Hjelmslevian sense of the term, in other words as a system that posits a plane of expression, form, and substance correlated with a plane of content, form, and substance. Now, speaking of the form of the content implies speaking of a systematic organization of the world. And this is not all: in principle there is nothing that is constitutionally expression or content, given that, if, in a function of signs based on a system A, x is the expression of y, in another system B, y can become the expression of x—or, to give another example, nothing prevents us conceiving of two semiotics, in one of which visual images stand for sequences of letters of the alphabet while in the other letters of the alphabet stand for visual images.

  Mnemotechnics that exhibit some aspects of a semiotics are those systems in which: (i) on the level of expression there appears a system of loci designed to accommodate figures that belong to the same iconographic field and exercise the function of lexical units; (ii) at the level of content, the res memorandae are in their turn organized into a logical-conceptual system such that, if this system could be translated in terms of another visual representation, the mnemotechnic could function as the plane of expression of a second mnemotechnic whose contents would become the system of loci and images that made up the plane of expression of the first mnemotechnic.

  For example, in the Thesaurus artificiosae memoriae of Cosma Rosselli (Venice, 1579), the theater of planetary structures, celestial hierarchies, infernal circles, organized in detail, that he presents is at the same time a lexical system and an organization of the world. Rosselli’s mnemotechnics is a semiotics because what institutes something as expression and as content is its sign function, not the nature of the thing. Anything at all can become functive expressive or content functive. A frequently recurrent expression in Rosselli is e converso and its equivalents: x may stand for y or e converso. Incidentally, entities previously placed among the loci may be used as figures and vice versa.47

  The objection could be raised that many mnemotechnics are not semiotics in the Hjelmslevian sense because their planes are conformal: the correlation between unit of expression and unit of content is not between one term and another and is in any case not arbitrary. There exists an isomorphic relationship between the planes, and therefore for Hjelmslev these mnemotechnics, more than semiotics, would be symbolic systems. Take, for instance, the convention (found in a number of authors) by which the system of grammatical cases is associated with parts of the body—in which it is not arbitrary, at least in the author’s intentions, to associate the nominative case with the head, the accusative with the breast that can be beaten, the genitive and the dative with the hands that hold and offer, and so on. Nevertheless, it should not be a matter of excessive concern whether a mnemotechnic is a symbolic system. In the first place, because we have reached a point where we consider as semiotic systems, albeit with their own particular characteristics, systems that Hjelmslev would have seen as symbolic, giving up on analyzing their possible articulations. And secondly, because the conformity of the mnemotechnical planes is either doubtful or weak and ambiguous; while the presumed iconic relations that they bring into play are fairly debatable. Rosselli, for example, claimed that the correlation must be based on similarity, but he failed to exhaust the many ways in which one thing may be similar to another (“quomodo multis modis, aliqua res alteri sit similis”) (Thesaurus, p. 107). There was, for instance, a similarity of substance (the human being as the microcosmic image of the macrocosm) and of quantity (the ten fingers for the ten commandments), correlation by metonymy and antonomasia (Atlas for the astronomers or for astronomy, the bear for irascibility, the lion for pride, Cicero for rhetoric),by homonymy (the four-legged dog for the constellation of the Dog Star), by irony and contrast (the fatuous individual for the wise one), by vestigial traces (the wolf’s spoor for the wolf, the mirror in which Titus admired himself for Titus), by a word pronounced differently (sanguine for sane), by similarity of name (Arista for Aristotle), by genus and species (leopard for animal), by pagan symbol (the eagle for Jove), by peoples (Parthians for arrows, Scythians for horses, Phoenicians for the alphabet), by the signs of the zodiac (the sign for the constellation), by relation between an organ and its function, by a common accident or attribute (the crow for the Ethiopian), by hieroglyphic (the ant for foresight).

  At this point the criteria become so vague that, as many mnemotec
hnical theorists recommend, it is advisable to commit to memory the relationship that connects a place or a figure to a res memoranda. Which is tantamount to saying that practically all mnemotechnics were based on relationships chosen almost arbitrarily and were therefore, more than symbolic systems, semiotics, however imperfect. But however imperfect they might be, they were, nonetheless, tentative semiotics. To say that in a functive of a symbolic nature the correlation is badly formulated does not exclude the possibility that the functive be proposed as such.

  It is the semiotic nature of mnemotechnics that makes it impossible to construct an art of forgetfulness on the mnemotechnical model, because it is a property of every semiotic system to permit the presentification of absence. It is a venerable topos to recognize that that all semiotic systems are characterized by their ability to actualize, if only in the possible world circumscribed by our assertions, the nonexistent. This is why, as Abelard points out, the sentence nulla rosa est (“there is no rose”) actualizes to some extent, at least in the mind’s eye, the rose.48

  Worth (1975) is the author of an essay entitled “Pictures Can’t Say Ain’t” in which he argues that no image in a mnemotechnic can act by canceling out what it refers to (illustrating just how provocative Magritte was when he painted a pipe with the caption “Ceci n’est pas une pipe” [This is not—or ain’t—a pipe]). But the fact is that the words of a verbal language too—whether or not they can say “ain’t,” since existence can be predicated only in a proposition and not in an isolated term—cannot say “do not take into consideration—or forget—what I am naming.”