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

by Umberto Eco

  In this labyrinth, which no longer presents itself as a logical division but as a rhetorical accumulation of notions and topics arranged under loci, the Latin verb invenire (= to find or discover) no longer means to find something one already knew existed, sitting in its proper place, ready to be used for the purposes of argument, but truly to discover some new thing, or the relationship between two or more things, that one was previously unaware of. Such a situation represents (as Rossi 1957, IV and V reminds us) the complete and radical refusal of any preestablished hierarchy among beings. Pursuing an idea that will be taken up again later by Leibniz, in the Advancement of Learning Bacon points out that, if a secretary of state is obliged to accumulate a series of records in his official place of business, he will classify them according to the nature of the document (treaties, instructions, etc.), whereas in his private study he will keep all the papers that require his immediate attention together, even though they may be of a heterogeneous nature. The Great Chain of Being is a thing of the past, and from now on every subdivision will invariably be made in context and directed toward a specific end.

  1.3.4. The Cannocchiale aristotelico of Emanuele Tesauro

  We have seen how with Bacon the idea of inventio (the noun derived from the verb invenire) undergoes a sea change and, instead of referring to the search for something already familiar, is transformed into the discovery of something not yet known. But in this case hunting through the repertory of knowledge is like rummaging through an immense warehouse whose extent is not yet known, and rummaging not simply to put what one finds, whatever it may be, to use, but to construct, so to speak, a bricolage, discovering new syntheses, connections and dovetailings among things that at first sight did not appear to have any reciprocal relationship.

  An encyclopedic model is paradoxically offered by Emanuele Tesauro’s Cannocchiale aristotelico (“Aristotelian Telescope,” 1665). I say “paradoxically” because, in the very century in which the model of Galileo’s telescope comes into its own as the paradigmatic instrument for the development of the natural sciences, Tesauro proposes a telescope named after Aristotle as an instrument for renewal of what today we would call the human sciences, and the instrument he proposes is metaphor. In the Cannocchiale, however, we recognize the fundamental nucleus of Aristotelian rhetoric (of which more in section 1.8.1), and the model of metaphor is proposed as a means of discovering unfamiliar relations among the elements of knowledge, though Tesauro’s interest, unlike Bacon’s, is rhetorical rather than scientific.

  To construct a repertory of known things, scrolling through which the metaphorical imagination may be led to discover unknown relationships, Tesauro develops the idea of a Categorical Index. He presents his index (with Baroque complaisance for the “marvelous” invention) as a “truly secret secret,” an inexhaustible mine of infinite metaphors and ingenious conceits, given that genius is nothing more or less than the ability to “penetrate the objects deeply hidden beneath the various categories and compare them among themselves”—the ability, in other words, to unearth analogies and similarities that would have passed unnoticed had everything remained classified under its own particular category.

  It is sufficient, then, to inscribe in a book Aristotle’s ten categories, the Substance and the nine Accidents, and then list under each category its Members and under each Member the Things “subject to it.”

  All we can do for our present purposes is to give a few meager examples of the extensive catalogue Tesauro provides (susceptible in any case of constant expansion). Thus, under the category of Substance, are to be recorded as Members the Divine Persons, Ideas, the Fabulous Gods, Angels, Demons, and Sprites; then, under the Member Heaven, the Wandering Stars, the Zodiac, Vapors, Air, Meteors, Comets, Torches, Thunderbolts, and Winds; and then, under Earth, Fields, Solitudes, Mountains, Hills, and Promontories; under Bodies, Stones, Gems, Metals, Herbs; under Mathematics, Orbs and Globes, Compasses, and Squares; and so on and so forth.

  Likewise, for the category of Quantity, under the Quantity of Size are listed the Small, the Large, the Long, and the Short; under the Quantity of Weight, the Heavy and the Light. For the category of Quality, under Sight we find the Visible and the Invisible, the Apparent, the Handsome and the Misshapen, the Bright and the Dark, the White and the Black; under Scent, Sweet Odor and Stench—and so on through the categories of Relation, Action and Passion, Site, Time, Place, and State.

  When we take a closer look at the Things subordinate to these Members, we find that, under the category of Quantity and the Member Size, among small things we find the angels (which fit within a point), the incorporeal forms, the pole as the unmoving point of the sphere, the zenith and the nadir; among Elementary Things the spark of fire, the droplet of water, the grain of sand, the scruple of stone, the gem, and the atom; among Human Things, the embryo, the abortus, the pigmy, and the dwarf; among Animals, the ant and the flea; among Plants, the mustard seed and the crumb of bread; among the Sciences, the mathematical point; in Architecture the tip of a pyramid; under Lanaria, the metal tip of a lace, and so on with a list that goes on for two pages.

  We have no need to ask ourselves just how congruous this list is. Incongruity seems to be typical of all of the efforts made in the Baroque period to give an account of the global contents of a field of knowledge, just as it is equally characteristic of many seventeenth-century projects for artificial languages. Gaspar Schott, in his Technica curiosa (1664) and his Joco-seriorum naturae et artis sive magiae naturalis centuriae tres (ca. 1666) gave notice of a work published in 1653, whose author’s name he claims to have forgotten. In fact the anonymous author seems to have been a certain Pedro Bermudo (1610–1648), a Spanish Jesuit who presented in Rome an Artificium or Arithmeticus nomenclator, mundi omnes nationes ad linguarum et sermonis unitatem invitans. Authore linguae (quod mirere) Hispano quodam, vere, ut dicitur, muto.25 It is doubtful whether Schott’s is a faithful description, but the issue is irrelevant, since, even if Schott had reworked the project after his own fashion, what interests us is the incongruity of the list. The Artificium provided for forty-four fundamental classes, which are worth listing here, giving only a few examples in parentheses:

  1. Elements (fire, wind, smoke, ash, hell, purgatory, and the center of the earth). 2. Celestial entities (stars, thunderbolts, the rainbow). 3. Intellectual entities (God, Jesus, speech, opinion, suspicion, soul, stratagem, or ghost). 4. Secular states (emperor, barons, plebs). 5. Ecclesiastical states. 6. Artificers (painter and sailor). 7. Instruments. 8. Affections (love, justice, lust). 9. Religion. 10. Sacramental confession. 11. Tribunal. 12. Army. 13. Medicine (doctor, hunger, clyster). 14. Brute beasts. 15. Birds. 16. Reptiles and fish. 17. Parts of animals. 18. Furnishings. 19. Foodstuffs. 20. Beverages and liquids (wine, beer, water, butter, wax, resin). 21. Clothing. 22. Silk fabrics. 23. Wools. 24. Canvas and other woven cloths. 25. Navigation and spices (ship, cinnamon, anchor, chocolate). 26. Metals and coins. 27. Various artifacts. 28. Stones. 29. Jewels. 30. Trees and fruit. 31. Public places. 32. Weights and measures. 33. Numerals. 39. Time. 40. Adjectives. 41. Adverbs. 42. Prepositions. 43. Persons (pronouns, titles such as Your Eminence). 44. Travel (hay, road, highway robber).26

  Around 1660 Athanasius Kircher had composed a Novum hoc inventum quo omnia mundi idiomata ad unum reducuntur (“New invention by which all the languages of the world can be reduced to one”) still surviving only in manuscript form,27 which proposed a fairly elementary grammar and a dictionary of 1620 “words,” in which he endeavored to establish a list of fifty-four fundamental categories capable of being represented by means of iconograms. His iconograms recall those in use today in airports and stations—sometimes they represent an object, such as a small wineglass, sometimes they are purely geometrical (a rectangle, a triangle, a circle), while some of them are superficially inspired by Egyptian hieroglyphs. Without going into detail (see Marrone 1986 and Eco 1993: 9), we may simply note that the fifty-four categories of the Novum Inventum also constitute a notably
incongruous list, including as they do divine, angelic, and celestial entities, elements, human beings, animals, vegetables, minerals, the dignities and other abstract concepts of Llull’s Ars Magna, beverages, clothing, weights, numbers, hours, cities, foodstuffs, family, actions like seeing or giving, adjectives, adverbs, the months of the year. But let us get back to Tesauro.

  Tesauro follows the bent of his time. But what seems to us a lack of the systematic spirit is on the contrary evidence of the effort made by the encyclopedist to avoid arid classification according to genera and species. It is the as yet unordered accumulation (or barely ordered, in Tesauro’s case, under the rubrics of the ten categories and their members) that will later permit the invention (in the Baconian sense, not of recovery but of discovery) of unexpected and original relationships between the objects of knowledge. This impression of a “hodgepodge” is the price we have to pay, not to achieve completeness but to eschew the poverty of any classification in the form of a tree.

  We have only to see what Tesauro makes of his warehouse of notions. If we were searching for a good metaphor for a dwarf (though for Tesauro discovering metaphors means, as it did for Aristotle, coming up with new definitions for things or discovering everything that can be said about a given object), from this repertory we could already derive the definitions of Myrmidon (the name is related to “ant”) or the little mouse at whose birth the mountains were in labor. But to this index there is added another that, for every small thing, depending on which of the ten categories we consider, decides, under Quantity, what the small thing is commensurate with or what parts it has; under Quality, whether it is visible or what deformities it has; under Relation, to whom or with what it is related, whether it is material and what form it has; under Action and Passion, what it can and cannot do, and so on. And once we have asked ourselves what the small thing is commensurate with, the Index ought to refer us, for example, to “the Measure of the Geometric Finger.”28

  Proceeding in this way through each category, we could say of the dwarf that he is shorter than his own name, more an embryo than a man, a fragment of humanity, far smaller than a thumb, so insubstantial as to be without color, sure to be the loser in a fight with a fly, so tiny you can’t tell whether he is sitting, standing or lying down, and so on.

  The Index, precisely because of its labyrinthine nature, allows us to make connections between each object and every other object—so that it seems that all Tesauro’s metaforeta or metaphor maker can do (and all he delights in doing) is deriving new knowledge from the deconstruction of a Porphyrian tree.

  Although, out of devotion to Aristotle, Dante’s “master of all those who know” (Inferno IV, 131), and his works, Tesauro opted to call his index “categorical,” what he in fact provides is a procedure to pursue the infinite paths of a labyrinth, in which the subdivisions according to categories are nothing more than provisional and ultimately arbitrary constructions designed to contain somehow or other material that is in a constant state of ferment.

  1.3.5. Wilkins

  The point of greatest tension between tree and labyrinth is reached in seventeenth-century England, in the ambit of the Royal Society, where various projects for an a priori philosophical language (such as Lodwick’s A Common Writing, Becks’s The Universal Character, Dalgarno’s Ars signorum or the Essay towards a Real Character and a Philosophical Language by Wilkins) are formulated, in which “characters” comprehensible to people who speak different languages are called upon to represent a global structure of the world.

  What these systems discuss is the possibility of representing the meanings of each term through a punctiliously exhibited hierarchical arrangement of subdivisions from genera to species, while at the same time giving an account of the nonregimentable multiplicity of notions that common speakers have at their disposal. The problem these systems find themselves having to face is that, if one chooses a tree classification, according to the dictionary model, it is impossible to give an account either of the meaning of the terms or the nature of the things designated, and therefore the nodes of every tree-like classification must be filled in with encyclopedic specifications, with sums of properties, in other words, that can neither be defined or classified.

  Referring the reader to Eco (1993) for a more detailed analysis of these systems and the relevant bibliography, we will confine ourselves in this context to considering briefly Wilkins’s Essay towards a Real Character, the most complete and fully worked-out project of them all. Wilkins conducted a kind of colossal review of all knowledge and produced a table of 40 major Genera, proceeding to subdivide them into 251 peculiar Differences, from which he derived 2,030 Species (presented in pairs). The table of 40 Genera (Figure 1.12) starts out with very general concepts like Creator and World and, by means of a division into substances and accidents, animate and inanimate substances, vegetative and sensitive creatures, arrives at Stones, Metals, Trees, Birds, or accidents like Magnitude, Space, Sensible Qualities, Economical Relations.

  Figure 1.12

  More detailed still are the tables that allow us to arrive at individual species, in which Wilkins proposes to classify, for instance, even a beverage like beer, in order to represent the entire notional universe of a seventeenth-century Englishman. With regard to this system of ideas (which Wilkins, clearly erring on the side of ethnocentricity, presumes to be common to all mankind), the “real characters” that he proposes are signs (which assume both a written form, almost hieroglyphic in nature, and an oral form, transcribed in pronounceable alphabetic characters). Thus, if De signifies Element, and Deb the first difference (Fire), then Deba will denote the first species, which is Flame.

  Here, however, we are not interested in Wilkins’s writing proposals (essential though they may be to his project for a universal language), but in the criteria he uses to organize the notions. Once again, the mere classification does not permit us to recognize a Flame or to assert that it burns. Even when we get down to the single species we find divisions according to which, given the category Viviparous Clawed Beasts, subdivided into Rapacious and Non-Rapacious, under Rapacious we find Cat-kind and Dog-kind, the latter being divided into European and Exotic, the European further divided into Amphibious and Terrestrial, the Terrestrial into Bigger (Dog/Wolf) and Lesser (Fox/Badger), as we see in Figure 1.13.

  Figure 1.13

  As usual, not only is it impossible to distinguish a dog from a wolf, but, in addition, the information that the “characters” of Wilkins’s alphabet transmit is simply that the dog (Zita in the universal language) is “the first member of the specific pair of the fifth difference of the genus Beasts.”

  It is not until we consult the extremely crammed encyclopedic tables which Wilkins places after his classifications that we learn that viviparous animals have feet with toes, rapacious animals usually have six sharp incisors and two long fangs to hold their prey, the Dog-kind have round heads which distinguish them from the Cat-kind whose heads are more oblong in shape, while the largest of the canines are subdivided into “domesticated-tame” and “wild-hostile to sheep”: this is the only way for us to grasp the difference between a dog and a wolf.

  Wilkins’s philosophical language taxonomizes but it does not define. In order to define, the system must have recourse to a miscellany of information expressed in a natural language that takes the form of an encyclopedia.

  The defect that becomes evident in Wilkins’s failure is the same defect that undermines any notion of a dictionary that sets itself the aim of being rigorous. In order for a dictionary to be totally independent of any additional knowledge of the world, its terms must be primitives not further definable—otherwise the tree would forfeit its nature as a device capable of guaranteeing the exactitude of the definitions it generates. But, in Wilkins’s case, it is clear that the mass of encyclopedic information underlying the organization of the tables according to supposed primitives is in fundamental contradiction with the compositional character according to traits that app
eared to be being realized in his “characteristic” language. The primitives are not primitives. Not only are Wilkins’s species combinations of genera and differences (a weakness already typical of a Porphyrean Tree, given that the differences are accidents not subject to hierarchization), but furthermore they are names used as hooks on which to hang encyclopedic descriptions.

  Nevertheless, precisely because it is impure, Wilkins’s system is susceptible of another reading, no longer as a dictionary but as a hypertext, in our contemporary meaning of the term. If a hypertext links every node or element of its repertory, by means of a multiplicity of internal cross-references, to a multiplicity of other nodes, one could imagine a hypertext regarding animals that inserts dog into a general classification of mammals, in a tree of taxa that also includes cats, oxen, and wolves. But if in that tree one points to dog (precisely in the modern computer sense of clicking on it), one is directed to a repertory of information concerning the properties and habits of dogs. Selecting another type of connection, one can also access a list of the various roles played by dogs in different historical periods, or a list of images of dogs in art history. Perhaps this is where Wilkins was headed, when he thought of considering Defense both in terms of the duties of the citizen as well as in terms of military strategy.