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

by Umberto Eco

Some historians (Dronke 1986: 14–16 and Bertini 2003: 35) have identified in Geoffrey a precise notion of the cognitive function of metaphor, in the sense in which it would be taken up by Dante in his Letter to Cangrande della Scala (Epistole XIII, 29), where he says of the things seen during his celestial journey that, since they cannot be expressed “sermone proprio” (“in everyday language”), they must be spoken of “per assumptionem metaphorismorum” (“by the employment of metaphors”). Nevertheless—setting aside Dante, to whom we will return—Geoffrey repeats that in constructing metaphors it would be wrong not to draw the properties from among those that are “expressissime et apparentissime similia” (“most expressly and apparently similar,” II, 3, 17–18). Obviously, milk and snow are white, and honey is sweet, but Geoffrey does not seem to advise identifying properties that are nonself-evident in order to create unexpected likenesses. Indeed (in II, 3) he denounces as “turgidus et inflatus” (“turgid and inflated’) that style “qui nimis duris et ampullosis utitur translationibus,” (“which has recourse to crude and bombastic metaphors”), such as saying “ego transivi per montes belli” (“I have crossed over the mountains of war”) instead of settling for “per difficultates belli” (“the hardships of war”).

  And in fact the same author, in his Poetria nova (765 et seq.), suggests the use of prefabricated metaphors, so to speak. Instead of “aurum fulvum, lac nitidum, rosa praerubicunda, mel dulcifluum, flammae rutilae, corpus nivis album” (“tawny gold, limpid milk, a rose redder than red, smooth-flowing honey, ruddy flames, a body white as snow”) it is better to say “dentes nivei, labra flammea, gustus mellitus, vultus roseus, frons lactea, crinis aureus” (“snow-white teeth, lips of flame, a taste like honey, rosy cheeks, a milk-white brow, golden hair). It is acceptable to say that spring paints the earth with flowers, that fair weather soothes, that the winds are sleeping, that deep valleys lie, because, by transferring human actions to nonhuman things, man sees himself in nature, as in a mirror. But this is still the canonical procedure of the anthropomorphization of the inanimate. Furthermore, though Geoffrey may venture a rule which we have dubbed logical-semantic, in point of fact he does not suggest any criterion for the proper identification of the relevant properties.

  3.2. References and Examples in Philosophical Thought

  We might expect greater commitment on the part of the philosophers, who deal with the correct meaning of terms and the difference between univocal and equivocal signs. In her essay “Prata rident”, Rosier-Catach (1997) examines a canonical topos in medieval doctrinal thought: the metaphor of the smiling meadow (already present in Ad Herennium 4). It is striking how this same example occurs over and over again in very different authors, from Abelard to Theodoric of Chartres and William of Conches, down to Thomas Aquinas, eventually spilling over into the discussions of analogy or translatio in divinis, in other words, the use of metaphors to speak of God.

  Abelard’s point of departure is an annotation in Boethius’s commentary on the Categories, according to which, if one calls the “gubernator” (helmsman) of a ship its “auriga”(charioteer), and if one does so “ornatus causa,” there is no ambiguity. Abelard says he agrees, because in that case the text assumes the transferred meaning only for a limited time, as occurs when one says “ridere” instead of “florere” of a meadow (Glossae super Predicamenta, in Geyer 1927: 121). The transferred meaning does not occur per institutionem but only in a specific context, “per abusionem translationis, ex accidentale usurpatione” (“for an abuse of metaphor, as a result of a casual inappropriate use.” Super Peri herm., in Geyer 1927: 364). What we have here is not an instance of translatio aequivoca based on penuria nominum. The case is instead somewhat similar to that of oppositio in adiecto (“opposition in the attribute”), as in homo mortuus, where homo signifies (here and here alone) “corpse.”

  William of Conches (Glosae in Priscianum) will speak of locutio figurativa more or less as Abelard does (Rosier-Catach 1997: 161–164). Robert Kilwardby says that in the case of the trope the expression is not understood as “intellectus primus” but as “intellectus secundus,” not “simpliciter” but “secundum quid.” The Flores Rhetorici (by the twelfth-century Master of Tours) speaks of words united in “decente matrimonio,” and there appears to be a timid allusion to the inferences that can be drawn from a metaphor, so that from “prata rident” one may proceed to “prata luxuriant floribus or prata floribus lasciviunt.” Here Rosier-Catach (1997) speaks of evidence of awareness of metaphorical productivity, but we personally find the allusion if anything quite tenuous. In the same vein the Dialectica Monacensis (II, 2, in De Rijk 1962–1967, II: 561) finds it extravagant and inappropriate to hazard the following syllogism: “Quicquid ridet habet os—pratum ridet—ergo habet os” (“Whatever smiles has a mouth—the meadow smiles—therefore it has a mouth”).

  From a logical point of view, the position could not be more reasonable. And yet, if we want to know how to go about making metaphor an instrument of new knowledge and invention, we have only to see what the Jesuit Emanuele Tesauro, in the baroque period, is able to make of a “fair flower of rhetoric” that by his day was beginning “to stink.” We have only to read the lengthy analysis in the Cannocchiale aristotelico (ed. Zavatta, 1670: 116 et seq.) dedicated to the smile of the meadows, where he demonstrates how many new ideas and revelatory images can spring from a productive development of the initial trope. For upward of five pages of variations by inference on the original nucleus, in a virtuoso pyrotechnic display of baroque wit, Tesauro shows how the metaphor can give rise to infinite ways of seeing the fecundity of the meadows: “Iucundissimus pratorum RISUS, RIDIBUNDA vidimus prata, RIDENTER prata florent, Pratorum RISIO oculos beat, RIDENTISSIME prata gliscunt …” (“The most delightful smile of the meadows, we have seen the meadows smile, the meadows smile in flowering, the smiling of the meadows delights our eyes, the meadows rejoice most smilingly”). Whereupon he proceeds to invert the metaphor, “Hac in solitudine MOESTISSIMA videres prata. Sub Canopo squalida ubique prata LUGENT” (“In such solitude you would see the meadows most mournful. Under the bright star of Canopus the mournful meadows are weeping”), or, by the subtraction of human properties, we get, “Prata RIDENT sine ore. RISUS est sine cachinno” (“The meadows smile without a mouth. The smile is without laughter”), and, by the extension of the metaphor to component parts of the meadow or to the whole earth, we get “Virides rident RIPAE. Laeta exultant GRAMINA, Fragrantissimi rident FLORES. Alma ridet TELLUS. Rident SEGETES” (“The verdant banks smile. The grasses exult joyfully. The most sweet-smelling flowers smile. The life-giving earth smiles. The crops smile.”) And Tesauro enthusiastically continues:

  Che se hora tu ligherai questa proprietà del rider de’ prati, con le cose Antecedenti, Concomitanti & Conseguenti: tante Propositioni, & Entimemi arguti, ne farai germogliare; che tanti fiori apunto non partoriscono i prati al primo tempo. Chiamo antecedenti le Cagioni di questo metaforico Riso; cioè: il ritorno del Sole dal tropico hiberno al Segno dell’Ariete. Lo spirar di Zefiro fecondator della terra. I tiepidi venti Australi. Le piogge di Primavera. La fuga delle neui. Le sementi dell’Autunno. Onde scherzando dirai: SOLI arridentia prata reditum GRATVLANTVR, Vis scire cur prata rideant? … Suavissimis AUSTRI delibuta suauijs, subrident prata, Dubitas cur prata rideant? IMBRIBVS ebria sunt. (Tesauro 1968, pp. 117–118)11

  And so on and so forth. And if we may grant a human smile to the meadows, why not grant them also the features that accompany the smile? Hence, “Pulcherrima pratorum FACIES. Et se la faccia ha le sue membra: ancor dirai; Tondentur falce virides pratorum COMAE, CRINITA frondibus prata virent. Micantes pratorum OCULI, flores” (“ ‘The FACE most fair of the meadows.’ And if the face has all its attributes, then you will say: ‘The green LOCKS of the meadows are mown by the sickle. The meadows are green with their COIFFURE of leaves. The flowers are the flashing EYES of the meadows’ ” (ibid., p. 118).

  This appeal to Tesauro, however, merely serves to underscore, by way
of contrast, the timidity of all medieval theories of metaphor.

  3.3. Metaphor, Allegory, and Universal Symbolism

  Why does the Middle Ages confine metaphor to a merely ornamental function and fail to recognize, at least on the theoretical level, its cognitive possibilities? The answer is twofold: (i) for the Middle Ages, our only teacher, who speaks through “real” metaphors (in rebus), is God, and all man can do is to uncover the metaphorical language of creation, and (ii) if man would speak of God, then no metaphor is equal to the challenge, and no metaphor can account for his unfathomable nature any more than literal language can.

  If we wish to study this aspect of medieval culture and its implicit semiotics, we must establish precise distinctions between metaphor, symbol, and allegory—which is what we did in Eco (1985), and to which we will return later.12 For now, we may speak generically of figural language for all those cases where aliud dicitur, aliud demonstratur, in which there is some kind of translatio from one term or a string of terms (or better, from the contents they express) to another, which somehow constitutes its secondary meaning.13 What interests us here is how the Middle Ages fixes its attention on phenomena of secondary or figural meaning, which are not those of literary metaphor.

  Our starting point is Paul’s First Epistle to the Corinthians 13:12: “Nunc videmus per speculum et in aenigmitate, tunc autem facie ad faciem” (“For now we see through a glass, darkly; but then face to face”). The most elegant solution poetically speaking is that supplied by the Rhythmus alter, formerly attributed to Alan of Lille (PL 210: 578C–579C):

  Omnis mundi creatura,

  Quasi liber, et pictura

  Nobis est, et speculum.

  Nostrae vitae, nostrae mortis,

  Nostri status, nostrae sortis

  Fidele signaculum.

  Nostrum statum pingit rosa,

  Nostri status decens glosa

  Nostrae vitae lectio.

  Quae dum primo mane floret,

  Defloratus flos effloret

  Vespertino senio.

  Ergo spirans flos exspirat,

  In pallorem dum delirat,

  Oriendo moriens.

  Simul vetus et novella,

  Simul senex et puella

  Rosa marcet oriens.

  Sic aetatis ver humane

  Iuventutis primo mane

  Reflorescit paululum.14

  The world is to be interrogated as if every item with which it is furnished had been put there by God to instruct us in some way. As Hugh of Saint Victor will remark, the sensible world “quasi quidam liber est scriptus digiti Dei” (“is like a book written by the finger of God”) (De tribus diebus VII, 4), and, according to Richard of Saint Victor, “habent tamen corporea omnia ad invisibilia bona similitudinem aliquam” (“and yet all corporal things bear some resemblance to the goods we cannot see”) (Benjamin major II, 13).

  The fact that the world is a book written by the finger of God is seen not so much as a cosmological notion as an exegetical necessity. In other words, this universal symbolism starts out primarily as scriptural allegorism and goes on to become what has been defined as “universal symbolism.”

  Commentators spoke of allegorical interpretations well before the birth of the patristic scriptural tradition: the Greeks interrogated Homer allegorically; in Stoic circles there arose an allegorist tradition which saw the classical epic as a mythical cloaking of natural truths; there existed an allegorical exegesis of the Jewish Torah, and in the first century Philo of Alexandria attempted an allegorical reading of the Old Testament.

  In an attempt to counterbalance the Gnostic overemphasis on the New Testament, to the total detriment of the Old, Clement of Alexandria proposes viewing the two testaments as distinct and complementary, while Origen perfects this position by insisting on the necessity of a parallel reading. The Old Testament is the figure of the New, it is the letter of which the other is the spirit, or, in semiotic terms, it is the expression of which the New is the content (or one of the possible contents). In its turn, the New Testament has a figural meaning, inasmuch as it is a promise of future things. With Origen the “theological discourse” is born, which is no longer—or no longer simply—a discourse on God, but on His Scripture.15

  Origen already speaks of a literal sense, a moral (psychic) sense, and a mystical (pneumatic) sense. Hence the triad—literal, tropological, and allegorical—that will later become the foursome expressed in the famous distich of Augustine of Dacia (thirteenth century): “littera gesta docet—quid credas allegoria—moralis quid agas—quo tendas anagogia” (”the letter tells us what went down—the allegory what faith is sound—the moral how to act well—the anagogy where our course is bound”).

  From the beginning, Origen’s hermeneutics, and that of the Fathers of the Church in general, tends to favor a kind of reading that has been defined as “typological”: the characters and events of the Old Testament are seen, because of their actions or their characteristics, as types, anticipations, foreshadowings of the characters of the New. Some authors (such as Auerbach 1944, for example) attempt to discern something different from allegory, when Dante, instead of allegorizing openly—as he does, for instance, in the first canto of the Inferno or in the procession in the Earthly Paradise—brings onstage characters like Saint Bernard who, without ceasing to be living and individual figures (in addition to being authentic historical personages), become “types” of superior truths on account of certain of their concrete characteristics. Some would go so far as to speak, apropos of these examples, of “symbols.” But in this case too, what we are probably dealing with is allegory: the vicissitudes, interpretable literally, of one character, become a figure for another (at best what we have is an allegory complicated by Vossian antonomasia, inasmuch as the characters embody certain of their outstanding characteristics).

  However we describe this typology, it requires that what is figured (whether a type, a symbol, or an allegory) be an allegory not in verbis but in factis. It is not the words of Moses or the Psalmist, qua words, that are to be read as endowed with an secondary meaning, even though they appear to be metaphorical expressions: it is the very events of the Old Testament that have been prearranged by God, as if history were a book written with his hand, to act as a figure of the new dispensation.

  A useful distinction between facts and words may be found in Bede’s De Schematibus et tropis, but Augustine had already addressed this problem, and he was in a position to do so because he had been the first, on the basis of a profoundly assimilated Stoic culture, to create a theory of the sign. Augustine distinguishes between signs that are words, and things that may function as signs, since a sign is anything that brings to mind something else, over and above the impression the thing makes on our senses (De Doctrina Christiana II, 1, 1).16 Not all things are signs, but all signs are certainly things, and, alongside the signs produced by man intentionally to signify, there are also things, events, and characters that can be assumed as signs or (as in the case of sacred history) can be supernaturally arranged as signs so as to be read as signs.

  In this way Augustine teaches us to distinguish obscure and ambiguous signs from clear ones, and to resolve the question of whether a sign is to be interpreted in a literal or in a figurative sense. Tropes like metaphor or metonymy can be easily recognized because if they were taken literally the text would appear meaningless or childishly mendacious, but what about those expressions (usually involving a whole sentence or a narration, and not a simple term or image) that have an acceptable literal meaning and to which the interpreter is instead led to assign a figurative meaning (as is the case, for example, with allegories)? A metaphor tells us that Achilles is a lion, and from the literal point of view this is a lie, but an allegory tells us that a leopard, a she-wolf, and a lion are encountered in a dark wood, and the statement could perfectly well be taken at face value.

  To get back to the author of the Rhythmus alter, more than a metaphor, what we have here is an allegory, indeed,
it represents a set of instructions for decoding allegories. He does not say life is a rose (an expression that would be absurd if taken literally). Instead, he lists all the qualities that pertain to the rose, qualities which (while still remaining literally comprehensible) become or may become (if the proper interpretive tools are provided) an allegory of human life. In fact, before listing the properties of the rose, he informs us that it is a depiction of our state (“nostrum statum pingit rosa”), and goes on to furnish the necessary elements to make the parallel clear.

  How do we understand that something that has an acceptable literal meaning is to be understood as an allegory? Augustine, discussing the hermeneutical rules proposed by Tyconius (De doctrina christiana III, 30, 42—37, 56), tells us that we must suspect a figurative sense whenever Scripture, even if what it says makes literal sense, appears to go against the truth of faith or decent customs. Mary Magdalene washes the feet of Christ with perfumed ointments and dries them with her own hair. Is it thinkable that the Redeemer would submit to such a lascivious pagan ritual? Obviously not. So the narrative must be a representation of something else.

  But we must also suspect a secondary meaning whenever Scripture gets lost in superfluitates or brings into play expressions poor in literal content. These two considerations are amazingly subtle and modern, even if Augustine found them already suggested by other authors.17

  We have superfluitas when the text spends an inordinate amount of time describing something that might have a literal sense, but without the textually economical reasons for this descriptive insistence being clear. We have semantically poor expressions when proper names, numbers, or technical terms show up, or insistent descriptions of flowers, natural prodigies, stones, vestments, or ceremonies—objects or events that are irrelevant from the spiritual point of view. In such cases, we must presume—since it is inconceivable that the sacred text might be indulging a taste for ornament—that aliud dicitur et aliud significatur, one thing is said and another is intended.