If, for example, I placed the Divine Comedy in the position of “q,” because the totality of assertions constituted by Dante’s text is false from the point of view of truth-function (it is not true that a Florentine ascended to Paradise while still alive, or that Charon exists), then by the same laws of material implication the inference would be false, despite having a true premise. Whereas everything would work if in place of “six times six makes thirty-six” we assigned to “p” the entire text of Mein Kampf, in that, according to the famous paradox of material implication, false plus false makes true. Semiotic balancing acts aimed at finding a relationship between symbolic logic and the obscure symbols of Romantic poetics are therefore pointless. They each have different ways of functioning, different syntax, and different natures with regard to truth.
Similarly, the use Cassirer makes of the term in his theory of symbolic forms has nothing to do with the sense that we attribute to it: instead, his is a culturological version of Kantian transcendentalism, and even Euclidean geometry is a symbolic form, where we breathe the sense of the infinite and the undecideable only in the continuation of the parallel lines that his fifth postulate promises us in a way that is itself undecideable in terms of the truth of the statement. We could stick to a sensible definition, which also applies to a whole series of daily experiences: the symbolic is identified by the existence, in every language, of levels of secondary meaning. This is the route taken by Todorov in his book on symbols. But to identify the symbolic with every instance of secondary meaning would lead us to confuse phenomena that are very different from each other.
There are two levels of meaning in any discourse that has two senses: the classic riddle is an example. But the two levels are structured, often on the basis of treacherous homonyms, according to two isotopies that can be traced without difficulty. Just like a message in code, the double sense needs to be deciphered, and once it has been decoded we have two senses that are indisputable, without any room for demurral.
Metaphor does not belong to the order of the symbolic. It can be open to multiple interpretations and can, as it were, be continued along the line of the second or third isotopy that it generates. But there are rules governing interpretation: that our planet is, as Dante says, “the threshing floor that makes us all so fierce” (Par. 22.151) might suggest thousands of poetic inferences, but it will not convince anyone, so long as there are cultural conventions we all agree on, that it is a place where peace and benevolence flourish. Moreover, I remain one of those who believe that the first signal of metaphorical usage consists in the fact that, taken literally, a metaphorical expression would appear false or weird, or nonsensical (the earth is not a threshing floor). This is not the case in the symbolic mode, which, as we shall see, conceals its own potential for meaning behind the deceptive appearance of something inexplicably obvious.
All the more reason, then, that allegory does not belong to the order of the symbolic either, since it is a continuous double sense based not on homonyms but on an almost heraldic codification of certain images.
The modern Western tradition is by now used to distinguishing allegory from symbolism, but the distinction is a rather late one: its articulation begins in Romanticism, and is particularly striking in Goethe’s famous aphorisms (Maximen und Reflexionen):
Allegory transforms the phenomenon into a concept and the concept into an image, but in such a way that the concept in the image is always to be considered circumscribed and complete in the image, and has to be given and to express itself through it (1.112).
Symbolism transforms the phenomenon into an idea and the idea into an image, in such a way that the idea in the image remains always infinitely effective and inaccessible and, even if articulated in every language, remains nevertheless inexpressible (1.113).
It makes a considerable difference whether the poet seeks the particular as a function of the universal or whether he sees the universal in the particular. In the first case we have allegory, where the particular is valid only as an example, as an emblem of the universal, whereas in the second case the true nature of poetry is revealed: the particular case is expressed without thinking about the universal or alluding to it. Now whoever catches this living particular seizes at the same time the universal without realizing it, or only realizing it later on (279).
True symbolism is that in which the particular element represents the more general, not as a dream or a shadow but as a living, instantaneous revelation of the inscrutable (314).
The classical and medieval world, on the other hand, understood “symbol” and “allegory” to be synonyms. Examples of this abound, from Philo to grammarians such as Demetrius, from Clement of Alexandria to Hippolytus of Rome, from Porphyry to Pseudo-Dionysius the Areopagite, and from Plotinus to Iamblichus, where the term “symbol” is also used for those didactic and conceptualizing representations that elsewhere would be called allegories.
It is true that scholars have described the world of the High Middle Ages as the “symbolic universe,” a universe where, ac-cording to the words of Duns Scotus Eriugena (De Divisione Naturae, V.3), “nihil enim visibilium rerum corporaliumque est, utar-bitror, quod non incorporate quid et intelligibile significet” (there is, I believe, no visible or corporeal thing that does not signify something non-corporeal and intelligible). The world, then, is apparently, as Hugh of St. Victor would say later, “quasi quidam liber scriptus digito Dei” (like some book written by the finger of God). So must it not have been the case that that world where “nostrum statum pingit rosa” (a rose depicts our condition) (Pseudo-Alain of Lille, Rhythmus Alter) was a world populated by symbols?
According to Huizinga (chapter 15 of his Waning of the Middle Ages), the medieval symbolic universe was very close to the universe of Baudelaire’s Corréspondences:
The medieval spirit was never so convinced of any great truth as it was of the words of St. Paul: “Videmus nunc per speculum in aenigmate, tunc autem facie ad faciem (now we see through a glass darkly, then we will see face to face).” The Middle Ages never forgot that any object would be absurd if its meaning were limited to its immediate function and its place in the world of phenomena, and that all things extend far into the world beyond. This idea is familiar also to us, as an unformulated sensation, when for instance the noise of the rain on the leaves in the trees or the light of the lamp on the table, in a moment of tranquillity, gives us a more profound perception than daily observation, which merely serves practical activity. It can sometimes appear in the form of a morbid oppression, which makes us see things already laden with personal menace or with a mystery that one ought to recognize but cannot. More often, however, it will fill us with the tranquil and reassuring certainty that our existence too participates in that secret sense of the world.
But this is the interpretation of someone who has already seen Verlaine and Rimbaud wandering, on the borders of his own country, as exiles searching for the absolute, listening to the sound of the same rain on the leaves and letting their hearts fill with languor, or hallucinating about the chromaticism of vowels. Was this really the symbolism of the High Middle Ages, not to mention the late medieval period?
In order to accept the Neoplatonic inheritance it was essential to conceive, as Dionysius the Areopagite does, an idea of the One as unfathomable and contradictory, where the divinity is called the “most luminous obscurity of the silence which arcanely teaches … the most luminous darkness” (Theologia mystica, passim). It is true that for Dionysius the concepts of the One, the Good, and the Beautiful are applied to God, as though they were on a par with Light, Lightning, and Jealousy; but these concepts will be used to describe him solely in a “hypersubstantial” way: he may be these things, but in a way that is commensurately but also incomprehensibly more intense.
What is more, Dionysius reminds us (and this is emphasized by his commentators), precisely in order to make clear that the names we attribute to God are inadequate, that it is important for them to be, as far as possible, different, incredibly unsuitable, almost provocatively offensive, extraordinarily enigmatic, as though the common quality that we are searching for between the symbolizing element and what is symbolized were indeed recognizable but only at the cost of inferential gymnastics and disproportionate proportions: and in order that the faithful, when naming God as Light, should not get the wrong idea that there exist celestial substances that are luminous and surrounded by haloes, it will be much better to name God using the names of monstrous beings, such as the bear or the panther, or through obscure dissimilarities (De Coelesti Hierarchia, 2).
Now this way of speaking, which Dionysius himself calls “symbolic” (e.g., De Coelesti Hierarchia, 2 and 15), has nothing to do with that illumination, that ecstasy, that rapid, lightning vision that all modern theories of symbolism see as peculiar to a symbol. The medieval symbol is a way of approaching the
