What is the real autonomy of these levels, when there is more than one of them? Can one read The Divine Comedy without grasping its anagogic message? I would say that that is what so much Romantic criticism did. Can one read the procession at the end of Purgatorio without apprehending its allegorical dimension? A good surrealist reading of it could do so. And as for Paradiso, Beatrice smiles and radiates so much that she would enchant any reader who ignores the upper levels of meaning, and some critics, inspired by criteria of pure lyricity, told us that we had to ignore these disturbing overlays of meaning in the poem as though they were totally unconnected and had to be removed.*
One could then say that those levels of reading that depend on overlays of sense can either be activated or not, depending on the historical epoch in question, and sometimes they become totally unfathomable, as can happen not only with texts from very ancient civilizations but, for instance, with many paintings of not so long ago, where—apart from those who are iconographers or iconologists—visitors to museums (and even critics who rely solely on the visual) enjoy Giorgione or Poussin without knowing what obscure mythologies their images refer to (though we are convinced that Panofksy enjoyed them even more, since he was able to read them at both levels, at the level of form and at the level of iconographical reference).
The answer to the second question is quite different. I have repeatedly built theories around the fact that a text (and particularly a text with an aesthetic aim, and in the present case a narrative text) tends to construct two Model Readers. It addresses in the first place a Model Reader of the first level, whom we will call the semantic reader, the reader who wants to know (and rightly so) how the story will end (whether Ahab will capture the whale, whether Leopold Bloom will meet Stephen Dedalus, after having accidentally crossed his path on a number of occasions in the course of 16 June 1904, whether Pinocchio will become a real flesh-and-blood boy, or whether Proust’s Narrator will manage to settle his accounts with Lost Time). But the text also addresses a Model Reader of the second level, whom we will call the semiotic or aesthetic reader, who asks himself what kind of reader that particular story was asking him to become, and wants to know how the Model Author who is instructing him step by step will proceed. To put it bluntly, the first-level model reader wants to know what happens, while the second-level model reader wants to know how what happens has been narrated. To find out how the story will end one usually just has to read the text once. To become a second-level model reader one has to read it several times, and some stories have to be read countless times.
There is no such thing as an exclusively second-level model reader; on the contrary, in order to become one, you have to have been a good first-level reader. Whoever has read The Betrothed and has not felt even the slightest shudder when Lucia sees L’Innominato appear in front of her cannot appreciate the way Manzoni’s novel has been constructed. But it is certainly the case that you can be a first-level reader without ever reaching the second level, as happens with those who are equally enthused by The Betrothed and. Gargantua without realizing that the latter is much richer in lexical terms than the former; or as happens with those who, not unreasonably, get bored reading Hypnerotomachia Poliphili because amid all those made-up words it is impossible to understand how things will turn out.
On close examination it is in the play between these two levels of reading that we can observe the two ways of understanding catharsis in Aristotle’s Poetics, and in aesthetics in general: for we know that there is either a homeopathic or an allopathic interpretation of catharsis. In the first case catharsis stems from the fact that the spectator of a tragedy is genuinely seized by pity and terror, even to the point of paroxysm, so much so that in suffering these two passions he is purged of them, and emerges liberated by the tragic experience; in the second case the tragic text places us at a distance from the passion that is represented in it, through an almost Brechtian kind of estrangement, and we are liberated from passion not by experiencing it but by appreciating the way it is represented. You can easily see that for a homeopathic catharsis a first-level reader is sufficient (this is, after all, the kind of reader that cries when the cavalry arrives in a Western), whereas for an allopathic catharsis one needs a second-level reader—and this is what, perhaps erroneously, makes people attribute a greater degree of philosophical dignity, a more purified and purifying vision of art, to allopathic catharsis, whereas the homeopathic theory becomes linked to the celebration of Corybantism and the Eleusinian mysteries with their perfumes and drugs, or to the celebration of Saturday-night fever.
We should beware of understanding this distinction of levels as though on one side there were an easily satisfied reader, only interested in the story, and on the other a reader with an extremely refined palate, concerned above all with language. If that were so, we would have to read The Count of Monte Cristo on the first level, becoming totally enthralled by it, and maybe even shedding hot tears at every turn, and then on the second level we would have to realize, as is only right, that from a stylistic point of view it is very badly written, and to conclude therefore that it is a terrible novel. Instead, the miracle of works like The Count of Monte Cristo is that, while being very badly written, they are still masterpieces of fiction. Consequently the secondlevel reader is not only he who recognizes that the novel is badly written but also the one who is aware that, despite this, its narrative structure is perfect, the archetypes are all in the right place, the coups-de-scène judged. to perfection, its breadth (though at times stretched to breaking point) almost Homeric in scope—so much so that to criticize The Count of Monte Cristo because of its language would be like criticizing Verdi’s operas because his librettists, Francesco Maria Piave and Salvatore Cammarano, were not poets like Leo-pardi. The second-level reader is then also the person who realizes how the work manages to function brilliantly at the first level.
However, it is certainly at this second level of critical reading that one is able to decide whether the text has two or more levels of meaning, whether it is worthwhile looking for an allegorical sense, or whether the tale is also saying something about the reader—and whether these different senses blend together in a solid, harmonious form or whether they can float about independently of one another. It is the second-level reader who will decide that it is difficult to decouple the literal from the moral sense in the fable of the wolf and the lamb (as though it were pointless without the moral sense to tell the story of that diatribe between animals). On the other hand, one can read with enjoyment and reverence the psalm “In exitu Israel de Aegypto / domus Iacob depopulo barbaro /facta est Iudaea sanctificatio eius, / Israel potestas eius” (When Israel went forth from Egypt, / the house of Jacob from a people of strange language, / Judah became his sanctuary, / Israel his dominion), even without knowing that in anagogic terms the verses mean, among other things, that the sanctified soul will emerge from the enslavement of earthly corruption toward the freedom of eternal glory—and then the second-level reader will go and find out if the psalmist’s text really did mean this as well.
There are certainly many analogies between the aesthetically and critically aware second-level reader and the reader who, faced with examples of intertextual irony, catches the references to the universe of literature. But the two positions cannot be identical. Let us take some examples. In the fable about the wolf and the lamb there are two senses (one literal, one moral), and certainly two readers: the first-level reader who understands not just the story (the literal sense) but also the moral, and the reader who recognizes the stylistic and narrative merits of Phaedrus as teller of fables. But there is no intertextual irony because Phaedrus is not quoting anyone—or if he does cite a previous fabulist, he simply copies him.
Homer’s Ulysses kills the suitors: just one meaning, but two readers—one who enjoys Ulysses’ revenge and one who enjoys Homers art —but no intertextual irony. In Joyce’s Ulysses there are two meanings in the biblical-Dantesque mode (Bloom’s story as an allegory of Ulysses’ story), but it is very difficult not to notice that the story retraces the steps of Ulysses’ wanderings, and if someone did not notice this, the tide would offer him a clue. The two levels of reading still remain open, since one could read Ulysses just to know how the story ends—even though such a limited and limiting form of reading is highly improbable (in fact, it is exaggeratedly wasteful), and it would be advisable
