On Literature

by Umberto Eco

  Both these modes can be practiced in two different ways, which, according to Croce, can be defined as the "artifex additus artifici" (an artist writing about an artist) approach, or the "philosophus additus artifici" (a philosopher writing about an artist) approach. In the first case, rather than explaining the work to us, the critic provides us with a diary of his emotions in reading it, and unconsciously seeks to outdo in virtuosity the object of his humble devotion. Sometimes he succeeds—we know very well passages on literature that are finer, in a literary sense, than the literature they discuss, just as Proust's pages on bad music are of the highest musicality.

  In the second approach the critic tries to show us, in the light of certain critical categories and criteria, why the work is beautiful. But in the case of a review he does not have sufficient space to tell us in depth how the work is made (and therefore to reveal to us the machinations of its style), while a history of literature has to maintain its analysis on a level of enforced generality. Unfortunately, it takes a hundred pages to lay bare the style of one page, and in a history of literature the proportions are inevitably the inverse.

  Let us now come to the third mode, textual criticism. In it the critic has to assume that the reader knows nothing about the work, even if it is as well known as The Divine Comedy. He has to make the reader discover it for the first time. If the text is not brief enough to be quoted in its entirety, and subdivided into sections of prose or verse, he has to presume that the reader has access to it in some other way, since the goal of this discourse is to lead him, step by step, in the discovery of how the text has been put together and why it functions as it does. This discourse can be put forward as a confirmation ("now I will show you why everyone considers this to be a brilliant text"), as a revaluation, or as the destruction of a myth. The ways in which one can show how a text is put together (and why it is right that it is put together in this way, and how it could not be composed in any other way, and why it has to be considered as sublime precisely because it is composed in this way) can be countless. No matter how these discourses are articulated, such criticism cannot be anything other than a semiotic analysis of the text.

  Consequently, if proper criticism is understanding and making others understand how a text is made, and if the review and the history of literature are unable to do this adequately, the only true form of criticism is a semiotic reading of the text.

  Like proper criticism (which must lead to an understanding of the text in all its aspects and potential) the semiotic reading of a text possesses a quality that is usually and indeed inevitably missing in a critical review or history of literature: it does not prescribe the various ways in which the text can be pleasurable, but, rather, it shows us why the text can produce pleasure.

  Because it has to make recommendations, a critical review cannot be exempt, except in cases of exceptional cowardice, from pronouncing a verdict on what the text says; historical criticism shows us at most that a work has enjoyed a varied and fluctuating critical fortune, and has aroused different responses. Textual criticism, by contrast, which is always semiotic even when it does not know it is, or even when it denies it is, fulfills that function which was admirably described by Hume in "Of the Standard of Taste," which cites a passage from Don Quixote:

  Two of my kinsmen were once called to give their opinion of a hogshead, which was supposed to be excellent, being old and of a good vintage. One of them tastes it; considers it, and after mature reflection pronounces the wine to be good, were it not for a final taste of leather, which he perceived in it. The other after using the same precautions, gives also his verdict in favor of the wine; but with the reserve of a taste of iron. You cannot imagine how much they were both ridiculed for their judgment. But who laughed in the end? On emptying the hogshead, there was found at the bottom, an old key with a leathern thong tied to it.*

  My point is that proper criticism always has the last laugh, because it allows everyone to have his own pleasure, but it also shows the reason for that pleasure.

  ***

  Of course, even textual criticism, carried out by a "philosophus additus artifici," can be aware of its own excesses, which thwart its very function. It will be useful to consider some errors of textual semiotics, which have occasionally caused the rejection syndromes I spoke of just now.

  There is often confusion between "the semiotic theory of literature" and "criticism that is semiotically oriented." I refer you to an old debate from the 1960s, which began with the famous catalogue issued by the publisher II Saggiatore dedicated to "Structuralism and Criticism," a debate in which there were roughly two speaking positions, one represented by Segre, the other by Rosiello. To put it briefly, for the first option, linguistic theory was to be used to shed light on the individual work; for the second, the analysis of the individual work was to be used to shed light on the nature of language. Therefore, when the first option was at work, a group of theoretical assumptions would be used to shed light on an author's personal style, whereas in the case of the second option, the personal style was felt to be a deviation from the linguistic norm, which reinforced knowledge of the norm as such.

  Now these two positions were and are equally legitimate. One can construct a theory of literature, and use individual works as documents, and one can read individual works in the light of a theory of literature, or, rather, in an attempt to make the very principles of a theory of literature emerge from the examination of individual works.

  Let us take the example of one area of literary theory such as narratology, which treats texts as examples and not as objects of analysis. If the role of criticism of a narrative text is to understand that text better, what role does narratology have? First and foremost, its role is to create narratology, just as philosophy essentially is used to philosophize. It helps to understand how narrative texts function, whether they are good or bad. Secondly, it is useful to many disciplines (such as artificial intelligence, semantics, and psychology) in helping us understand how the totality of our experience is structured (maybe) always and in every case in the shape of "narrations": a narratological theory that was useful in understanding only how stories are told would not amount to very much, but if instead it teaches us how we organize our approach to the world in narrative sequences, then it is something more. Finally, it also teaches us to read better, and even (take the case of Calvino) to invent new forms of writing. So long as we know how to make it interact with a "natural" way of reading, that is to say with a critical reading that is not set in stone at the outset by certain narratological prejudices.

  Now there are two ambiguities here, one of production and one of reception. The first is when the semiotician is not clear, or does not make clear, whether he is using the text to enrich his theory of narrative, or whether he is working with certain narratological categories in order to understand the particular text better. The second is when the reader (often prejudiced) takes as an exercise in criticism a discourse that was aimed instead at deriving general principles of narrativity from one or more individual texts. This would be like a psychologist who is interested in the motives that make someone kill reading a statistical essay on crime in the last twenty years and complaining that statistics have not provided an explanation of individual motivation.

  We could restrict ourselves to saying to these prejudiced readers that narratological theories are of no use either to the reading or to the criticism of a text. We could say that they are simply protocols of multiple readings, and that they serve the same purpose as the theory of physics, which explains how bodies fall according to one single law without telling us whether this is good or bad, nor what the difference is between a stone falling from the Leaning Tower of Pisa and an unhappy lover plummeting from a wuthering height. We could say that their purpose is to understand not texts but the function of storytelling in its totality, and that they therefore seem more like a chapter in psychology or cultural anthropology than like a chapter in literary criticism.

  A
nd yet we would have to explain also that they have, in addition, and if nothing else, a pedagogical value. They are an instrument used by those who are teaching others to read, in order to instantly identify the crucial points to which the pupil's attention has to be drawn. And so, if nothing else, they would be useful for teaching people how to read. But since one has to teach the skill of reading even to those who are no longer illiterate, they are of use also to the mature reader, the critic, and even the writer, in order to generate reliable observation.

  In short, we really need to make people understand that although a dictionary is not enough to make a good writer, nevertheless good writers often consult dictionaries. Without claiming thereby that the Zingarelli dictionary and Leopardi's Canti belong to the same kind of discourse.

  And yet, perhaps also through the semioticians' own fault, the enemies of textual semiotics cannot distinguish between the two genres of discourse (of textual semiotics and a textual criticism that is semiotically oriented). And in confusing them they lose the sense of the third type of discourse I spoke about, the only one that can help us understand the way a text reveals how it has been given form.

  When preconstituted theory precedes the reading of a text, we often mean to lay bare the tools of inquiry, which are already known, in order to show that we are familiar with them, and that we have been ingenious in constructing them, instead of possessing the artistry to conceal the art, and to make what the text finally reveals to us come directly out of the text and not out of the pyrotechnics of theoretical metalanguage. It is obvious that this method of proceeding scares the reader who wants to know something about the text, and not about the metalanguage that establishes codes of reading.

  Since a textual theory sketches out the constants, while textual criticism should highlight the variables, it often happens that, having realized that the world of intertextuality is made up of constants and variables of invention, and that a work of literature is a miracle of invention that keeps at bay and conceals the variables on which it nevertheless plays, semiotic inquiry is reduced to the discovery of the same constants in every text, and thus loses sight of the inventions.

  As a result of this we find people researching the structure of the Tarot cards in Calvino (as though the author had not already revealed everything on this subject to us), or formulaic exercises on life and death, identified in every text, which as a result reduces Hamlet to "to be / not to be," "not wanting to be / wanting not to be." In these cases, we should note, the procedure may be excellent in didactic terms, and may even succeed in showing how, whereas all of us struggle each day between wanting to be and wanting not to be, Shakespeare presents an eternal dilemma in a new way. But it is precisely from this "novelty" that our discourse should start, and narratological leveling is only a preamble to discovering the "peaks" of art.

  If literary theory exists to uncover the constants in different texts, when the critic applies the theory he should not limit himself to finding the same invariables in every text (for doing so goes no further than the work of the theorist), but if anything he should start from the awareness of the constants to see how the text calls them into question, makes them interact, and covers this skeleton with different skin and muscles in each different case. The drama of Oedipus's not wanting to know in Sophocles' play is given not by this modal structure (which can be found even in the most banal scenes where the wife who has been betrayed says to her gossipy female friend: "Please, I don't want to know") but by the strategy through which the revelation is delayed, by what is at stake (parricide and incest, as opposed to a banal example of conjugal betrayal), and by the surface of the discourse.

  Finally, semiotic discourse often fails to distinguish between "manner" and "style," which Hegel identifies in the first case as a repetitive obsession in the author who always continually writes the same way, and in the second case as the capacity constantly to outdo himself. And yet it is textual semiotics that is the only critique capable of bringing out such differences.

  If numerous different excesses can be imputed to textual semiotics, what should we say about the defects of those who oppose it? It is certainly not our job to complain about the orgasms we are made to witness by the " artifices additi artifici," who invariably provide us with a diary of their moments of languor as readers, so much so that one of their passages dedicated to author A and republished by mistake in a book dedicated to author B would go unnoticed by both editor and reviewer.

  In fact, we could leave these orgasmic critics to their own delights, which do not harm anyone, and after a while show how those who are so orgasmic in words are in fact very unlibertine in reality, and abhor alterity, since in every one of their critical embraces they are simply making love to themselves. And we could also leave to their own devices those who want to write social criticism, or the history of literary institutions, or criticism of good and bad behavior, all of which are often useful and praiseworthy.

  Except that in Italy, in the last ten years, there has been a kind of competition to see who can cast the fiercest anathemas against what they call "formalist-structural-semiotic" readings, as if the latter—and some people have even claimed this—were responsible for Tangentopolis, the Mafia, the collapse of the whining Left, and the rise of the triumphant Right.

  This could become an embarrassing incident, in the sense that these complaints could lead the young, including young teachers, astray, diverting them from a number of directions people had taken in the last twenty years, and to good effect.

  If you go into the ground floor of the Presses Universitaires bookstore in Paris and go to the second table on the right, you will find scores of manuals for all kinds of schools and at all levels on how to write an " analyse de texte." Even the very pioneers of structuralism in the 1960s have been forced to rediscover not the Russian formalists or the Prague school but the legions of good old Anglo-Saxon empiricist critics and theorists, who for decades had analyzed in depth the strategies of point of view, of narrative montage, of actants and subjects in action (as in Kenneth Burke).

  My generation, which was the first post-Croce generation, delighted in the revelations of Wellek and Warren, in the readings of Dámaso Alonso and Spitzer. We began to understand that reading was not just a picnic where one gathered here and there, almost by chance, the hawthorns and buttercups of "poesia," which were hidden amid the manure of structural fillers. Rather, one looked at the text as a whole, as something animated with life at every level. It seemed that even our culture had learned this.

  Why is it now forgetting all this? Why are young people now being taught that to discuss a text they do not need a strong theoretical repertoire, and a capacity to examine all levels? Why are they being taught that the long and lasting labors of a critic like Contini were damaging (just because—and this is true—he overrated Pizzuto), whereas the only ideal critic (now famous again!) is one with a free mind that reacts freely to the occasional solicitations provided by the text?

  Personally, I see in this tendency a reflection of other sectors of communication: criticism is being leveled down to the rhythms and rate of investment of other activities that have been proved to guarantee a profit. Why bother with reviewing, which forces one to read the book, if it sells more copies of a paper to have the literary section comment on the interview given by an author to a rival paper? Why put Hamlet on television, as the much-criticized TV of the 1960s used to do, when you can obtain higher ratings by putting on the same talk show, and treating the village idiot and the academic idiot on the same level? And why on earth read a text year after year if you can achieve the ecstasy of the sublime by chewing a few leaves, without wasting your nights and days discovering the sublimity of leaves in the sublime workings of chlorophyll in photosynthesis?

  For this is the message that is propagated daily by the high priests of the New Post-Antique Criticism: they repeatedly tell us that whoever knows about chlorophyll and photosynthesis will for the rest of his life be insensitive to
the beauty of a leaf, that whoever knows anything about the circulation of the blood will never be able to make his heart palpitate with love. And this is totally wrong, and we must say it and repeat it from the rooftops.

  This is a life-or-death battle between those who love texts and those who are simply in a hurry.

  But I shall allow an unimpeachable authority on the matter to speak, one so wise and reliable that we do not even know his real name, which should win over to his side the supporters, who are now everywhere, of wisdom that is traditional, unknown, and occult, or those sophisticated publishers who only publish one-book authors or authors who have not even managed that. We know this authority as Pseudo-Longinus, of the first century A.D., and we are inclined to attribute to him the invention of a concept that has always been the standard for those who proclaim that art cannot be discussed because it arouses ineffable emotions; we can record the ecstasy it produces, and at best we can retell the experience in other words, but it cannot be explained.

  The concept is that of the Sublime, which in certain epochs of the history of criticism and aesthetics has been identified with the specific effect of art. And indeed, Longinus (or whoever he was) states immediately that "the Sublime does not lead its listeners to persuasion, but drives them to ecstasy." When the Sublime emerges from the act of reading (or listening) "in some sense it scatters everything, like a bolt of lightning."

  The only thing is that at this point (the only point that has become famous throughout the centuries, and we are only at the end of the first section) Longinus wonders whether the Sublime can be thought about, and notes immediately that many people in his unfortunate times believe that it is an innate skill, a natural talent. But Longinus believes that natural talent can be preserved and made to bear fruit only by method, in other words by artistry, and thus he proceeds with his enterprise, which, as many people have forgotten or never known, is to define semiotic strategies that produce in the reader or listener the effect of Sublimity.