From: Cervantes: Bulletin of the Cervantes Society of America 8 special issue (1988): 117-26.
Copyright © 1988, The Cervantes Society of America

The Archeology of Fiction in Don Quijote


Contemporary philosophy of language views discourse as a form of action, as a human practice not simply regulated but constituted by conventions we collectively accept or “agree upon.” Truth is no longer defined as something to be measured exclusively by objects or events outside of language, but as residing as well in those shared procedures and binding conventions by which we agree to talk to each other in certain ways and name the world with certain words. In his Philosophical Investigations, Ludwig Wittgenstein has an imaginary listener ask: “So, you are saying that human agreement decides what is true and what is false?” To which he answers. “It is what human beings say that is true and false, and they agree in the language they use. That is not agreement in opinions but in form of life.”1
     Within this view, sincerity or even sanity is no longer a matter of inner thoughts or feelings holding the ultimate truth or rationality of our words. It is rather a matter of discursive adequacy, of linguistic competence as a form of life. Knowing how to use a language, knowing how to talk is, among other things, having learned what thoughts and feelings are conventionally represented in the words we use. Contemporary philosophy of language holds a view of human reality in which self, thoughts, feelings come neither before nor after language but in it.

     1 Ludwig Wittgenstein, Philosophical Investigations, transl. by G. E. M. Anscombe (New York: The Macmillan Co., 1958) I, par. 241.


118 INÉS AZAR Cervantes

     The ideas I want to sketch and briefly explore are the following. First, that this conception of language and discourse is contemporary but not new: it is presupposed in everything we call fiction, and it was extensively explored in Don Quijote. Second, that Don Quijote explores this view by embodying the opposite view in the figure of its hero and by letting his rather solipsistic practice of discourse literally bump against the community that surrounds him and against the shared procedures and binding conventions that hold that community together. Third, that Don Quijote's madness is precisely his inability to recognize and honor the agreements that constitute our inevitably shared forms of life. Fourth, that the object of Don Quijote's undertaking is, in some fundamental ways, language and the practice of discourse. Fifth, that the key to Don Quijote's undertaking is the question of telling, which is the question of language and also the question of fiction.
     In another, well-known passage of his Philosophical Investigations, Wittgensten puts the question of telling in its simplest, most compelling form:

     Something happens —and then I make a noise. What for? Presumably in order to tell what happens. — But how is telling done? When we said tell anything? —What is the language-game of telling?
     I should like to say: you regard it much too much as a matter of course that one can tell anything to anyone (I, 363).

     There are many different things we can tell: we can tell what happens or where to find something or how to go somewhere; we can tell the time, the difference, the truth or a lie; we can tell a story. These different kinds of telling cover some of the fundamental things we are able to do in language and, in most cases, only in language: referring, describing, giving instructions, distinguishing and identifying, representing and, of course, understanding. All these uses of language that make telling possible and fiction intelligible are also the stuff of which the world of Don Quijote and Don Quijote's madness are made.
     Now, Don Quijote's madness is highly implausible: in the world we know, most people do not go crazy from reading books of fiction, although some are suspected to have committed suicide after reading Werther or Thus Spoke Zarathustra. But if it is hardly credible that somebody would go mad Don Quijote's way, it is not altogether impossible. This “implausible” possibility engages us in the story,

8 special issue (1988) The Archeology of Fiction in Don Quijote 119

while keeping us somewhat distanced from the hero; it teases us into feeling reassured that, unlike Don Quijote, we can certainly tell what is the case and what is not. Yet, on some occasions, we cans' tell, and we dons', which accounts for all the instances in which we are mistaken, duped or taken in.
     The text of Don Quijote exploits precisely the duplicity of language which, by the exercise of the same abilities, allows somebody to identify windmills as windmills but also misidentify them as giants, to get the right idea but also the wrong idea, to refer to something that actually exists but also to something that does not exist anywhere. With gestures, we are restricted to pointing only to what is present and perceptible ‘here’ and ‘now.’ With language, we are free to refer not only to what is present and perceptible but also to everything that is absent. What is absent may simply be somewhere else or may have happened at another time; but it may neither be ‘here’ nor anywhere else; it may never have happened. This ability to refer to absent things endows us with a world, a conception of an objective space that we share with others. It allows us to know other people and the world. It makes both truth and lies possible. It opens to us the possibility of fiction.
     One way to see how this possibility works and also how it works in the text and in the world of Don Quijote is to consider both the writing and the reading of fiction as a dialogic game of make-believe not unlike the mud-pie games we used to play in childhood.2 Fiction shares with the mud-pie game the ‘seriousness’ of its pretense and the need to understand conventions and respect procedures, not so much because misunderstandings and abuses could spoil the fun of the game, but simply because they would abolish the game altogether.
     The name of the children's game gives away the material substitution on which the game is based: dry mud instead of pie. The children act out a cross between metaphor and collage, in which mud is placed in the customary context of pies and is handled as a pie: it is literally decorated, sometimes with pebbles, sometimes with ‘real’ raisins; it gets metaphorically baked and on occasion burned; it can be literally served, cut into pieces, offered and exchanged; usually it is only metaphorically tasted. While playing, children are constrained as well as protected by the material limits of the game: in order to have

     2 I have borrowed the concept of games of make-believe and some general ideas about mud-pies from Gareth Evans, The Varieties of Reference (London and New York: Oxford University Press, 1982), pp. 353-63.

120 INÉS AZAR Cervantes

their ‘pie’ they have to remember all along that it is dry mud. The game calls for discrimination and care in the handling of things; it enforces duplicity but makes no room for confusion: only those who really know the difference between mud and dough can play the game, and play it well. The pretense cannot be properly upheld unless it is consciously acknowledged. The general rule of the game is that, by agreeing to abuse the normal procedures for reference in ordinary language, we will point to a piece of dry mud and call it a pie, we will pretend it is a pie, and act accordingly. In order to enter the game we must understand and agree to its conventionalized abuse of reference, to its pretense. Conversely, an abuse of the pretense, say, the ‘realistic’ demand that the make-believe pie be edible, that it actually taste like a real piece of pie and that we act accordingly, will immediately shift us from the game into reality, where we will find ourselves with plain mud in our mouth.
     While engaged in their game, children may be accurately described as willingly and knowingly misidentifying the mud they call ‘pie’ and treat as a pie, the pebbles they call ‘raisins’ and handle as raisins, the board they call ‘oven’ and use as an oven; they just pretend that something that exists and is there is other than it is. The abuse of reference that fiction invites us to indulge in is of a more radical nature than mere substitution of misidentification, since, among other things, neither pie nor mud ‘is there.’ And as soon as something is named or described in the text, we get all ‘pies,’ as it were, and no ‘mud.’ Unlike the children's game, the pretense of fiction is not based on material substitution but on the creation of existence from scratch. I can imagine no single rule which could account for the creative power of our game. But for the sake of symmetry and contrast I would suggest, very tentatively, the following: the general rule of fiction is that by the conniving abuse of reference we are able to say and understand that there is something which in fact is not there, something which plainly does not exist.
     The game that Don Quijote invites us to play is not only of fiction, but also about fiction. This much we all know and agree upon. Our disagreements arise when we try to account, for instance, for the relationship between the novel and the books of chivalry, between their respective heroes, worlds and modes of narration. This relationship has come to us already framed in terms of opposition: Amadís de Gaula is purported to be the negative model of Don Quijote: the adventures of the hidalgo are viewed as the adventures of the world of chivalry turned upside down.

8 special issue (1988) The Archeology of Fiction in Don Quijote 121

     Yet Amadís de Gaula and the books of chivalry are not simply the ‘negative models’ for Don Quijote. And, in spite of Cervantes' own words, the novel does not, or not simply, “demolish the authority and influence that books of chivalry have in the world and among common people.” First, because in Don Quijote “common people” become readers, characters who interpret texts, understand them rightly and wrongly and, in doing so, make them their own; second, because the books of chivalry call up in Cervantes' novel the presence of other books. And all these books, passionately read, believed in, misunderstood, critically examined or censored and burned, evoke in the end Literature itself. This is why, rather that “demolishing,” Don Quijote questions. And what it questions is not the fleeting authority of one particular genre, but the way in which readers confer authority upon the institution itself, upon literature and the written word.
     Thus, books of chivalry become the exemplary form of the contract between reader and institution that Don Quijote, the new text, presents as ‘the old model.’ The classical genres, the ‘old models’ are closed systems of rules that prescribe and set up a particular form of reality, an order of things. These rules direct and fulfill the reader's expectations; they promise and in effect deliver the prescribed order.
     The model that Amadís and the books of chivalry offer to Don Quijote is a space with the structure of one of those board-games where the player has to move his pieces along a fixed path. The path has its points of arrival and departure and its ‘accidents,’ all perfectly determined. In a board-game, each piece —like the knight-errant— has its destiny: to reach the pre-established point of arrival. Monsters, giants, rival knights, all the “other” characters in the books of chivalry are the obstacles that the hero must confront and overcome. The road is no more than the apparently contingent order in which these obstacles meet the knight. The characters and the road are the devices by which the text of chivalry displays and at the same time conceals its paradoxical coupling of chance and necessity, of destiny —already established— and action —free only in appearance.
     The general rule of the game of chivalry is that the knight-errant must be on the road, that his encounters must take place, that the obstacles he confronts must in the end surrender to his heroic virtue. The sallies and encounters, the unequal distribution of victory (the hero) and defeat (all his opponents), the implausible workings of magic itself must appear as ‘natural’ events that call for no explanation.

122 INÉS AZAR Cervantes

     Cervantes' novel does not stand in opposition to the model of chivalry. It confronts the old genre with the paradigm of difference. The strategy of this difference is, like dialogue itself, all inclusive and seemingly paradoxical: it assimilates without repeating and questions without demolishing; it alters objects, concepts, relationships and other texts in order to make them intelligible. Thus, the books of chivalry are ever-present in the obsessive reading of the protagonist. But Don Quijote begins when the ‘idle’ hidalgo has just stopped reading, which is precisely when we, the ‘idle’ readers of the novel, begin to read. The contact between the protagonist and his books has been a process of mutual transformation: the books of fiction he reads drive Don Quijote mad, and in his madness he turns these fictions around and forces upon them the responsibilities of history and the authority of truth. From this transformed and transforming act of reading comes the project of yet another transformation: Don Quijote intends to become a knight-errant and to make his world a world of chivalry.
     Don Quijote's enterprise begins when he realizes that his contemporary world seems to show no traces of the world of chivalry he found depicted in his books. The lack of traces of that ‘old’ world may be explained in two different ways: either the state of affairs depicted in the books of chivalry never was that of the actual world and therefore the ‘reality’ of chivalry is a mere fiction, or the state of affairs depicted in those books no longer is but was once that of the actual world. Don Quijote believes in the necessary truth of this second explanation. For him the discourse of chivalry books is referential, Amadís actually existed, and the world once was as his books say it was. Moreover, the world of Amadís —the past— had an order of its own, previous to language and unchanged by discourse, which is only its reflection. For Don Quijote, books of chivalry are not only historical evidence, they are all that now remains of the past, its only testimony.
     Now, the traditional archaeologist begins his undertaking with material objects; old coins and shards, for instance, out of which he reconstructs cities and civilizations. The new “archaeology of knowledge” proposed by Michel Foucault begins its reconstruction not with coins and shards but with texts from the past.3 These texts

     3 Michel Foucault, Les mots et les choses: une archeologie des sciences humaines, Paris: Gallimard, 1966; see also his L'Archeologie du savoir, Paris: Gallimard, 1969.

8 special issue (1988) The Archeology of Fiction in Don Quijote 123

are fragmentary evidence of a past practice of discourse (parole). The undertaking of the new archaeologist is to reconstruct, out of this fragmentary evidence, the whole system of conventions (langue) that once made that form of discourse possible. Like the new archaeologist, Don Quijote begins his undertaking with texts, in his case, books of fiction which he believes to be accurate descriptions of past objects and events. He certainly does not plan to describe the objects he may find but to find in his immediate context the objects which fit the descriptions in his books. In order to accomplish his task, Don Quijote needs now to reverse the relationship he himself had established between language and reality: if the discourse of chivalry was tailored to the old order of things, it is now the world that must be adjusted to the order of the old discourse. The description of things past becomes for Don Quijote the necessary prescription for things to come. His task is then a monumental attempt to reconstruct language and discourse, and with language, the world.
     Like the new archaeologist, Don Quijote has evidence of a prior form of speech; but his archaeology of discourse is founded in the belief that his books are direct and complete evidence of the old abstract system and that his duty is to reconstruct the old practice itself and its context of reality. His inevitable failure is written into the terms of his undertaking: a past form of life is as impossible to recover as the circumstances and the acts that once brought it into being. Thus, Don Quijote's enterprise can only be carried on in a displaced way. His new circumstances, his acts and his words will no doubt bring some practice of discourse into existence, but it will be a new practice, never the one he was trying to recover. Don Quijote's undertaking is fated to produce almost any kind of discourse, any personal identity, except the ones he was searching for. As evoked by this paradoxical enterprise, the books of chivalry are the privileged objects of the world of Don Quijote. Yet they are nothing but a pretext. Amadís is Don Quijote's goal but, in fact, he is only Don Quijote's point of departure.
     From the model of chivalry, Don Quijote derives another paradoxical aspect of his undertaking: he not only must find identity through difference and permanence through change but he must also make chance the exclusive instrument of his necessary reconstruction. Thus, Don Quijote imagines his world as a vast collection of “tuertos que enderezar” and appoints himself as the agent who will bring about the perfect order of things that his world is calling for. Having imagined this much, he only needs to be on the road.

124 INÉS AZAR Cervantes

Adventures will come to him, as they necessarily came to Amadís. It should not matter where he goes, or rather it should matter that he be on the road without destination, because part of his task is to make chance predictable, as, according to his books, chance must be. And if Don Quijote is ever going to control chance, he must, paradoxically, let chance control his journey. To do otherwise would be to admit that specific times, places, persons and circumstances —all the things that are contingent— do make a difference, that chance can never be controlled.
     Like the mud-pie game, Don Quijote's archaeology is based on material substitution, on a cross between metaphor and collage, in which a barber's basin is put in the customary context of a knight's helmet and is handled accordingly. Don Quijote abuses reference as the children do: he points to a windmill and calls it a giant, or to a pair of Benedictine friars and calls them evil enchanters; and having named them so, he acts accordingly. His problem is, of course, that while he acts according to the new identities he has assigned to people and things, these things and people dons' they remain, as it were, in a world where pies are pies and mud is simply mud. And this is necessarily so, because Don Quijote has no shared or public conventions to invoke, no game to play. He plans to succeed by prescription, since he expects that, as long as his behavior honors his words, other people and the world will oblige. But he faces reality with the utterances of a strictly private language, which he could never bring into agreement with other people or with the world. Don Quijote's enterprise shares in only the ‘seriousness’ of the children's game but not its controlled pretense. Thus, like some untutored children, Don Quijote demands, on every occasion, that his mud taste like a pie, which explains how he ends up so many times and in so many ways with ‘plain mud’ in his mouth.
     Cervantes places Don Quijote's private language in an exemplary context of shared conventions and public procedures, where even those who break some rules (e.g. “los galeotes”) acknowledge their existence and their binding force. The novel builds its own archaeology, its reconstruction of literature and fiction, by letting the deranged archaeology of the hero unfold in all its possibilities and consequences.
     The novel organizes Don Quijote's world in such a way that nothing in it will ever perfectly match the words of his books. Thus, Don Quijote must necessarily misidentify objects, persons and situations or give up his undertaking altogether. In order to

8 special issue (1988) The Archeology of Fiction in Don Quijote 125

misidentify, as he must do, Don Quijote needs to break the conventional tie between words and the world. Which opens the possibility that anything could match his words but also that nothing will. Unlike the children's game, his enterprise promotes confusion and transforms his world into a disquieting mix of games of make-believe that are neither fiction nor mud-pie. Don Quijote's enterprise questions the identity of persons and things, the authority of the past, the function of intention, of memory and desire. None of these aspects, neither Don Quijote himself nor his confrontation with reality is the key to his project. The protagonist of Don Quijote's undertaking is language, not language as an abstract structure, but language as discourse, as a form of life.
     In contrast to Don Quijote's private ‘game,’ the game of fiction abuses reference not by breaking the tie between words and the world but by invoking that tie while changing, temporarily, the direction of reference. Only because the words of Don Quijote are the words of a public language, the novel can depict characters and events and bring them into fictional existence. The fundamental gesture of Cervantes' novel, the gesture of fiction, is that those characters and events exist only in the words that name them, in the discourse we read. Fiction calls for discrimination and care in the handling of words; it enforces duplicity but makes no room for confusion, although it often explores its possibilities and dangers, only those who can tell, in all the senses of telling, can play the game of fiction, and play it well.
     By now it should be clear that my analogy between fiction and mud-pies has broken down. But having followed it this far, we should be able to see more sharply how the two games differ and what it is that makes fiction an inexhaustible game and therefore so difficult to describe. The first reason for our difficulties is that fiction is not just a game, but a language-game, and institution like language itself, or family, or culture. We are introduced, sometimes ritually, sometimes not, to most games and to some institutions. But we are born into some kind of family, some form of culture, some practice of language, and therefore of fiction. Unlike the mud-pie game, the game of fiction is neither simple, nor predictable, except in trivial ways. It has no material base and cannot be explained in terms of any equivalence. Its procedures have no fixed limits, except those of language, and its conventions are, like Don Quijote's enchanters, everywhere present but nowhere visible.
     So, I am unable to draw a definite picture of fiction along the lines

126 INÉS AZAR Cervantes

of the mud-pie game. And I believe that it is impossible to draw such a picture both accurately and completely, because our doing so depends on some basic questions about language for which we dons' have, and may never have, any satisfactory answers. Thus, for instance, we could eventually describe how fiction abuses reference, but only by begging the question of reference; and we could describe —as we indeed do— many things that are brought into make-believe existence by our fictional modes of saying and understanding, but without being able yet to tell how anyone can tell anything at all. Yet we know how to tell and understand all kinds of things, and we actually spend a good part of our lives telling and understanding them, and we know that we do so, even though we do not know how it is that we can do it. The archaeology of Don Quijote, the character as well as the book, suggests that fiction finds its foundation and its fundamental strategies precisely in this possibility of knowing language and know how to use it, without knowing how we know.


Prepared with the help of Ruth Hyndman
Fred Jehle Publications of the CSA HCervantes