Selecciona una palabra y presiona la tecla d para obtener su definición.

  —[116]→     —117→  

ArribaAbajoCervantes as Narrator of Don Quijote

Howard Mancing

Es una premisa casi unánimemente aceptada en la teoría literaria contemporánea que un autor histórico no puede aparecer de manera alguna en un texto ficticio. Esta actitud ejemplifica, a veces, lo que Bajtín ha llamado «teoreticismo»: el confiar más en una bella teoría que en una realidad pragmática. En este ensayo se propone lo contrario: Miguel de Cervantes es el narrador de Don Quijote. Se critica especialmente la actitud binaria: algo puede ser o verdad o ficción (o autor histórico o personaje) pero no algo entre ambas cosas, ni una combinación de ellas.

Department of Foreign Languages and Literatures
Purdue University
640 Oval Drive
West Lafayette, IN 47907-2017

For Jay Allen

I. The Author

In this essay I would like to take a look at some questions often contemplated by literary scholars and casual readers alike: What is the relationship of the author to his or her text? Is the author or, at least, the author's voice, to be found in the text and, if so, where? Who tells the story; who is the narrator? As we try to identify the disembodied narrative voice that speaks to us from within the text, we might recall the image used by the Victor Talking Machine company, later RCA Victor, in the early days of sound recordings: Nipper the dog, sitting with head cocked, listening curiously to the bell-shaped speaker of a phonograph, as though wondering where his master is92. As has often been noted, it seems that when we read a novel we want to have the experience of communicating with another human being, with the author93.


But modern literary scholarship would seem to have banished the historical author from any role whatsoever in the text. Today most literary scholars labor under the influence of New Critical formalism, where the text stands alone and the author is irrelevant; or structuralism, where the text embodies culture and the author is irrelevant; and/or some form of poststructuralist theory, where the reader creates, or deconstructs, the text and the author is irrelevant. The author is dead, proclaims Roland Barthes; not quite dead, responds Michel Foucault, but merely a function of the text, and an eighteenth-century invention, at that. It is often (condescendingly) assumed to be critically naïve to talk and write of authors in serious scholarly discourse94.

I would like to bring the author -in this case Cervantes- back into consideration as I look closely at the narrative structure of Don Quijote. Few issues in Cervantine scholarship have attracted more attention than the identification of the narrative voices in the novel and the clarification of relationships among them95. A high point of sorts is achieved by James A. Parr in his book Don   —119→   Quijote, An Anatomy of Subversive Discourse, in which he describes no fewer than eleven «voices» and «presences» in Cervantes' novel and draws an elaborate chart to illustrate their relationships (30-36).

Without reviewing in detail the history of scholarly approaches to the subject of the narrative structure of Don Quijote, I would like to examine the issue in some detail, tracing the reader's perception of narrative voices from the beginning of the book to the end. In particular, I want to call into question the nearly universal assumption that the author himself, Miguel de Cervantes, is absolutely absent from the text. In doing this, I hope to illustrate the simultaneous simplicity and complexity of the narrative strategies in the text and to consider the nature of the narrative achievement of Cervantes. My method will be to take the book as it exists and read it from cover to cover, commenting in turn on each narrative element as it is perceived.

II. The Title Page

It would seem unnecessary to state the obvious: the fact that on the title page of the book it is announced that the work was «compuesto por Miguel de Cervantes Saavedra». It may be obvious, but one sometimes gets the impression that the poststructuralist assertion that authors do not exist is taken quite literally by some very sophisticated readers. There are times when one searches in vain for a reference to «Cervantes» in a great deal of current critical commentary. Parr, for example, constructs an elaborate, eleven-point «hierarchy» of what he calls «narrative voices and presences» in Don Quijote, beginning with «the extratextual historical author, a presence». First of all, it would have been easier and more direct simply to say «Cervantes» than to utilize the somewhat awkward phrase «extra-textual historical author». More to the point, though Parr goes on to discuss the role and significance of each of the voices and presences he perceives in the text, he dismisses «the historical Cervantes» in a single sentence, while the ten others all receive at least a full paragraph of attention.


It is important, however, to recognize -explicitly- that Cervantes wrote Don Quijote and that this fact is acknowledged on the title page of the book, as well as elsewhere. As we will see, the claim -the fact- of authorship by the historical human being, Miguel de Cervantes, is significant in the narration of the fictional story of Don Quijote.

III. The Prologue to Part I

Susan Lanser's The Narrative Act (1981) is one of the most comprehensive, readable, and convincing of all studies of narration. In it, Lanser describes and evaluates better than any predecessor the role of what she calls the «extrafictional voice»:

In every text, ...even a fictional text, an authorial voice does communicate historical information. This authorial voice is an extrafictional entity whose presence accounts, for example, for organizing, titling, and introducing the fictional work. This extrafictional voice, the most direct textual counterpart for the historical author, carries all the diegetic authority of its (publicly authorized) creator and has the ontological status of historical truth. This is the voice of scientific and «utilitarian» discourse as well as cultural communication. It is a voice that, but for the degree to which literary criticism has removed the author's historical presence from the text, could simply be called the authorial voice.


It seems clear that, as Lanser says, we routinely ascribe to the empirical human being called «the author» the words contained in texts of scientific or utilitarian discourse. When we read, for example, The Origin of Species, we do not ascribe the assertions in the text to any sort of narrative persona, but simply to Charles Darwin. We say that Toynbee and Prescott are the authors of the historical works that bear their names; we attribute to Hayden White or Jacques Derrida the statements in the critical texts they write. Even in autobiographical texts that consist of a mix truth and fiction, such as Benjamin Franklin's Autobiography or Vladimir   —121→   Nabokov's Speak Memory, there is no need to talk of the «narrator», the «textual persona», or the «implied author» of the text, but, simply, of Franklin or Nabokov. Similarly, if Isaac Asimov writes an introduction or a preface to a collection of his science fiction short stories, no one I have ever encountered attempts to distinguish between the voice of that preface and the voice of the historical figure of Asimov. No one routinely talks of anything like the «extrafictional voice of the introduction». There would be no point in it. Authors normally speak in their historically authorized «real» voice in order to introduce a fictional text.

In Don Quijote, the prologue of Part I is of particular interest to the reader, even though its function in the narrative scheme of Don Quijote is very often overlooked. Many literary scholars do not -indeed, feel that they must not- assume that the voice of the prologue of Part I of Don Quijote is that of the historical man, Miguel de Cervantes. For example, Parr attributes the voice of the prologue to what he calls a «dramatized author» (33), a voice separate from all others in the text, while John Weiger identifies that voice as that of an unnamed fictional character whom he calls the «prologuist». But, as I have just suggested, most common readers (and some literary critics and theorists, such as E. C. Riley and Jay Allen) assume that this voice is that of the author, Cervantes. To make matters more complicated, this supposedly fictional voice of the prologue claims some of the historical truths known to be associated with the historical author of the work. Surely it is both easier and more consistent with human psychology to accept the voice of the prologue as that of an author (Cervantes) who sometimes fictionalizes himself, than it is to conceive of it as a fictional character who sometimes «factualizes» himself by claiming, for instance, to be the author of the book we are reading, whose author we know is Cervantes.

Alberto Porqueras Mayo has studied the prologue as a literary genre in sixteenth- and seventeenth-century Spain, where it was more important and more prominent than it was in other Renaissance literatures. Porqueras maintains that the prologue is the author's direct link to the reader, the place where the author   —122→   justifies his or her writing of the text and transmits to the reader valuable information about, and important clues on how to read, the text that follows. Elias Rivers, too, has defined the prologue as a «well-defined semiacademic genre in which writers presented their works to readers» (Quixotic Scriptures 108). And this, in fact, is what most of us assume to be true of other prologues of the period. In the first part of Guzmán de Alfarache, for instance, there are two prologues to the first part, one addressed to the «vulgar» reader and the second addressed to the «discrete» reader. These, together with the single prologue to the second part of Guzmán (1604), are devices employed by the author, Mateo Alemán, in his explicit didactic aim and serve to instruct the reader on how to read the fictional text that follows.

It is worth noting some of the strategies used by Cervantes in the prologues to his other books:

  • In the prologue to his Ocho comedias y ocho entremeses, Cervantes reminisces about having watched avidly the early theatrical productions of Lope de Rueda. I do not know anyone who does not attribute this claim to the historical Cervantes.
  • In the prologue to his Novelas ejemplares, Cervantes describes his physical appearance in some detail. I do not know anyone who does not take this to be an authentic self-portrait of the historical Cervantes. He even identifies himself by his full name, Miguel de Cervantes Saavedra, in this text.
  • In his prologue to Part II of Don Quijote, Cervantes announces the forthcoming publication of his allegorical novel Persiles and promises a sequel to his pastoral novel La Galatea. I do not know anyone who does not assume that the historical Cervantes makes these statements.
  • In the prologue to Los trabajos de Persiles y Sigismunda, Cervantes tells of a recent meeting with a student who was overjoyed to be in the company of the famous writer Miguel de Cervantes. Is this a strictly factual account, some   —123→   sort of mixture of fact and fantasy or wishful thinking, or a purely fictional (and self-serving) account? Whatever we take it to be, it is a statement made by the historical Cervantes (who may or may not be narrating a fiction).

It seems quite clear to me that the extrafictional voice of the prologues that the reader perceives is that of Miguel de Cervantes96. The text of the prologue to the first part of Don Quijote, in fact, contains some claims that only the historical Cervantes could make. The claim to be the author of the book -recall the title page- or the worry about how to introduce the work to the public in a prologue can -logically- only by made by Cervantes, since authors write prologues to books they have written. The yo, then, who frets about this last pre-publication hurdle, the prologue, is Cervantes97. The problem comes in with the «friend» who enters the picture to discuss the author's dilemma.

It is commonly assumed that the scene Cervantes describes -he sits at his desk with writer's block, unable to compose his prologue, when a good friend arrives and discusses the matter with him- is a fiction. Therefore, for scrupulous, hair-splitting, modern literary scholars addicted to absolute binaries (see the discussion below in section VII), the entity who utters these words   —124→   must be a fictional character and not Cervantes. As John Weiger puts it: «Few doubt that the prologuist's friend is a fictional character. Virtually no one doubts that the conversation is fabricated. It follows that the fictitious friend's interlocutor, the prologuist, is equally fictitious» (135). Logically, however, this scene has no less formal claim to authority than the recollection of the days of Lope de Rueda in the prologue to the Ocho comedias. Who can say for a fact that Cervantes ever saw Lope de Rueda? Why could that autobiographical statement not be just as much an invention as this dramatic scene? And what if Cervantes and some clever friend really did have the conversation recorded here, or, at least, one something like it? How can we «know» the «truth» of any autobiographical text? In fact, contemporary theorists of autobiography all stress the fact that any recreation of the past, inevitably dependant on fallible memory, is ultimately an inseparable mixture of fiction and fact98.

Logically, then, the yo of all of Cervantes' prologues are either fictional entities and what they claim cannot be ascribed to the historical author, or all of them are extratextual voices -authorial voices- of substantial (but never absolute) diegetic authority. Why do we have so much trouble dealing with adjacent factual and fictional statements? Why do we have to resort to the invention of new narrative entities, personae, or masks, in an attempt to separate fictional and factual voices?

When I explain a scar on my arm by telling my daughters about the time the rustlers nearly stole the herd of cattle and how, in the ensuing gunfight, I was lightly wounded but we managed to fight off the bad guys, they have -especially now that they are older and more schooled in narrative conventions- no difficulty in distinguishing among a real scar, their real father, and a fictional tale. When I speak facetiously, sarcastically, or ironically to a group of my departmental colleagues, they do not   —125→   need to resort to any concept of narrative masks or implied department heads in order to distinguish between truth and fiction and understand what I have to say. We all know (at least most of the time) how to interpret a wink, a smile, a nod, a tone of voice, or a posture, in order to separate fact from fiction in oral discourse. And we can interpret a wide variety of contextual cues in order to do the same in many written texts (although some, such as Lazarillo de Tormes, present more difficult and perplexing problems than Don Quijote).

In the world outside the realm of literary theory there is no pretense that we need a special term for talking or writing about our concepts of real people, historical figures, or authors of books. No one speaks of an «implied president» in political discourse, even when we know that what a Bill Clinton or George W. Bush says or writes may be quite different from what the «real» person thinks. By the same token, I know Jay Allen very well, but the version of him that I understand cannot be the version of himself that holds, not is it identical to that held by any other individual. «My» Jay Allen is not «the» Jay Allen. Yet I can say «Jay Allen» (and not «the implied cervantista») and everyone knows that I am making no claim to essentialist knowledge but as simply referring to the person I know. As Michael Toolan says specifically of authors, «Even if we know an author personally, we still perform the same process of forming a mental picture or representation (itself a kind of narrative) of that author to ourselves as an integral part of the activity of knowing a person. In short, the pictures we have of authors are always constructions, so that all authors are, if you like, 'inferred authors'» (78). The conflict between reality and illusion has perhaps been the main theme of the novel since Don Quixote, and it is nothing if not ironic if we should fail to see that narrative theory deals centrally with the same problem. Do we really have to work so hard to convince students that a speaker's, writer's, or narrator's words are to be taken with a grain of salt in many -even most- contexts? Is skepticism in general, and textual skepticism in particular, really so difficult to convey? Surely we teachers and professors of literature should be able to do that job   —126→   without recourse to an arcane vocabulary and a set of imprecise concepts that no one else in any other discipline needs to employ when dealing with precisely the same issues. In effect, all of us are versions of ourselves, both to ourselves and to everyone else. As Bakhtin says, «It is customary to speak about the authorial mask. But in which utterances (speech acts) is there a face and not a mask, that is, no authorship?» (Speech Genres 152).

In the prologue to Part I, then, Cervantes pretends to discuss the nature of literary prologues with his friend99. He also states that he is not the padre but the padrastro of the book. The latter assertion is usually, and logically, linked with his claim to have searched the archives of La Mancha for the raw data he assembled into his account, just as historians, reporters, or editors assemble information from a variety of sources and write their versions of some events. For all we (absolutely) know, maybe there were some sort of local archives in La Mancha that Cervantes actually investigated. Formally, this statement has every bit as great a claim to being factual as his claim of having seen Lope de Rueda in his youth. Much more likely, however, and as is generally assumed by those who have written on the subject, this statement is presented with another textual wink, evoking the convention of the books of chivalry that were traditionally written in the guise of historical documents. So Cervantes says in effect: «I consulted all the sources and wrote this version of the story of Don Quijote; though I seem to be the author (padre), I am really only the editor (padrastro) of the text».

What this ironic posturing, intertextual allusion, and clever metatextual play does is establish the tone for the work: festive, satiric, intellectually subtle. The extrafictional voice -the voice of Miguel de Cervantes- may be engaging in a bit of fiction of its own, but it accomplishes the aims that Porqueras says prologues traditionally set out to accomplish. It prescribes the readers' horizon of expectations and contextualizes the story that is to follow,   —127→   suggesting how readers might approach the text.

IV. Part I, Chapters 1-8: The Editor and his Text

Susan Lanser distinguishes between «public» and «private» narrators (basically the same as Genett's «extradiegetic» and «intradiegetic» narrators) in a fictional text. The latter, the private narrator, is usually a fictional character and is always subordinate to the former, the public narrator. The public narrator «generally defines for the reader the story world in which s/he will function as creator and authority» (137-38). Then, breaking with the scholars and theoreticians who would insist on a mandatory and absolute separation between the public narrator and the author, Lanser proceeds to make the following assertion:

Ordinarily, the unmarked case of narration for public narrators is that the narrating voice is equated with the textual author (the extrafictional voice or «implied author») unless a different case is marked -signaled- by the text. In other words, in the absence of direct markings which separate the public narrator from the extrafictional voice, so long as it is possible to give meaning to the text within the equation author = narrator, readers will conventionally make this equation. This does not mean there are numerically more cases where author and narrator are equivalent than cases where they are not; it merely means that the separation between the two voices must be marked.


What is particularly significant here is Lanser's insistence that there must exist a textual marker in order to distinguish between the author and the narrator. Again, in «nonfiction» texts, this presents no problem: we never need, nor are we given a signal, to distinguish between the historical Roland Barthes and the voice of The Pleasure of the Text, or between the historical Jay Allen and the voice in Don Quixote: Hero or Fool? But in fictional texts the problem is more complex. Normally, the textual marker is perfectly obvious. For example, Mark Twain is the author of The Adventures of Huckleberry   —128→   Finn, but there is no temptation whatsoever to call him the narrator of the fictional text because there is a obvious and unavoidable textual marker: the narrator's name, Huck Finn. Or, again, Francisco de Quevedo is the author of El buscón but the narrator is identified as Pablos, an entity clearly distinct from Quevedo. In Daniel Defoe's Moll Flanders the narrator is Moll, not Defoe; in J. D. Salinger's The Catcher in the Rye the narrator is Holden Caulfield, not Salinger; in Ernesto Sábato's El túnel it is Juan Pablo Castel, not Sábato; and so forth. This, in fact, is the normal case: narrators explicitly have names or, at least, some characteristics that clearly distinguish them from their authors. Only in the absence of this sort of signal or marker (as is the case with the narrator «Marcel» in Marcel Proust's Remembrance of Time Past and the narrator «Carmen» in Carmen Martín Gaite's El cuarto de atrás100) do we have a problem101.

I believe that Lanser's distinctions here are clear, logical, and convincing, and that they have significant implications for Don Quijote. The greatest single stumbling block for overly sophisticated readers of the text has been the mind-set that says: public narrator ≠ [does not equal] author. Thus, the invention of a number of terms to assume the function of the author: a «persona», a «mask» or -most commonly- an «implied author». The classic articulation of this doctrine is that of the exemplary theorist, Wayne Booth, who, in his famous study The Rhetoric of Fiction (1961), asserts unequivocally that the «implied author is always distinct from the 'real man' -whatever we may take him to be- who creates a   —129→   superior version of himself, a 'second self', as he creates his work» (151)102.

When a reader moves from the extrafictional prologue of any text to the fictional chapter 1, typically there is, according to both Rivers and Porqueras, some degree of uncertainty. In the fictional text of Don Quijote, the first words encountered are: «En un lugar de la Mancha, de cuyo nombre no quiero acordarme» (I, 1; 97). Quite clearly, there is no textual marker, no signal, at this point that distinguishes the yo narrating here from the yo of the prologue. In fact, the references in the first chapter to «los autores que deste caso escriben» (98) and «los autores desta tan verdadera historia» (102) tend to reinforce the equation. The identification of the public narrator of the fictional text and the Cervantes of the prologue is then made explicit in the following passage from Chapter 2:

Autores hay que dicen que la primera aventura que le avino fue la del Puerto Lápice; otros dicen que la de los molinos de viento; pero lo que yo he podido averiguar en este caso, y lo que he hallado escrito en los anales de la Mancha, es que...


It seems clear and consistent, then, to read the text in such a way as to identify the yo who reconstructs the «history» of Don Quijote, searches the archives of La Mancha, takes note of oral tradition, and pieces together a coherent story -the editor (the voice from the prologue, the padrastro) of the story- as Cervantes. Both in the prologue and in the fictional text the whole matter is treated lightly, obviously a spoof on the explicit literary models of the books of chivalry. Readers are invited to join in the fun and   —130→   smile as they read the text, confident in the guidance provided by the author-editor-historian-narrator Cervantes.

The only passage that clouds the issue, and the one that has caused the greatest variety of responses in modern critical readers, is the final paragraph of chapter 8 where «el autor desta historia» breaks off his narration in the midst of the battle between Don Quijote and the Basque squire, claiming to have found nothing more written about the subject. Commenting on this, «el segundo autor desta obra» says that he finds it hard to believe that there is not more material available in the writings of «los ingenios de la Mancha» (153). The problem comes in identifying and distinguishing between the autor and the segundo autor. Some (such as George Haley) have seen in the figure of this «second author» a shadowy presence that ultimately controls the text; others perceive him as the translator of the Arabic text; and still others (like Parr) view him as an independent transitional figure between the archival narrator of Chapters 1-8 and the «supernarrator» of the remainder of the text from I, 9 to the end of Part II.

Much simpler and more consistent, it seems to me, is to read this passage so that el autor refers to the most recently employed historical source from the annals of La Mancha, while the term el segundo autor -who, among other things, searches again the archives of La Mancha- refers to the author/editor of the text we are reading, i. e., Cervantes. There is a shift in person from first to third, but this narrative strategy is extremely common in Renaissance texts (e. g., in Lazarillo de Tormes) and should cause no reader, whether from the seventeenth or the twenty-first century, any conceptual problem. The text, though, is ambiguous. It invites -and obviously has received- a multiplicity of readings. In accord with Occam's razor, I believe that a simple consistent answer is superior to a complex consistent answer. It is simpler -and perfectly consistent with the other evidence of the text- to read this passage as another reference to the public narrator Cervantes and his playful, metafictional search for sources.

V. Part I, Chapters 9-52: The Manuscript of Cide Hamete Benengeli

In I, 9 Cervantes tells the (fictional) story of his quest for more information about Don Quijote in order to continue beyond the point where the text was truncated in mid-adventure at the end of the previous chapter. This fictional search results in the discovery of the manuscript written in Arabic by Cide Hamete Benengeli, which, after its translation by the morisco hired to do the job, forms the main text of the remainder of Part I. The public narrator remains the same Cervantes as before; there is no textual signal that there has been a change in narrative voice, no marked indication that a new voice has assumed this role, and therefore there is no reason to assume that this is a new narrative voice.

From now on Cervantes's primary source is the translated manuscript of Cide Hamete Benengeli, but occasionally he consults other sources and/or hearsay, and from time to time includes in his edited version of the translation some marginal comments by the Muslim historian or the translator. The definitive narrative hierarchy, then, is as follows:

  • The adventures of Don Quijote and Sancho Panza,
  • as written in Arabic by the historian Cide Hamete Benengeli,
  • as translated into Spanish by the unnamed morisco,
  • as edited and written by Miguel de Cervantes,
  • as read by the Reader.

There is again some difficulty in the final paragraphs of the last chapter of Part I when we read the following: «Pero el autor desta historia, puesto que con curiosidad y diligencia ha buscado los hechos que don Quijote hizo en su tercera salida, no ha podido hallar noticia de ellas, a lo menos por escrituras auténticas» (I, 52; 591). The narration goes on to talk again about other sources such as the Manchegan archives, hearsay evidence, oral tradition, conjecture, and so forth, and ends by reproducing some festive poetry, mostly in praise of a defunct Don Quijote. Along the way, the translator adds interpretive comments and, paradoxically, Cide   —132→   Hamete even complains about the fidelity of the translation. The truth of the narrative is subverted when Cervantes, the main characters, and the translator call Cide Hamete's honesty into question, leaving the reader to contemplate a true history written by a congenital liar.

Although the term el autor has usually been employed in the text to refer to Cide Hamete, it seems clear that he is not the referent here in I, 52, for that normally omniscient historian would have no reason to search the archives of La Mancha for source material nor is it ever recorded that he had recourse to such archival sources. More reasonable is the reading that the editor -Cervantes- carried out yet again such a search after Cide Hamete's manuscript came to an end. We have seen before that Cervantes has been careless or, at least, inconsistent, in his use of terms, referring to himself as father, stepfather, author, second author, and so forth. Assuming that el autor here refers again to Cervantes as editor of the text is, I believe, both the simplest and the most consistent reading.

VI. Part II: Miguel de Cervantes vs. Cide Hamete Benengeli

Again, the title page of Part II of the novel includes an affirmation of authorship: «por Miguel de Cervantes Saavedra, autor de su primera parte». The extrafictional prefatory material this time contains the uncharacteristically long approval to print the book by the Licentiate Márquez Torres, a text that has been considered a fiction by some, and the witty and charming dedication -which consists of yet another fiction- by the historical author, Miguel de Cervantes, to his historical patron, the Count of Lemos. Since the voice of this dedication and the voice of the subsequent prologue both claim to be in the final stages of composition of Los trabajos de Persiles y Sigismunda, Cervantes' posthumous allegorical romance, it is quite clear that both belong to Cervantes. Since in the prologue Cervantes assures the reader that «esta segunda parte de Don Quijote que te ofrezco es cortada del mismo artífice y del mesmo paño que la primera» (26), it is quite clear that he again proposes himself as the public narrator of the text to follow.

In fact, all of Part II maintains the same narrative structure as   —133→   was outlined above for the bulk of Part I: Cide Hamete narrates, in Arabic, the events of Don Quijote and Sancho; the morisco translates this into Spanish; and Cervantes edits the translation and actually writes the version that we read. As before, there are times when other sources and hearsay are consulted; there are more direct citations of Cide Hamete's words and the translator's opinions are recorded more frequently than in the first part. Allen («The Narrators») and Mancing («Cide Hamete») have traced the growing tension between the public narrator, Cervantes, and the private narrator, Cide Hamete Benengeli, showing how the latter misunderstands and misinterprets characters and events and becomes himself an increasingly comic character, while the former calls the reader's attention to the historian's shortcomings and manipulates the metatextual narrative humor.

There is still one difficult and confusing passage that must be addressed; characteristically, it comes at the end of Part II, in the very last chapter of the entire novel. Having just narrated the death of Don Quijote and having just brought his story to a conclusion, Cide Hamete turns to his pen and, symbolically hanging up this tool of his trade, instructs it -the pen- to proclaim: «Para mí sola nació don Quijote, y yo para él; él supo obrar y yo escribir; solos los dos somos para en uno» (II, 74; 578). Since this passage goes on with no apparent shift in narration, we must assume that the pluma also says the following, the last words of the book: «yo quedaré satisfecho y ufano de haber sido el primero que gozó el fruto de sus escritos eternamente, como deseaba, pues no ha sido otro mi deseo que poner en aborrecimiento de los hombres las fingidas y disparatadas historias de los libros de caballerías, que por las de mi verdadero don Quijote van ya tropezando, y han de caer del todo, sin duda alguna. Vale» (II, 74; 578). Evidently some sort of change or blurring has taken place here. Since satisfecho, ufano and primero are all masculine, it is clearly not the feminine pen, la pluma, speaking at this point (as it was earlier in the paragraph when the pen was instructed to say sola, a feminine adjective). As Allen (Hero or Fool? II) has observed, there seems to be some kind of grammatical and narrative collapse at this point and   —134→   all the voices -the pen, Cide Hamete, Cervantes- fuse together (see also Anderson). Some have made much of the role of the pen, but since its appearance is so brief, at the very end of a long book, and in such an ambiguous passage, I cannot see how it has any substantial significance at all in the narrative scheme of things.

VII. Theoreticism

One of the main reasons why scholars have had so much difficulty with the narrative structure of Don Quijote is that they have too often fallen into the trap of binary thinking103: something must be either fact or fiction; you can be either a human being in real life or a fictional character or textual narrator; but you can never be something between these mutually exclusive extremes or some combination of both of them. Those scholars -formalists, structuralists, semioticians, and poststructuralists alike- who take it as a matter of faith, or as some sort of absolute and universally accepted law, that the historical author can never, under any circumstances, enter into his or her own fictional narrative are the ones who are forced by their theory to go to the greatest extremes in separating and multiplying narrative voices, always exiling Cervantes himself104. The binary assumption is what forces some to insist that the yo of the prologue cannot be Cervantes, because the prologue is a fiction and the historical human being of flesh and bone cannot be part of a fiction. It is then further assumed and asserted as fact that   —135→   the «friend» of the prologue is a fiction, and if the friend is a fiction the person he talks to must also be a fiction: fiction with fiction, and fact with fact; never the twain shall meet. So it is suggested that the prologue is written not by Cervantes but by an unnamed and totally fictional entity of some kind. The voice perceived in the prologue in turn cannot be the narrator of the fictional text (although the reason why is not clear), and so the narrators (or narrative voices) begin to multiply. In all this we have an example of what Bakhtin (Towards a Philosophy) calls «theoreticism»: the condition that exists when one's beautiful abstract theory is more important than dealing realistically and pragmatically with life, and so life is sacrificed to theory.

VIII. The Achievement of Cervantes

In conclusion, let me recall the words of one of the more famous fictional narrators of the twentieth century: Holden Caulfield, in Salinger's Catcher in the Rye. At one point in his narrative, as Holden reflects on a book he has particularly enjoyed reading, he says, «What really knocks me out is a book that, when you're all done reading it, you wish the author that wrote it was a terrific friend of yours and you could call him up on the phone whenever you felt like it» (19).

What is, I believe, so wonderful about the narration of Don Quijote is the diabolical complexity of the simple narrative scheme Cervantes employs. The increasingly complicated structures and relationships that scholars have contrived to see in the text, culminating in Jim Parr's eleven-level schema and complex conceptual graph, do not seem justified, as I have attempted to demonstrate here. Don Quijote is not a text with multiple public narrators, and its many private narrators are easily perceived for what they are.

Much more intriguing, I believe, is the magical way in which Cervantes moves us from the title page and its empirical claim of authorship, through the clever extratextual prologue that is clearly his, and into the fictional text without skipping a beat and without changing narrative voice. The ease with which he transforms himself from a historical person into a narrator of fiction, and the   —136→   inevitability of the reader's making this smooth transition from fact to fiction, is a stunning achievement. We move with Cervantes through the looking-glass into a narrative wonderland of unrivaled genius. I suggest that Cervantes' personal intimacy with readers, as he accompanies and guides us on a magical narrative adventure through the text, one of the reasons why Cervantes has had such an appeal. The historical Cervantes, the author and the narrator of Don Quijote, enchants not only general readers such as Holden Caulfield, but also to some of the finest scholars who have dealt with the text105.

Works Cited

Allen, John Jay. «The Narrators, the Reader and Don Quijote». MLN 91 (1976): 201-12.

——. Don Quixote: Hero or Fool? A Study in Narrative Technique. Gainesville: U of Florida P, 1969.

——. Don Quixote: Hero or Fool? Part II. Gainesville: U Presses of Florida, 1979.

Anderson, Ellen M. «His Pen's Christian Profession: Cide Hamete Writes the End of Don Quixote». RLA: Romance Languages Annual   —137→   6 (1994): 406-12.

Bakhtin, M. M. Speech Genres and Other Late Essays. Ed. Caryl Emerson and Michael Holquist. Trans. Vern W. McGee. Austin: U of Texas P, 1986.

——. Towards a Philosophy of the Act. Ed. Michael Holquist and Vadim Liapunov. Trans. and notes by Vadim Liapunov. Austin: U of Texas P, 1993.

Bal, Mieke. Narratology: Introduction to the Theory of Narrative. Trans. Christine van Boheeman. Toronto: U of Toronto P, 1985.

Barthes, Roland. «The Death of the Author». In Image, Music, Text. Trans. Stephen Heath. New York: Hill and Wang, 1977. 142-48.

Booth, Wayne C. The Company We Keep: An Ethics of Fiction. Berkeley: U of California P, 1988.

——. The Rhetoric of Fiction. Chicago: U of Chicago P, l961.

Bruner, Jerome. «The Autobiographical Process». In The Culture of Autobiography: Constructions of Self-Representation. Ed. Robert Folkenflik. Stanford: Stanford UP, 1993. 38-56.

Burke, Seán. The Death and Return of the Author: Criticism and Subjectivity in Barthes, Foucault and Derrida. Edinburgh: Edinburgh UP, 1992.

Canavaggio, Jean. «La dimensión autobiográfica del Viaje del Parnaso». Cervantes 1.1-2 (1981): 29-41. 8 April 2003.

Cervantes Saavedra, Miguel de. El ingenioso hidalgo Don Quijote de la Mancha. 2 vols. Ed. John J. Allen. Madrid: Cátedra, 1977.

Close, Anthony. «The Empirical Author: Salman Rushdie's The Satanic Verses». Philosophy and Literature 14 (1990): 248-67.

Couser, G. Thomas. Altered Egos: Authority in Autobiography. New York: Oxford UP, 1989.

Eakin, John Paul. Fictions in Autobiography: Studies in the Art of Self-Invention. Princeton: Princeton UP, 1985.

Eisenberg, Daniel. La interpretación cervantina del Quijote. Trad. Isabel Verdaguer. Madrid: Compañía Literaria, 1995. March 6, 2003.

El Saffar, Ruth. Distance and Control in Don Quixote. A Study in Narrative Technique. North Carolina Studies in the Romance   —138→   Languages and Literatures, 147. Chapel Hill: U North Carolina P, 1975.

Ferguson, Suzanne. «The Face in the Mirror: Authorial Presence in the Multiple Vision of Third Person Impressionist Narrative». Criticism 21 (1979), 230-50.

Fernández Mosquera, Santiago. «Los autores ficticios del QuijoteAnales Cervantinos 24 (1986): 47-65.

Flores, R. M. «The Rôle of Cide Hamete in Don Quixote». Bulletin of Hispanic Studies 59 (1982): 3-14.

Foucault, Michel. «What Is an Author?» In Language, Counter-Memory. Practice. Ed. Donald F. Bouchard. Trans. Bouchard and Sherry Simon. Ithaca: Cornell UP, 1977. 113-38.

Freeman, Mark. Rewriting the Self: History, Memory, Narrative. London: Routledge, 1993.

Genette, Gérard. Fiction and Diction. Trans. Catherine Porter. Ithaca: Cornell UP, 1993.

——. Narrative Discourse Revisited. Trans. Jane E. Levin. Ithaca: Cornell UP, 1988.

Haley, George. «The Narrator in Don Quijote: Maese Pedro's Puppet Show». MLN 80 (1965): 146-65.

——. «The Narrator in Don Quixote: A Discarded Voice». In Estudios en Honor a Ricardo Guillón. Ed. Luis T. González-del-Valle and Darío Villanueva. Lincoln, NE: Society of Spanish and Spanish-American Studies, 1984. 173-83.

Herman, David. «Autobiography, Allegory, and the Construction of the Self». British Journal of Aesthetics 35 (1995): 351-60.

Hix, H. L. Morte d'Author; An Autopsy. Philadelphia: Temple UP, 1990.

Juhl, P. D. Interpretation: An Essay in the Philosophy of Literary Criticism. Princeton: Princeton UP, 1980.

Keefer, Donald. «Reports of the Death of the Author». Philosophy and Literature 19 (1995): 78-84.

Kerr, Lucile. Reclaiming the Author: Figures and Fictions from Spanish America. Durham: Duke UP, 1992.

Killham, John. «The 'Second Self' in Novel Criticism». British Journal of Aesthetics 6 (1966): 272-90.


Kosko, Bart. Fuzzy Thinking: The New Science of Fuzzy Logic. New York: Hyperion, 1993.

Lanser, Susan Sniader. The Narrative Act: Point of View in Prose Fiction. Princeton: Princeton UP, 1981.

Lathrop, Thomas A. «Cide Hamete Benengeli y su manuscrito». Cervantes: Su obra y su mundo. Actas del I Congreso Internacional sobre Cervantes. Ed. Manuel Criado de Val. Madrid: EDI-6, 1981. 693-97.

——. «Who is the Narrator in Don QuijoteHispanic Studies in Honor of Joseph H. Silverman. Ed. Joseph Ricapito. Newark, DE: Juan de la Cuesta, 1988. 297-304.

López Navia, Santiago Alfonso. La ficción autorial en el «Quijote» y en sus continuaciones e imitaciones. Madrid: U Europea de Madrid-CEES, 1996.

Mancing, Howard. «Against Dualisms: A Reply to Henry Sullivan». Cervantes 19.1 (1999): 158-76. 8 April 2003.

——. «Cide Hamete Benengeli vs. Miguel de Cervantes: The Metafictional Dialectic of Don Quijote». Cervantes 1.1-2 (1981): 63-81. 8 April 2003.

Martín Gaite, Carmen. El cuarto de atrás. Barcelona: Destino, 1978.

Martín Morán, José Manuel. «La función del narrador múltiple en el Quijote de 1615». Anales Cervantinos 30 (1992): 9-64.

Parr, James A. Don Quixote: An Anatomy of Subversive Discourse. Newark, DE: Juan de la Cuesta, 1988.

Paz Gago, José María. Semiótica del «Quijote»: Teoría y práctica de la ficción narrativa. Amsterdam: Rodopi, 1995.

Porqueras Mayo, Alberto. El prólogo como género literario: Su estudio en el Siglo de Oro español. Madrid: Consejo Superior de Investigaciones Científicas, 1957.

Presberg, Charles D. Adventures in Paradox: Don Quijote and the Western Tradition. University Park: Penn State UP, 2001.

Riley, E. C. «El Viaje del Parnaso como narración». Cervantes: Estudios en la víspera de su centenario. Ed. Kurt Reichenberger. 2 vols. Kassel: Reichenberger, 1994. 2: 491-507.


——. «Three Versions of Don Quixote». Modern Language Review 68 (1973): 807-19.

Rivers, Elias L. «Cervantes' Journey to Parnassus». MLN 85 (1970): 243-48.

——. Quixotic Scriptures: Essays in the Textuality of Hispanic Literature. Bloomington: Indiana UP, 1983.

Salinger, J. G. Catcher in the Rye. New York: Signet, 1953.

Schippers, J. G. «On Persuading (Some Notes on the Implied Author in Critical Discourse)». Dutch Quarterly Review of Anglo- American Letters 11 (1981): 34-54.

Spacks, Patricia Meyer. Imagining a Self: Autobiography and Novel in Eighteenth-Century England. Cambridge: Harvard UP, 1976.

Stanley, Liz. The Auto/biographical I: The Theory and Practice of Feminist Auto/biography. Manchester: Manchester UP, 1992.

Stecker, Robert. «Apparent, Implied, and Postulated Authors». Philosophy and Literature 11 (1987): 258-71.

Tavris, Carol. The Mismeasure of Woman. New York: Simon & Schuster, 1992.

Toolan, Michael J. Narrative: A Critical Linguistic Introduction. London: Routledge, 1988.

Turner, Mark. Reading Minds: The Study of English in the Age of Cognitive Science. Princeton: Princeton UP, 1991.

Weiger, John G. «The Prologuist: The Extratextual Authorial Voice in Don Quixote». Bulletin of Hispanic Studies 65 (1988): 129-39.

Woodmansee, Martha. The Author, Art, and the Market: Rereading the History of Aesthetics. New York: Columbia UP, 1994.

Woodmansee, Martha, and Peter Jaszi, eds. The Construction of Authorship: Textual Appropriation in Law and Literature. Durham, NC: Duke UP, 1994.