In his first Panizzi Lecture, Don McKenzie characterized bibliography as the discipline which “can show the human presence in any recorded text”, as “the discipline that studies texts as recorded forms, and the processes of their transmission, including their production and reception”. The first aim of my lecture today is to show how some Early Modern texts, literary or not, have appropriated such processes and referred to the techniques, machines, and individuals involved in the production of “texts as recorded forms”.
For understanding these processes we have to enter into the place where books were composed, printed and corrected. For our first visit in a printing shop we can accompany a king, Philip III of Spain whose visit is related by Melchor de Cabrera Nuñez de Guzman, who was lawyer in the Royal Council. In 1675 Melchor de Cabrera published a pamphlet aimed at proving the eminent dignity of the Art of Printing and the radical difference existing between the mechanical trades and printing considered as a liberal art and, even, as the “Art of Arts”. The argumentation was destined to justify the fiscal privileges and exemptions granted by the monarch to the master printers, the correctors, the compositors or the pressmen. In order to show that the kings had always protected the liberal art of printing, Cabrera recalled that some princes have honoured the printers and compositors by visiting their workplace. It was the case of Philip III : “The king was in Lerma where the Duke [his “válido” or principal minister] had a printing shop. The king desired to see it. His daughter the infant Doña Ana (who became after queen of France) accompanied him. She went to a compositor’s case and asked that someone wrote her proper name in the palm of her hand. Then she took the types in the case according to the compositor’s indications and she put them in the composing stick. Then the king entered into the printing shop and asked that the workers [“artifices” in the Spanish text] stay seated at their place and go on with their task as they were doing. When he stopped at a case, during the time he was watching the compositorial work, he put his hand on the compositor’s left shoulder”.
By this royal gesture, Philip III expressed the dignity of an Art which was practiced, according to Cabrera, by another kink, Louis XIII of France who “had a press in his palace in Paris and who was a famous compositor”. This dignity of the art of printing is recognized by Philip III’ succesor, Philip IV, who knowing that the Duke of Medina de las Torres owned a press in the palace of the Retiro he used to print the documents necesary for lawsuits in which he was involved, “went to the place and honoured the compositors and pressmen in the same manner than his father and grandfather [Philip II].
Quite at the same time than Philip III his king, Don Quixote also visited a printing shop, not in Lerma, but in Barcelona. Let us follow him in the streets of the Catalan city in the Chapter LXII of the Second Part of the novel: “He and Sancho went out for a walk with two servants [...] Now, as they were going down a street, Don Quixote happened to raise his eyes and saw written over a door in very large letters: ‘Books printed here’ [‘Aquí se imprimen libros’], which greatly pleased him, for he had never before seen any printing and longed to know how it was done”. It is not the first time that a printing-house is the setting for a fictional work since in the 1550's William Baldwin used it as the place for the stories told “by the fire” in Beware the Cat. But in Cervantes' novel the presence of the printing-shop is more than a simple setting for the plot. It inscribes in the book itself the place and process which made possible its publication. If the work done within the actual printing-houses is a condition for giving existence to the illusions of the fiction, in Cervantes's history the terms are turned upside down since the actual and prosaic world of the workshop becomes one of the places where the imaginary narrative fuses, as wrote Borges, “the objective and the subjective, the world of the reader and the world of the book”.
Entering into the shop Don Quixote “saw all the processes of a large printing-house” [“toda aquella máquina que en las emprentas grandes se muestra”]. Cervantes introduces immediately the reader to the division and multiplicity of tasks which characterize the printing process: Don Quixote “saw them [i.e. the workers of the shop] drawing off the sheets in one place [“tirar”], correcting the proofs in another [“corregir”], setting up the types in a third [“componer”] and revising in yet another [“enmendar”].”
The acuteness of Cervantes's description of the work in the printing-house can be confirmed by a comparison with the first handbook on printing composed in a vulgar language (except if we consider as such the the 1634 German translation of Hornschuch’s Orthotypographia published in 1608 as an Instructio operas typographicas correcturis et Admonitio scripta sua in lucem edituris. It was directly set and printed in a few copies around 1680 by Alonso Víctor de Paredes who was compositor and then printer in Madrid and Sevilla. In the chapter X of his book, dedicated to “De la Corrección, y obligaciones que deven observar, assi el Corrector, como el Componedor, y el de la prensa”, that is to say to proof-reading and emendation of printed texts, Paredes distinguished between four types of “correctores”: the university graduates who know Grammar, Theology, or Law but who are not printers and ignore the technical aspects of printing, the master printers who can fulfilled rightly the task when they are learned in Latin and vulgar tongue, the more skilled of the compositors (“componedores” in Spanish) although they do not know Latin, and finally, when the printing shop belongs to a printer's widow or to a bookseller, these ignorant proof-readers “who hardly know how to read” she or he employs.
Whoever the corrector may be (except the last ones who are too ignorant) his duties are the same. First, he must detect the misprints by listening to the reading aloud of the copy (“escuchar por el original”) while he is checking the proofs. Secondly, he has the obligation to refuse the printing of a text even after it has received a “licencia”, a license, if he discovers in the work something prohibited by the Inquisition or against the Religion, the King or the “República”. Finally, the «corrector» is the one who completes the book by adding the right punctuation, by making up for the carelessness (“descuidos”) of the author, by correcting the mistakes (“yerros”) of the compositors. Such a task required that the proof-reader, whoever he is, is able to “understand the intention of the author” (“entender el concepto del Autor”) and to play a fundamental role as a necessary intermediary betwen the author and the reader.
A few year after Paredes's handbook, Moxon splits such a role between compositor and proof-reader. According to him, “A good Compositor is ambitious as well to make the meaning of his Author intelligent to the reader, as to make his Work shew graceful to the Eye and pleasant in reading. Therefore if his copy be Written in a language he understands, he reads his Copy with consideration; that so he may get himself into the meaning of the author, and consequently considers how to order his Work the better both in the title Page, and in the matter of the Book: As how to make his Indenting, Pointing, Breaking, Italicking, etc. the better sympathize with the Authors Genius, and also with the capacity of the Reader”. But all the decisions taken by the compositor are subject to the corrections of the proof-reader who is also involved in the process of publication since “he examines the Proof and considers the Pointing, Italicking, Capitalling, or any error that may through mistake, or want of Judgement, be committed by the Compositor”.
In the same manner Melchor de Cabrera splitted such duties between the compositor and the proof-reader (“corrector”). The first one “works more with his mind (“entendimiento”) than with his hands” and he has to understand the “authorial intention and discourse”. He must be expert in Castilian language, in spelling and punctuation, in the casting off of his copy “since the books are not composed following the order of the text but by formes (“alternando el original”). He must be able to punctuate rightly the discourse “for allowing his clear understanding (“su clara inteligencia”), to mark the different periods, to put in the right place accents, parentesis, question marks and exclamation points “since often the author’s mind is confused and are missing all the elements necessary for the understanding of what is written or printed. And when they are absent the meaning is changed, altered and distorted”. But as later for Moxon all the decisions taken by the compositor are submitted to the corrections of the proof-reader who must know “Grammar, Spelling, Etymologies, Pointing, and how to put accents, and this both in Latin and vulgar tongue (“romance”).
For Paredes, for Moxon, for Cabrera, for Cervantes, textual production supposes different stages, different techniques, different human agencies. Between “the author's genius and the capacity of the reader”, as wrote Moxon, a multiplicity of operations defines the process of publication as a collaborative process in which the materiality of the text and the textuality of the object cannot be separated. Paredes expressed this double nature of the book - as a material object and as a literary work - thanks to an original image. He turned upside down the classical metaphor which described the human body as a book, as in Romeo and Juliet or Richard the Second since he considered, not the human face or soul as a book, but the book as a human creature : “Asimilo yo un libro a la fabrica de un hombre”, “I compare a book to the making of a man”. Both, the book and the man, have a rational soul (“anima racional”) and a body which must be elegant, handsome and harmonious (“un cuerpo galan, hermoso, y apacible”).
If the book can be compared to a man, it is because God created human nature in the same manner than a printer print a book. For Cabrera mankind is one of the six books written by God. The other ones are the starry sky compared to an immense chart, the World itself which is a universal library encompassing all the beings, things and kwnoledges created by God, Life, “el Libro de la Vida” which is book in the format of a register containing all the names of the disciples of Christ, Christ himself (“el libro de Dios”) who is both “exemplar” and “exemplum”, and the first of all the books, the Virgin, whose creation was decided even before the Creation of the World, in the “Mente Divina”, in the Divine Mind. The Book of human nature is a printed one : “God put his image and seal on the press in order that the copy would be true to what it had to be” and “he desired rejoice himself with agreat number and a great variety of copies of his mysterious Original”.
For Paredes if the book has a soul and a body, its soul is not only the text as it was imagined, written or dictated by the author, the “buena doctrina”; it is this text given in a “acertada disposición”, an adequate presentation. If the physical body of the book is the product of the work done by the pressmen or the binders, its soul is not molded only by the author's invention, but also by the printers, compositors, or proof-readers who take care of the punctuation, spelling, or lay-out of the text. For Paredes, as later for Don McKenzie, the “substantive essence” of the work and the “accidentals” of the printed texts cannot be separated.
A small episode of the history of the publication of Don Quixote, printed in 1604 with the date of 1605 by Juan de la Cuesta in Madrid, ilustrates the evidence and also the risks of the collaborative dimension of any process of publication. In chapter XXV of the novel, Sancho refers to the theft of his ass when he declares: “God bless the man who has saved us the trouble of unharnessing Dapple [the donkey]”. Indeed, four chapter further Sancho is walking whereas his master is riding on Rocinante: “Then Don Quixote mounted Rocinante, [...] leaving Sancho on foot; which made him grieve afresh for the loss of Dapple [the donkey], whom he now missed”. But without any explanation, Sancho's donkey reappears in chapter XLII as if he had never been stolen: “Sancho [...] made himself more confortable than any of them, throwing himself down on his ass's harness”. 
Cervantes was aware of his carelessness and wrote for the second edition of his novel, printed in 1605 only some months after the publication of the first one, a brief story relating how Ginés de Pasamonte stole Sancho's donkey when this one was aslept. For the coherence of the narrative, Cervantes composed another brief story telling how Sancho recognized his thief and recovered his ass. In the second edition the first episode was inserted into chapter XXIII and the second one into chapter XXX. All seemed right but, alas, the first sentence of chapter XXV was not corrected and still read: “Don Quixote took leave of the goathered and, remounting Rocinante bade Sancho follow him, which he did on his ass [“con su jumento”], most unwillingly”. Once again, Sancho was perched on his stolen donkey... It is only in the edition printed by Roger Velpius in Brussels in 1607 that this textual inconsistency disappeared whereas curiously the third edition printed by Juan de la Cuesta in 1608 kept the misplaced allusion to Sancho's ass at the beginning of chapter XXV.
The tribulations of the stolen but always present donkey teach a double lesson. First of all, they introduce us to the instability of the texts. Their variants, oddities or vagaries result from the plurality of decisions, or blunders spread out over the different stages of their publication. The author's carelessness, the compositors's mistakes, the proof-readers's inattention : all contributed to the construction of the successive texts of the same work. How can contemporary editorial practices and literary criticism deal with such a mobility ? For Francisco Rico, the last editor of the Quixote, it is necessary to recover the text as Cervantes wrote it, composed it, or dreamt of it. Consequently, as for classical philologists, the confrontation of all the variants proposed by the different editions of the novel must lead to the choice of the more probable authorial reading - even if sometimes this reading seems betrayed by all printed editions. For some Shakespearean scholars like Margreta de Grazia, Peter Stallybrass, Stephen Orgel or Leah Marcus, on the contrary, the successive forms in which a work has been published must be considered as its different historical embodiments. As an effect of the printing-house practices and the collaborative work of many agents, each variant, even the oddest and more inconsistant, must be understood, respected and possibly edited as transmitting the work in one of the multiple modalities of its inscription. In such a perspective, he concept of an ideal original text considered as an abstract linguistic entity existing behind or beyond the different printed instantiations of the work is taken as a complete illusion. Thus to edit a work does not mean to recover this inexistent text but to make explicit both the preference given by the editor to one or another of its different “recorded forms” and the choices concerning the “materiality of the text” - that is to say its divisions, spelling, punctuation, lay-out, etc. The opposition between these two perspectives refers back to the tension underlined by Don McKenzie between the traditional idea of “a master-text, a kind of ideal-copy text, transcending all the versions and true to the essential intention of the ‘work’” and the historicity of the the different recorded versions of the work.
The compromise between the texts as recorded forms and the work in its singularity and unity is not an easy one. Let us take one Shakespearean example. In the 1598 Quarto of Love's Labour's Lost the couples of lovers are not constituted since the beginning of the play since the first flirtatious exchanges involved Biron and Katherine (and not Rosaline) and Dumaine and Rosaline (and not Katherine). It is only in the third act (according to the modern divisions of the play) that the two young lords fall in love with the woman to whom they will be linked until the end of the comedy. The version printed in the 1623 Folio proposes a very different situation since it joined since their first encounter Biron and Rosaline and Dumaine and Katherine. Is this exchange of names between the Folio and the Quarto a correction of Shakespeare's want of attention who mistook in the foul papers of the play the names he gave to characters ? Or is it a correction of a confusion made by a compositor of the Quarto ? Or is the version of the Quarto more faithful to the authorial intention since the abrupt switching of lovers's interest is a topic present in other Shakespearean plays ? The tension between the editorial decision and the literary comment as proposed by Stephen Greenblatt in the recent Norton Shakespeare seems to me a clear indication of the difficulty for answering such questions. Greenblatt wrote: “Although the version printed here is based on the near consensus among recent textual scholars, Q may provide the most accurate rendition available of the romantic relations in Love's Labour's Lost”. Thus, on the one hand, the edition respects the more coherent text as it is accepted by tradition, but, on the other hand, the aesthetic preference expresses a regret for a more interesting and exciting version...
The textual episode of the stolen donkey teaches us another lesson. With other inconsitencies, it stresses the strong relationship existing between Cervantes's mode of composition or style and the practices of orality. As Francisco Rico points out, Cervantes “revolutionized the fiction by conceiving of it not in the artificial style of literature but by using the domestic prose of life [“en la prosa doméstica de la vida”]”. For the first time a novel was written according to the syntax and rythm of spoken language and against the requirements of literary conventions and grammatical rules. Moreover, the composition of the narrative which multiplies digressions, parenthetical remarks, loose associations of words, ideas, or topics, is shaped, not by the constraints of learned rethorics, but by the freedom of oral exchanges and conversation. Mistakes, omissions, confusions did not matter a lot for such a form of writing which ignored the rules, imitated the ease of spontaneous orality, and used the printed word as if it was a way of speaking.
But let us return to the printing-house in Barcelona. Don Quixote meets there an “author” who has translated into Castilian language an Italian book entitled Le bagatele. The dialogue between Don Quixote and the translator goes on by mobilizing different elements: first, the classical debate about the exactness or inaccuracy of translations (“It seems to me that translating from one tongue to another, unless it is from those queens of tongues Greek and Latin, is like viewing Flemish tapestries from the wrong side” declares Don Quixote), secondly, an amusing reference to the great success of Ariosto's poems in Spain (“I know a little Italian [“toscano”], said Don Quixote, and pride myself on singing one of Ariosto' stanzas”), and thirdly the comic effect produced by Don Quixote's ironic admiration for very evident translations (“I will lay a firm wager than where the Italian has ‘piace’ you say in Castilian ‘place” [please], and where it has ‘piu’, you say ‘más’ [more], and ‘su' you translate ‘arriba’ [above] and ‘giù' ‘abajo’ [beneath]”.
But beyond these explicit or immediate references to common knowledge, the dialogue is built on the literary displacements of three fundamental realities of printing and book-trade. The first one recalls the two different processes for publishing a book in Early Modern Europe: “But tell me, sir, [said Don Quixote] is this book printed on your own account [“por su cuenta”] or have you sold the copyright [“el privilegio”] to a bookseller ?”. ”I am printing it on my own account” replies the translator who hopes to gain a thousand ducats with the first edition of the book. In such a case the author commissioned the printing of the edition to a printer and sold directly or through some bookseller the copies of his work. If he preferred to sell the “privilegio” to a publisher the author abandoned to him all the profit of the possible success of the book. For Le bagatele, says his translator, the press-run of the first edition has to be of two thousand copies, or “cuerpos”. This figure corresponds to one of the two situations described by Paredes’ manual since he indicated that during one sole working day a press can print either one thousand and five hundred or two thousand copies of two formes or one thousand copies of three formes. It is clear that Cervantes knew very well the practices and constraints of the economy of printing and used his knowledge for constructing some details of his novel - here, for example, the self-complacency of the translator so confident in the success of his work that he has commissioned a press-run above the average. We can recall, indeed, that in all probability Juan de la Cuesta printed only one thousand and five hundred or perhaps one thousand seven hundred and fifty copies of the first edition of the Quixote in 1604.
The second tension alluded to by the dialogue between Don Quixote and the translator of Le bagatele opposes two “economies of writing”. The first one supposes that the author can make his livings by his own or thanks to the protection of his patron. Both situations lead to write for fame. But it is not the translator's aim. He works for another goal : profit. He declares to Don Quixote: “I don't print my books to win fame in the world [“para alcanzar fama en el mundo”], for I am already known by my work. I want profit [“provecho quiero”], for fame isn't worth a bean without it”. Here too Cervantes demonstrates a precise perception of the literary world of his time since the translators were the first “authors” to receive for their works not only copies, but also a monetary remuneration – and this perhaps because they were considered as simple copists or scribes. As Don Quixote declares to the translator : “Translating from easy languages indicates no talent or power of words any more than transcribing or copying one paper from another” The opposition between “fama” and “provecho”, fame and profit, was a literary commonplace in Golden Age Spain - Cervantes handled it in the Prologue of the Second Part of his novel. It expressed a first stage in the process of professionalization of writing allowed by certain successfull genres and practices, and, by the same token, it confirmed the traditional link between desinterestedness and literary reputation.
In this same pasage Cervantes is also playing with another commonplace: the stigma of the nasty guiles and fraudulent alliances which existed between printers and booksellers for cheating the authors. Don Quixote warns the translator against such frauds as concealment of the actual press-run or the falsification of the account-books : “It is very clear that you do not know the tricks of the printing trade or the arrangements printers make with one another» [“las entradas y salidas de los impresores y las correspondencias que hay de unos a otros”]. Don Quixote's warning recalls the contemporary condemantions of printing and book-selling considered as corrupting at the same time the integrity of the texts, distorted by ignorant mechanicals, the literary ethics corrupted by cupidity and piracy, and the evidence of meaning destroyed by an uncontrolled circulation and possible misunderstandings of the works. “Dios le dé a vuesa merced buena manderecha”, “God send you good luck” are the last words addressed by Don Quixote to the too credulous and self-confident translator unconscious of the pitfalls that threaten him.
Two books (among others perhaps) are printed and corrected in the printing-house visited by Don Quixote. The first one is entitled Luz del alma (Light of the Soul). According to Francisco Rico, such a title cannot refer back to the book actually published by Felipe de Meneses in Salamanca en 1556 with the same title, Luz del alma cristiana. For Cervantes, to quote this old book inhabited by a Christian humanism would have been anachronistic and improper in relation with the Catholic commitments of the last part of his life. The title is more likely either a generic reference to the kind of books which constituted the dominant part of the Spanish printed production at the beginning of the XVIIth century at the time of the high waters of the Counter Reformation or a more precise allusion to a book which was a best-seller in Golden Age Spain : the Obras de Ludovico Blosio [i.e. Louis de Blois abbot del monastery of Liesse]. This book had more than a dozen of editions between 1596 and 1625. Juan de la Cuesta printed it several times: en 1604 for Diego Guillén, at the same time and with the same types than the First Part of Don Quixote, and, again, in 1608 y 1611 for Francisco de Robles who was the publisher of the Quixote. Whence the possible knowledge of this book by Cervantes and the implicit reference to it during Don Quixote’s in the printing house in Barcelona. Whence Don Quixote’s remark, which has to be understood seriously : “Books like this, numerous though they are, are the kind that ought to be printed, for there are many sinners nowadays, and there is need of infinite light for so many in the dark (“tantos deslumbrados”)”.
The second title discovered by Don Quixote in the printing house still is more interesting for us. It reads as Segunda Parte del ingenioso hidalgo don Quijote de la Mancha (Second Part of the Ingenious Gentleman Don Quixote de la Mancha) composed, according to the proof-reader who is correcting it, "by someone or other, native of Tordesillas”.
“I have heard of this book already” says Don Quixote. He is not the only one since the reader of the Second Part knows, at least by the allusions made in the Prologue, the existence of this apocryphal continuation of Cervantes’ novel, entitled Segundo tomo del Ingenioso Hidalgo Don Quixote de la Mancha que contiene su tercera salida, given as composed by “el licenciado Alonso Fernández de Avellaneda, natural de la villa de Tordesillas” and printed, according the title-page, in 1614 in Tarragona by Felipe Roberto. The alalysis of the types used for Avellaneda’s book suggests that this address very likely hides the real one which could the printing house of Sebastián de Cormellas in Barcelona. Such a clue reinforces the hypothesis according to which the printing house visited by Don Quixote is Cormellas’ one – and not as proposed before Pedro Malo’s printing house.
In the text itself of the Second Part, the first mention of Avellaneda's continuation appears in Chapter LIX when two patrons of the inn where Don Quixote and Sancho have stopped evoke both Cervantes's first Part and Avellaneda's Second Part. Don Juan says: “I beg you, Don Jerónimo, till they bring in the supper let us read another chapter of the second part of Don Quixote de la Mancha”. Don Jerónimo replies: “Why, Don Juan, do you want us to read this nonsense (“estos disparates”) ? Can anyone who has read the first part of the history of Don Quixote de la Mancha possibly take any pleasure in reading the second ?”.
Such a dialogue reminds to the reader the third chapter of the Second Part in which different characters (Don Quixote, Sancho, the Bachelor Sansón Carrasco) comment upon the reception of the first part of the novel, alluding to its success, proven by the great number of editions and copies (“It is my opinion there are more than twelve thousand copies of this history in print to-day” says Carrasco), but alluding also to the criticisms against the carelessness of its author: “For as there are an infinite number of fools in the world, an infinite number of people have enjoyed that history. But here are some who have found fault, and taxed the author's memory for forgetting who it was that robbed Sancho of his donkey”. The mistake made in the first editions is thus transformed into a literary motive of the narration itself. In the following chapter Sancho tells himself how and when his ass was stolen and found again and he replies to Sansón Carrasco who noticed that the mistake is “when the author says that Sancho was riding on this same Dapple, before the ass turned up again” : “I don't know how to answer that [...] All I can say is that perhaps the history-writer (“el historiador”) was wrong, or it may have been an error of the printer (“descuido del impresor”)”. Long before the visit to the printing-house, the working-place where books are printed and corrected (or poorly corrected) is present in the narrative thanks to Cervantes's irony about his own want of care - attributed to the printer.
The fact that the protagonists of the Quixote are also readers or commentators of the Quixote was for Borges one of the “partial enchantments” of the novel. For him, this literary device constituted one of the most powerful inventions thanks to which Cervantes fused the world of the text and the world of the reader : “Why does it disquiet us to know that Don Quixote is a reader of the Quixote, and Hamlet is a spectator of Hamlet ? I believe I have found the answer: those inversions suggest that if the characters in a story can be readers or spectators, then we, their readers or spectators, can be ficticious».
But the protagonists of the Quixote do not know only the novel of which they are the heroes. They have also read Avellaneda's continuation. In this sense, Cervantes’ novel does not appropriates only the techniques of the printing-house but also the literary and book-trade practices of its time. Let us return to the inn where Don Jerónimo and Don Juan have carried a copy of the apocryphal Second Part. Hearing Don Juan saying that “what most displeases me in this [book] (i.e. Avellaneda's continuation) is that it depicts Don Quixote out of love (“desenamorado”) with Dulcinea del Toboso”, Don Quixote intervenes in the conversation, denies the insulting affirmation, and makes oneself known to the two “hidalgos”. From this moment on, Cervantes plays a subtle and dazzling game with the book published one year before his own Second Part and which is supposed to tell “la tercera salida”, the “third expedition” of the knight-errant which was announced and summarized in the last chapter of the First Part by these words : “Fame has preserved a tradition in La Mancha that the third time Don Quixote left his home he went to Saragossa, and took part in some famous jousts in that city”.
In Chapter LIX of the Second Part, Don Quixote does not only refute Avellaneda's statements (he affirms that he is and will be always in love with Dulcinea), but he declares that the events the continuation recounts as already happened will never take place. When he learned that Avellaneda has related how he behoove in the tournament in Saragossa, Don Quixote decides not to go there: “I will not set foot in Saragossa. Thus I will publish this modern historian's lie to the world, and people shall see that I am not the Don Quixote he writes of”. As in Karl Popper's epistemology, Cervantes's narrative “falsifies” Avellaneda's one by giving out as something which will not happen in the future what was related as a event already occured. Don Quixote, indeed, will not go to Saragossa, but to Barcelona where he will visit a printing-house and the galleys.
In this same Chapter LIX, Don Quixote leafs through Avellaneda's book: “Don Quixote took it and without a word began to turn over the pages”. By an ironic reversal, he accused its author of carelessnes: “the third [thing], which most confirms him an ignoramus, is that he stupidly wanders from the truth in the most essential point in the whole history. For he says here that the wife of Sancho Panza, my squire, is called Mari Gutiérrez, and that is wrong, for her name is Teresa Panza; and it is much to be feared that anyone who is mistaken on so important a point will be mistaken throughout the history”. The comic criticism addressed to the negligent author of the continuation is an amusing way by which Cervantes is mocking both at his own critics, who have denounced, as Lope de Vega has done in his play Amar sin saber a quién, the episode of the stolen donkey as a lack of deference towards his readers, and at himself since in the First Part of the novel Sancho's wife appeared with several different names : Juana Gutiérrez, Marí Gutiérrez, Juana Panza, Teresa Panza, Teresa Sancha, etc. .
Multiple are the allusions to Avellaneda's continuation in the last chapters of the novel. In Chapter LXII Don Quixote leaves the printing house with these words : “I have heard of this book already [...] but truly, on my conscience, I thought it had been burnt by now and reduced to ashes for its presumption”. In Chapter LXX, in Altisidora's infernal vision, the devils are playing tennis (“jugar a la pelota”) with books in place of balls. One of these book is “the second part of the History of Don Quixote de la Mancha, not composed by Cide Hamete, its original author, but by an Aragonese who styles himself as a native of Tordesillas”. One devil cries : “‘Away with it, out of here [...] Throw it into the pit of hell, and never let me set eyes on it again’. ‘Is it so bad ?’ asked the other. ‘So bad’, replied the first ‘that if I were to set myself deliberately to make it worse I couldn't’”.
The play with the continuation culminates in Chapter LXXII where Don Quixote and Sancho met with Don Alvaro Tarfe, one of Avalleneda's original characters. After denying that the Don Quixote he meets is the true Don Quixote, Don Alvaro is obliged to recognize that the Don Quixote he knew in the novel of which he is one of the protagonists is not the right Don Quixote. This one never took part in the tournament in Saragosa; he was never shut up in a madhouse in Toledo. Don Alvaro accepts to make a declaration before the mayor of the village, which takes the judicial form of a “petición”, in which “as a matter of right” (“de que a su derecho convenía”), he declares ”that he did not know Don Quixote de la Mancha, also here present, and that it was not he who was written of in a history entitled Second Part of the Exploits of Don Quixote de la Mancha, composed by a certain Avellaneda, native of Tordesillas”. Using in a parodic way legal vocabulary and formulas, Don Alvaro's declaration is one of the devices by which Cervantes transforms into a ficticious motive Avellaneda's plagiarism. In this sense the “effets de réel” of the novel are not produced only, as noticed Borges, by its setting in the “dusty roads and sordid inns of Castile”, but, in first instance, by the permanent exchanges between the fiction and the technical or literary conditions which govern its composition (in both sense of the word) and its circulation.
The last allusion to Avellaneda's continuation is made in the last chapter of the novel and given as the last clause of Don Quixote's will : “Item, I beseech the said gentlemen, my executors, that if by good fortune they should come to know the alleged author of a history circulating hereabouts under the title of The Second Part of the Exploits of Don Quixote de la Mancha, they shall beg him on my behalf, with the greatest earnestness, to forgive the occasion I unwittingly gave him of publishing so many gross absurdities as are therein written; for I quit this life with an uneasy conscience at having given him an excuse for writing them”.
This ironic forgiveness granted to Avellaneda cannot be separated from Don Quixote's return to reason : “My judgment is now clear and free from the misty shadows of ignorance with which my ill-starred and continuous reading of those detestable books of chivalry had obscured it”. The more immediate sign of such a recovery of judgment is the recovery of the name : “Congratulate me, good sirs, for I am Don Quixote de la Mancha no longer, but Alonso Quixano, called for my way of life the Good”. This reappropriation of his true name by Don Quixote marks the real end of the fable, even before his death, since it is the exact reversal of the gesture which opened it when the “hidalgo” Quixana, as he is named in the second edition of 1605, “resolved to call himself Don Quixote”, “se vino a llamar don Quijote”. 
What is perhaps the more extraordinary moment in all the novel is then proposed to the reader when the other characters want to perpetuate the fable, to turn shepherds as they have decided and still live in a pastoral world much more pleasant than the actual one and which would allow the parody of another genre. Sansón Carrasco, Sancho, the narrator himself try desperately to uphold the fable and they keep on naming Alonso Quixano Don Quixote. A second time Don Quijote must assert he has recovered his name: “I was mad, but I am sane now. I was Don Quixote de la Mancha, but to-day, as I have said, I am Alonso Quixano the Good”. From this moment on, the two names designate the hero's double identity : on the one hand, the identity of “one of those [gentlemen] who have always a lance in the rack, an ancient shield, a lean hack and a greyhound for coursing”, and, on the other hand, the self-granted identity of a fabulous knight-errant travelling “through the world with horse and armour in search of adventures”. By using the hero's naming and “de-naming” as a key element for a “mise en abîme” of a fable - Don Quixote's adventures - within the fiction – the history as a whole -, Cervantes once again gives a profound meaning to a common characteristics of Early Modern literary works: the instability of the names. Shakespeare's Love's Labour's Lost (as we has seen) or Lazarillo de Tormes, in which, in spite of the title, the hero is never called Lazarillo but Lázaro (except in a pun on the words “lacerado” / “Lazarillo”), are examples among many others of this mobility of the names which can be assigned to the author's want of care, to the decisions or mistakes made by the master printer, the compositor, or the proof-reader, or to a shared conception of the characters in literature (and individuals the real social world) which did not suppose any fixed, stable and determined identity. Cervantes's genius transforms these social, literary or typographic variations of names, whatever their reasons may be, into an essential criterion for delineating the time and space of enchantment.
The attention paid by Cervantes to the complex and collaborative processes through wich a text becomes a book can lead us to more general conclusions. During too long a time the history of the book has firmly--and paradoxically--separated the technical conditions of the production or diffusion of printed objects from the texts they transmit, which were assumed to have a linguistic or semantic stability regardless of the variations of their material forms. In the Western tradition there are several reasons for this dissociation : the permanence of the opposition between the purity of the idea and its corruption by matter; the invention of copyright, which establishes authorial ownership over a text always identical to itself, regardless of its material support; or the development of an aesthetic that considers works independently of their particular and successive incarnations. It is the very possibility of reproducibility and the dispersion of texts in multiple states that leads, in Neoplatonic philosophy, justifications of literary propriety, or aesthetic judgment to such an abstraction of discourse.
The two critical approaches that have pleaded most strenuously for consideration of the material modalities of textual inscription have, by way of a new paradox, done more to reinforce this process of textual abstraction than to challenge it. Analytical bibliography has used an exacting attention to different states of the same work (editions, impressions, individual copies) in order to recover an ideal text, purified of the adulterations of the process of publication and true to the work as written, dictated, thought or dreamed up by its author. Hence the obsession with lost manuscripts in a discipline quite exclusively devoted to the comparison of printed objects, and the radical distinction between the perfect, ideal and unique work and its multiple, ever distorted and corrupted reproductions.
Deconstruction has, in a different way, insisted on the materiality of writing and the various forms of linguistic inscription. But in its attempt to abolish the most immediate oppositions (between orality and writing, for instance, or between the singularity of speech acts and the reproducibility of writing) this approach has proposed metacategories (arche-writing, iterability) that can only distance us from the perception of the differences they subsume. This leads to the inevitable effacement of textual materialities by conceptual categories developed at a distance from empirical evidence.
Cervantes’ acute consciousness of the constraints and possibilities of print culture and book-trade allows us to overcome such risks. He reminds us of the historicity of the diverse operations that contribute to the collective production, not only of the books, but of the texts themselves. Moreover he shows that the transactions between literature and the social world are not only purchases of objects, appropriation of languages, or symbolic acquisition of ritualistic or everyday-life practices. They are also permanent negotiations between works as poetic, immaterial, creations and the prosaic world of presses, ink, and types thanks to which they become material artefacts and commodities. In this process what is at stake is not only the circulation of social energy, but more fundamentally the modes of inscription of textual vitality.
At the same time Cervantes’enchantments caution us against the temptation to invert the inherited hierarchies and thereby to privilege the materiality of symbolic productions at the expense of their meaning. As Josph Leo Koerner has observed ocusing attention on the modalities of textual inscription might be “a way of saving the soul by looking at material but finding it haunted by subjectivity”. This is a forceful reminder that the understanding of the meanings invested in the works by their readers, listeners or spectators remains the first aim of our interpretations. In this sense the hermeneutic project remains essential. But, as Don McKenzie wrote, “new readers make new texts, and their new meanings are a function of their new forms”. We know, then, that the recovery of such a plurality of meanings can be fully achieved only if we are able to retrieve in all their singularity and differences the conceptual categories and material forms that give any text its successive historical identities.
 D.F. McKenzie, Bibliography and the Sociology of Texts, The Panizzi Lectures 1985, London, The British Library, 1986, p. 20 and p. 4 (reed. Cambridge, Cambridge University Press, 1999).
 Melchor de Cabrera Nuñez de Guzman, Discurso legal, histórico y político en prueba del origen, progressos, utilidad, nobleza y excelencias del Arte de la Imprenta, Madrid, 1675.
 Miguel de Cervantes, Don Quijote de la Mancha, Edición del Instituto Cervantes dirigida por Francisco Rico, Barcelona, Instituto Cervantes-Crítica, 1998, pp.1142-1146 . I quote the novel from J.M. Cohen's translation, Miguel de Cervantes Saavedra, The Adentures of Don Quixote, London, Penguin Books, 1950, pp. 876-878.
 William Baldwin, Beware the Cat and Funerals of King Edward The Sixth, New London, Connecticut, Connecticut College, 1963, pp. 25-63. I thank Joshua Phillips for having drawn my attention to this text.
 Jorge Luis Borges, «Magias parciales del Quijote», Otras Inquisiciones, (1952), Madrid, Alianza Editorial, pp. 74-79 [English translation: «Partial Enchantments of the Quixote», Other Inquisitions 1937-1952, Tranlated by Ruth L.C. Simms, Austin, University of Texas, 1964, pp. 43-46].
 Alonso Víctor de Paredes, Institución y origen del arte de la imprenta y Regla generales para los componedores, Edición y prólogo de Jaime Moll, Madrid, El Crotalón, 1984 [reed. Madrid, Calambur, Biblioteca Litterae, 2002].
 Ibid., pp. 42-45.
 Joseph Moxon, Mechanick Exercises on the Whole Art of Printing (1683-4), Edited by Herbert Davis and Harry Carter, London, Oxford University Press, 1958.
 Ibid., pp. 311-212.
 Ibid., p. 247.
 Jeffrey Masten, «Pressing Subjects or, The Secret Lives of Shakespeare's Compositors», in Language Machines: Technologies of Literary and Cultural Production, Edited by Jeffrey Masten, Peter Stallybrass and Nancy Vickers, New York and London, Routledge, 1997, pp. 75-107.
 Ibid., pp. 83-92.
 Miguel de Cervantes, Don Quijote de la Mancha, op. cit., 280 (English tr. p. 206].
 Ibid., p. 339 [English tr. 254] .
 Ibid., p. 499 [English tr. p. 386].
 Ibid., p. 270 [English tr. p. 199].
 Francisco Rico, «Historia del texto» and «La presente edición», in Miguel de Cervantes, Don Qujote de la Mancha, op. cit., pp. CXCII-CCXLII and CCLXXIII-CCLXXXVI.
 See for a magnificent example, Jean Bollack, L'Oedipe roi de Sophocle. Le texte et ses interprétations, Lille, Presses Universitaires de Lille, 1990, tome I, Introduction. Texte. Traduction, pp. XI-XXI et pp. 1-178.
 Magreta De Grazia and Peter Stallybrass, «The Materiality of the Shakespearean Text», Shakespeare Quarterly, Volume 44, Number 3, 1993, pp. 255-283; Leah S. Marcus, Unediting the Renaissance. Shakespeare, Marlowe, Milton, London and New York, Routledge, 1996, and Stephen Orgel, «What Is a Text», in Staging the Renaissance. Reinterpretations of Elizabethan and Jacobean Drama, Edited by David Scott Kastan and Peter Stallybrass, New York and London, Routledge, 1991, pp. 83-87.
 D. F. McKenzie, Bibliography and the sociology of texts, op. cit., pp. 29-30.
 Stephen Greenblatt, «Textual Note», The Norton Shakespeare Based on the Oxford Edition, Stephen Greenblatt General Editor, New York and London, W.W. Norton & Company, 1997, «Love's Labour's Lost», pp. 733-802 (quotation p. 739-740).
 Francisco Rico, «Prólogo», in Miguel de Cervantes Saavedra, Don Quijote de la Mancha, Nueva edición anotada al cuidado de Silvia Iriso y Gonzalo Pontón, Barcelona, Galaxia Gutenbeberg / Círculo de Lectores, 1998, pp. 9-29 (quotation p. 22).
 See other examples, particularly from Milton's tracts and speeches, in D. F. McKenzie, «Speech-Manuscript-Print», The Library Chronicle of the University of Texas at Austin, Volume 20, Numbers 1/2, 1990, pp. 87-109.
 Alonso Víctor de Paredes, op. cit., pp. 43-44.
 See for example the contracts between publishers and translators published by Annie Parent, Les métiers du livre à Paris au XVIe siècle (1535-1560), Genève, Droz, 1974.
 Ferando Bouza, «Para qué imprimir. De autores, públicos, impresores y manuscritos en el Siglo de Oro», Cuadernos de Historia Moderna, 18, 1997, pp. 31-50.
 Francisco Rico, Visita de imprentas. Páginas y noticias de Cervantes viejo, s.l., En la casa del lago, 1996.
 Cf the modern edition of Alonso Fernández de Avellaneda, El Ingenioso hidalgo Don Quijote de la Mancha que contiene su tercera salida y es la quinta parte de sus aventuras, Edición de Fernando García Salinero, Madrid, Clásicos Castalia, 1971. Cf. Edward C. Riley, «Three versions of Don Quijote», Modern Language Review, 68, 1973, pp. 807-819.
 For the identification with Cormellas’ printing house, cf. Francisco Rico, Visitas de imprentas, op. cit., pp. 48-49, who indicates that in the same year 1614 Sebastián de Cormellas printed a new edition of the Obras de Ludovico Blesio ; for the identification with Pedro Malo’s printing house, cf. Lluís C. Viada i Llluch, «L’estampa barcelonina d’En Pere i d’En Pau Malo davant de la rectoria del Pi : una conjectura cervàntica» , Bulletí de la Biblioteca de Catalunya, IV, 1925, pp. 225-237.
 Miguel de Cervantes, Don Quijote de la Mancha, op. cit., p. 1110-1111 [English tr. p. 850].
 Ibid., p. 647 [English tr. p. 486].
 Ibid., p. 655 [English tr. p. 491] .
 Ibid., p. 657 [English tr. p. 493].
 Jorge Luis Borges, «Magias parciales del Quijote», Otras Inquisiciones, op. cit., p. 79 [English tr. p. 46].
 Miguel de Cervantes, Don Quijote de la Mancha, op. cit., p. 591 [English tr. p. 457].
 Ibid., p. 1115 [English tr. p. 853].
 Ibid., p. 1112 [English tr. p. 851].
 See the essay by Edward C. Riley, “Who’s Who in Don Quijote? Or Approach to the Problem of Identity”, Modern Language Notes, 81, 1966, pp. 113-130.
 Miguel de Cervantes, Don Quijote de la Mancha, p. 1195 [English tr. p. 918].
 Ibid., p. 1208 [English tr. pp. 928-929].
 Ibid., pp. 1220-1221 [English tr. p. 938].
 Ibid., p. 1217 [English tr., p. 935-936].
 Francisco Rico, «Quexana», Euphrosyne. Revista de Filología Clásica, Volume XXII, 1994, pp. 431-439.
 Miguel de Cervantes, Don Quijote de la Mancha, op. cit., p. 42 [English tr., p. 34].
 Ibid., p. 122O [English tr., p. 938].
 Ibid., p. 35 [English tr. p. 31].
 Ibid., p. 40 [English tr. p. 33].
 Francisco Rico, «La princeps del Lazarillo. Título, capitulación y epígrafes de un texto apócrifo», in Rico, Problemas del Lazarillo, Madrid, Cátedra, 1988, pp. 113-151.
 Peter Stallybrass, «Shakespeare, the Individual, and the Text», in Cultural Studies, Edited, and with an introduction by Lawrence Grossberg, Cary Nelson, Paula A. Treichler, New Tork and London, Routledge, 1992, pp. 593-612; and Random Cloud, «“The very names of the Persons”: Editing and the Inventions of Dramatick Character», in Staging the Renaissance: Reinterpretations of Elizabethan and Jacobean Drama, David Scott Kastan et Peter Stallybrass (eds.), New York and London, Routledge, 1991, pp.88-96.
 Stephen Greenblatt, Shakespearean Negotiations : the Circulation of Social Energy in Renaissance England, Berkeley and Los Angeles, University of California Press, 1988. pp. 10-11.
 Joseph Leo Kerner, “Commentary III and Postscript”, Word & Image, Volume 17, Numbers 1 & 2, January-June 2001, “Printing Matter”, pp. 177-18O (quotation p. 180)
 D.F. McKenzie, Bibliography and the Sociology of Texts, op. cit., p. 20