From: Cervantes: Bulletin of the Cervantes Society of America 2.1 (1982): 69-87.
Copyright © 1982, The Cervantes Society of America

ROFESSIONAL EDITORS, vis-à-vis academic editors, are in an enviable position because they can deal directly with authors or the authors' executors or publishers, and thus have their editorial interventions sanctioned, or at least accepted, by an authorized person. Because they revise, prune, and correct usually new works, as yet unpublished, they have various options in the degree of intervention: they may make only minor changes and necessary corrections to the text, they may make suggestions of a very general character, they may restructure the work throughout and advise or request specific, major changes, or they may rewrite almost the whole work.
     The task of the academic editor, on the other hand, is not as clearly laid out. He edits works which have already been accepted as being worthy of notice or which are being rescued from archives or repositories where they have lain accidentally or unjustly forgotten; hence he can seldom secure authorial consent or approval for those changes which he might wish to introduce in his editions. His duties, rights, and responsibilities are not at all clear cut or necessarily obvious. These various difficulties prompt most academic editors to approach their texts, not with the detached professionalism of the man who has made a career of editing, but with a mixture of

     * For a response to this article, see John J. Allen, “A More Modest Proposal for an Obras completas Edition,” Cervantes 2.2 (1982): 181-84. -FJ.

self-defeating excess of reverence for the written word, preconceived critical ideas about previous interpretations of the work, fear of disagreeing with colleagues, and a defeatist concern for what future reviewers might say about their work. Also, they usually must take account of old-fashioned editorial policies, textual precedents, scholarly prejudices inherited from previous critical editions and manuscripts and/or early editions of questionable authority.¹ But all these obstacles, whether imaginary or all too real, can be overcome once the purposes and goals of the academic editor are understood and the editorial rules that would govern the specific edition under consideration are spelled out.
     The purpose of this article is threefold: to outline what in my opinion should be the overriding aim of scholarly editions, to give an example of how this general objective can be attained, and to emphasize the need there is to produce a scholarly, modernized edition of Cervantes' works.
     Every work of literature is unique. This uniqueness is the result of various circumstances peculiar to each work; the period when it was written, the existence or absence of authorial revisions, the language and style of the author, the literary genre of the work, and how and by whom it was either transcribed or set in type and printed. Therefore, even though academic editors may transfer methodology from one text to the other, they should nonetheless approach and deal with each text differently. Moreover, the aim of an academic editor should be simple and the same in all cases: to give readers and critics scholarly, reliable texts of his author. An editor of works printed before the twentieth century should, ideally, be familiar with press policies and compositorial habits of the period when his

     ¹ In this study I differentiate between scholarly, critical, and annotated editions. A scholarly edition is, in my opinion, any edition prepared following rigorous editorial standards and up-to-date bibliographical methods. A critical edition is an edition prepared from a holograph manuscript or a copy (or copies) of the earliest extant edition. An annotated edition is an edition having text and critical apparatus together in the same volume (or volumes). It follows, then, that critical and annotated editions may or may not be scholarly, and that a modernized, regularized, “popular,” or pocket-book edition may or may not be critical or annotated, but it can very well be a scholarly edition. I would like to thank Professors E. C. Riley and E. L. Rivers for useful suggestions. I am especially indebted to Professors W. F. Hunter, J. A. Lavin, and L. A. Murillo for their invaluable help and advice.

text was printed. He should be well acquainted with the orthography of all those who had anything to do with the original manuscript and the printed book. These acquired abilities are especially necessary when editing works written and printed during the sixteenth and seventeenth centuries.
     Unfortunately, we know very little about Spanish presses or press policies of this period, or about the spelling preferences and setting habits of the compositors who worked in these centuries, and, as a consequence, about the differences between the orthographies of individual compositors and the authors whose works they set. Cervantes and the compositors who set his works are a case in point. We know almost nothing of the particularities of Cervantes' orthography. The compositors of the first editions of Parts I and II of Don Quixote did not hesitate to alter Cervantes' text and orthography throughout their stints because these did not conform to their own styles, spellings, and punctuation habits. We can distinguish seven distinct orthographies in Don Quixote. Which is Cervantes'? None, of course. But this apparently hopeless situation in fact contains the very elements which enable us to recover the original spellings. We are indeed fortunate in having more than one or two strong-minded compositors between Cervantes' manuscripts and us.² In the first editions of Cervantes' works we have extant records which tell us not only of the countless changes that his orthography underwent at the hands of the compositors, but also, and just as important, of the creative process itself.³

     ² Had Don Quixote been set by one compositor alone, or by two compositors (one for each Part), and given the fact that we do not have an holograph manuscripts of Cervantes' works, it would be impossible to tell whether the compositor (or compositors) respected the orthography of the printer's copy, imposed his own orthography, or accepted some characteristics of the original but rejected others.
     ³ See my articles “Cervantes at Work: The Writing of Don Quixote, Part I,” Journal of Hispanic Philology, 3 (1979), 133-60, and “The Lose and Recovery of Sancho's Ass in Don Quixote, Part I,” The Modern Language Reriew, 73 (1980). 301-10. My book The Compositors of the First and Second Madrid Editions of “Don Quixote,” Part I, London: Modern Humanities Research Association, 1975, and my article “El caso del epígrafe desaparecido: Capítulo 43 de la edición príncipe de la primera parte del Quijote,” Nueva Revista de Filología Hispánica, 28 (1980), 352-60, show that the compositors of the first and second editions of Don Quixote, Part I, altered the orthography of their copy-text. Some forthcoming studies and projects now in progress will demonstrate that the compositors who set the first editions of Don Quixote, Part II, Galatea, and the [p. 72] Novelas, also imposed their own orthographic preferences. The present article is based on the theories advanced and conclusions reached in these studies.

     The tasks of editors of Cervantes' works should be, therefore, to familiarize themselves with all those bibliographical and philological techniques which will help them find out and understand, through careful textual and typographical analyses of the first editions, what exactly happened during the writing, setting, and printing of these works, and to produce editions that may be called “ideal copies” based on the first editions. An “ideal copy” is the sort of copy a professional editor would nowadays aim to produce after having had the opportunity to correct and clean the text of all imperfections, regularize, if necessary, the orthography of the original manuscript, and discuss all these editorial interventions with the author himself. Few, if any, authors would refuse to make minor changes in their work once factual inaccuracies and unseemly contradictions had been pointed out to them. Hence, an edition of Cervantes' works does not need to reproduce mechanically or uncritically the texts of the first editions, nor to coincide one hundred per cent with Cervantes' manuscripts, should we have them. Holograph manuscripts and first or subsequent editions should be respected punctiliously only when we can be certain that they are “ideal copies.” We now know that this is not the case with the early editions, nor with any other later edition, of Cervantes' works.
     A brief study of the vicissitudes undergone by the word vuestra and its masculine and plural forms at the hands of the compositors of the first editions of Parts I and II of Don Quixote will illustrate the tangled orthographic web we now have in these editions and the possibilities of unraveling it.
     In Don Quixote the reading vuestra and its related forms appear set as vuea(s), vuetra(s), v(s)., V(s) or va(s).

		Don Quixote, I				Don Quixote, II
vuea(s)		3		1%		215		31%
vuetra(s)		428		90%		125		19%
Abbreviations		38		9%		338		50%

     Traditionally, most editors of annotated editions have retained both spellings as they occur in the first editions (modernizing vuea[s] to vuesa[s]), and have resolved abbreviations to vuestra(s). But the percentages of the two forms in the different Parts are totally different (see above), suggesting either that Cervantes changed his spelling preference sometime between the writing of Part I and Part II, or that the compositors tinkered with the text.
     Fortunately, this puzzle is not difficult to unravel. Three characteristics of the distribution of the word and its spellings can be noted initially: 1) the per-gathering percentages of the form vuestra(s) and abbreviations in Part I, and the forms vuestra(s), vuessa(s) and abbreviations in Part II, are strikingly even within each Part; 2) in Part I only gathering ¶¶ has the form vuessa(s) (three occurrences); 3) in Part II only gatherings ¶ (preliminaries), A, B, D, and 2N (index) have no occurrences of the vuessa(s) form (gathering 2N has no occurrences of any form). The fact that the use of these forms varies between particular gatherings matching the changes of compositor, and from Part to Part clearly implies that the variations are the result of strong compositorial preferences. Cervantes, of course, had nothing to do with the division into gatherings. Whereas the compositors of Part I and the compositors of gatherings A, B, and D of Part II apparently considered the form vuessa(s) improper and avoided setting it throughout their stints (this precise demarcation cannot be of Cervantes' making), the other two compositors and the apprentice who set copy for Part II accepted this form without too much misgiving.⁴ It is also apparent that the compositors of the first edition of Part I shunned setting abbreviations of these forms, in spite of the fact that on numerous occasions they were short of space to met their stints due to the casting off of too much copy (the only exception is compositor F in the inner sheet of gathering X, which has eight abbreviations, a saving of approximately one and a quarter lines of type set, but this is the exception that proves the rule). The compositors of Part II, on the contrary, set a great number of abbreviations per gathering though they were in no need of extra space.<sup>5

     4</sup> Gathering ¶ of Part II has two occurrences of the possessive pronown vuestra(s) and nine occurrences of the abbreviated form, but this gathering was set by compositor I, who used the form vuessa(s) in his other gatherings.
     ⁵ Another compositorial practice evident in both Parts was to set comparatively few abbreviations of the plural feminine form vuestras [p. 74] (seventeen abbreviations out of a possible eighty-five instances), and of the singular and plural masculine forms vuestro(s) (twenty-six abbreviations out of a possible two hundred and two instances).

     The next step is to ascertain whether or not there is any sort of internal coherence in the use of either the abbreviations or the different spellings in the two Parts. If we check Part I alone we find that the form vuestra(s) is used almost exclusively (the only exceptions are three occurrences of the form vuessa(s) —to which I shall return later, with other exceptions noted below— and thirty-eight abbreviations —which are not used in any characteristic way); but if we examine Part II we find that, with three exceptions, all occurrences of the form vuessa(s) are in the title vuessa(s) merced(es) (see the table below, items 5-7), and all but three occurrences of this title contain the form vuessa(s) (see items 8-10) or an abbreviated form.⁶ With all other nouns (singular and plural, masculine and feminine), excepting the three instances mentioned above (items 5-7), the forms used are vuestra(s) and vuestro(s). Given the strength of compositorial preferences, as noted earlier, and the facts that the grammatical patterns vuessa(s) merced(es) over against vuestra(s) or vuestro(s) + noun appear only in Part II, and in Part II are not followed by the compositors of gatherings A, B, and D, who instead abbreviated all fifty-one occurrences, we could assume that the form vuessa(s) was a compositorial rather than an authorial spelling. Some additional textual evidence, however, shows such an assumption to be premature, and indeed erroneous. But first we must restate one of the previous premises. Even though the word vuestra was frequently pronounced vuessa —initially only the form vuestra was considered proper in writing— and the grammatical pattern vuessa(s) merced(es) is characteristic only of Part II, the fact that the vuessa form appears in both Parts alongside the vuestra form suggests that Cervantes might have used both spellings in his manuscript, and that the compositors of Part I and of gatherings A, B, and D of Part II indeed considered that the form vuessa(s) should not be used in writing. The problem therefore, as it now appears, is whether or not Cervantes, aware as he must have been of this linguistic phenomenon, used both forms purposely and systematically.

     ⁶ For a detailed study of the development of the title vuestra merced see José Pla Cárceles, “La evolución del tratamiento ‘vuestra-merced’,” Revista de Filología Española, 10 (1923), 245-80. For a list of other monographs written on this subject see Rafael Lapesa, Historia de la lengua española, 8th edition, Madrid, 1980, p. 393, Footnote 62.

The following quotations from Don Quixote will permit us to come to some conclusions in this respect.⁷

VUESTRA(S) + [any noun except merced(es)]  — >VUESSA(S)

VUESTRO —> VUESSO

1. “Seran vueas fazañas los joezes” (I, ¶¶ 8v, line 13; sonnet “De Solidan, a don Quixote”).
2. Y i la vuea linda Dulzinea” (I, ¶¶ 8v, 7; sonnet “De Solidan a don Quixote”).
3. “Ni a vueas cuytas muetra buen talante” (I, ¶¶ 8v, 9; sonnet “De Solidan a don Quixote”.
4. “En tal deman vueo conorte sea” (I, ¶¶ 8v, 10; sonnet “De Solidan a don Quixote”.
5. “Dezid amiga mia, a vuea eñora, [dixo don Quixote]” (II, 23, M2v, 21-22).
6. “la carta que vuea Grandeza me ecriuiô” (II, 52, BbA, 17-18 Teresa's letter to the Duchess).
7. “uplico a vuea excelencia” (II, 52, Bb8v, 1; Teresa's letter to the Duchess).
   

   VUESSAS(S) MERCED(ES) —> VUESTRA(S) MERCED(ES)

8. “vuetras mercedes eñores e pongan en cobro, [dixo Sancho]” (II, 17, H5v, 1-2).
9. “eta [merced] que la vuetra merced oy me ha fecho, [dixo don Quixote]” (II, 32, Q7v, 25-26).
10. “no auria navaja que con mas facilidad rapae a vuetras mercedes como mi epada raparia de los ombros la cabeça de Malambruno, [dixo don Quixote]” (II, 40, T8-T8v, 32-34, 1).

It is obvious that Cervantes had a clear stylistic purpose in mind when he used the form vuessa(s) before nouns other than merced(es) (items 1-3, and 3-7). In Solisdán's sonnet we find, in addition to the four occurrences with an improper use of vuessa(s) the only instance of the masculine reading of this form that occurs in Don Quixote (vuesso, item 4), several archaisms (see Murillo's critical edition, Madrid: Clásicos Castalia, 1978, I, 67, footnote 24), and the colloquialism “alcaguete” (I, ¶¶, 8v, 11). Although the compositor of gathering ¶¶ of the first edition of Part I was apparently firm on changing Cervantes' vuessa(s) for the compositorial form vuestra(s) throughout all the

     ⁷ These quotations include the only ten exceptions to the grammatical rule discussed above.

other gatherings he set (gatherings I, K, M, N, Q, R, V, Y, Bb, Ff, Kk, Nn, *, and **), he could not avoid realizing that in this sonnet Cervantes was purposely using the form vuessa(s) before nouns other than merced(es).⁸ Nor is it coincidental that the three other occurrences of the “improper” form (items 5-7) are used by Don Quixote when he is talking to a peasant girl, servant of the enchanted Dulcinea (episode of the Cave of Montesinos), and by Teresa Panza in, pointedly, her letter to the Duchess. In these excerpts Cervantes is making gentle sport of the archaic language used in romances of chivalry, the condescending attitude of Don Quixote towards the peasant girl, and the monazillo who wrote Teresa's letter to the Duchess (as in the case of vuestra merced, the proper written forms of the titles vuestra grandeza and vuestra excelencia probably required the learned form; see Cervantes' dedication to the Duque de Bejar, “vue[]tra Excelencia,” I (¶ 4, 10-11, 19-20, and ¶ 4v, 9). Furthermore, these readings escaped the misplaced zeal of the compositors only because the passages where they occur are laden with other archaisms, colloquialisms, and rustic constructions, and the compositors could not be blind to this fact.⁹
     When studying the exceptions listed in the second group, one could make a similar argument to that made for the exceptions listed in the first group (Sancho uses one of them, and the other two are uttered by Don Quixote in two passages where he is in his imaginary

     ⁸ Pla Cárceles considers that the form vuessa derived directly from the learned (written) form vuestra: (Lat.) vostra > vuestra > vuessa; but he studies this form only as an element of the title vues(tr)a merced (his only mention of the form vuestra alone —the excerpt he quotes has the form vuessa— appears on page 252, footnote 2), leaving out the likely possibility that the forms vuessa(s) and vuesso(s) might have appeared earlier, independent of the title. The uses and exact origin of this form are in fact not clear. In his Manual de gramática histórica española (sixth edition, Madrid, 1941) Ramón Menéndez Pidal notes: “En el caso de la [combinación] STR hay una solución ss que se halla en algunas voces hoy desusadas; . . . vuesso . . . puede remontar al latin vulgar; vuesa merced” (p. 145, item 51, 1), and, “La lengua antigua y vulgar conoce otra forma: vuesso” (p. 258, item 97, 1); and Juan de Luna (Diálogos, Paris, 1619) states: “si (el que habla) es de los más ladinos dize vuesasté, el común, vuesa merçed, y los mis rústicos vuestra merçed,” quoted by Pla Cárceles, p. 259.
     ⁹ If one wishes to find further textual evidence in other works of Cervantes to support these theories and conclusions, Galatea offers excellent text ground. In this work Cervantes had, of course, no intention of dwelling on or stressing ironically the differences between proper and improper forms because, from a linguistic point of view, all his characters are equals —shepherds shepherdesses, villagers— hence we would look in [p. 77] vain for an occurrence of the title vuestra merced in any form, or of any improper use of the forms vuessa(s), vuesso(s), or vuestra(s) even though the first edition of Galatea was set by more than one compositor and their setting habits and spelling preferences varied from each other. In the Novelas, in the Entremeses, and in Cervantes' poetry we will, on the other hand, find all these form again, and also some occurrences of the syncopated vulgarism boacé (Ginés de Pasamonte is the only character in Don Quixote who uses the form boacé, Part I, N8, 12, and N8v, 2; did the disguised Ginés use the vulgarism also in Part II in his character of Maese Pedro?).

world of chivalry), but in my opinion this textual evidence is insufficient and inconclusive. I would tend to assign these three readings to unconscious, compositorial standardization (as opposed to the conscious standardization that took place in Part I) rather than to any stylistic purpose on Cervantes' part.
     From all this textual and typographical evidence we can safely conclude that it is the compositors who are responsible for the striking variations in the use of the forms vuessa(s), vuestra(s) and abbreviations between gatherings and between Parts I and II of Don Quixote, rather than Cervantes, who is very unlikely to have changed his spellings so radically and so precisely between the writing of the two Parts. We can conclude also 1) that Cervantes used the spellings vuessa(s), vuesso(s), vuestra(s), and vuestro(s) in his works, 2) that in all probability he did not abbreviate these forms in his manuscripts (otherwise the linguistic differentiation would have been lost), 3) that he acknowledged and accepted the grammatical rule which prescribed the use of the form vuestra(s) in writing, 4) that he used the most common speech-form vuessa(s) in his dialogues, but only as an element of the title vuessa merced, and 5) that he regarded the forms vuessa(s) and vuesso(s)+noun as incorrect and used them to denote improper or archaic constructions.
     It is now obvious that the compositors' standardization was imposed on top of the distinctions of usage that Cervantes was deliberately employing. The seven readings of group one appear in passages that show they are meant to represent the linguistic idiosyncracies of the characters who use them, and because Cervantes' intentions am evident we cannot consider them either typographical errors, compositorial whims, or mere coincidences. Textual evidence is so clear in this respect that we cannot but wonder in how many other instances the compositors might have remained blind to the linguistic craftsmanship of Cervantes and might have obliterated in a matter of seconds what had taken him so many years to create. It needs to be

emphasized, nonetheless, that even though the first editions of Cervantes' works reflect only vaguely the original punctuation and orthography of the now lost manuscripts and are a hodgepodge of several different spelling and punctuation preferences, they should be the only text used as the basis for editing Cervantes' works. At no time should editors allow themselves to base their editions on texts other than those of the first editions.
     What is needed, then, are editions which will give readers and critics texts which in a thorough, scholarly, and systematic fashion regularize the countless compositorial vagaries found in the first editions, without, on the other hand, falsifying or ignoring Cervantes' masterly handling of the language, and without giving undue authority to later editions or to non-Cervantine manuscripts.
     These goals can be attained in three ways: either by recovering Cervantes' orthography and producing an old-spelling edition of his works with accidentals of punctuation, written accents, layout, etc., based on the first editions; by recovering Cervantes' language, modernizing the accidentals only, and producing a scholarly, annotated, regularized edition based on the first editions; or by updating the orthography and language of the first editions to present-day usage without obliterating the various linguistic levels intended by Cervantes and producing a scholarly, modernized edition. This is not to say, of course, that future editors have carte blanche to do with the text of the first editions what they will, introducing or retaining only their own preferred readings, or forcing their own critical bias in what should be, above all, a bibliographical and philological analysis of the texts. Editors of old-spelling, regularized, and modernized editions should strive to retain Cervantes' various linguistic levels and styles. Old-spelling regularization should not mean out-and-out archaization, nor ankylozation of the language. Regularization of accidentals should not be carried out without taking into account possible stylistic and linguistic variations. Modernization should not mean vulgarization of the language, nor partial obliteration of any textual or stylistic characteristics. I am already at work on an old-spelling edition of Don Quixote and of the other works of Cervantes, but this project will require many more years of work before reaching completion; and a scholarly, annotated, regularized edition cannot be produced safely until after Cervantes' orthography and lexicon have been recovered. There should be, however, an impending sense of urgency in doing justice to Cervantes now that we know that his orthography and style were drastically and systematically altered by

the compositors who set the first editions of his works. This aim could be accomplished now by producing a scholarly, critical, modernized edition of his works. What follows is a brief summary of the general editorial policies and conventions I would propound for producing such an edition.
     The editors of a scholarly, critical, modernized edition should update both the orthography and the lexicon of the first editions without sacrificing any of the linguistic levels of the “ideal copies” made up from these editions, for example, although the title vuestra merced is still used in some rural areas, it is nowadays unacceptable in every-day social intercourse, having been almost completely replaced by the polite shortened from usted. Hence, in a truly modernized edition most occurrences of this title and of the related form su(s) merced(es) (used to refer to third person singular and plural) should be replaced by the appropriate updated form (usted[es] and ella[s], él [ellos], el señor oidor, etc.). But forms of address still in use (vuestra grandeza, vuestra señoría, etc.), related humorous forms (vuestra pomposidad), forms which occur in passages where archaisms are used (vuestra should be replaced by vuesa to differentiate this use from forms of address still in use and related humorous forms), and incorrect uses of the forms vuessa(s) and vuesso(s) should be retained. It follows that only when the title vuestra merced is replaced by the word usted, should corresponding possessive pronouns and adjectives (vosotras, vos, vuestro, etc.) be replaced by the appropriate form su(s), suya(s), or suyo(s). When these pronouns and adjectives belong to the second person plural familiar (vosotras, vosotros), they should remain. The voseo should also remain untouched. Ginés' colloquialism boacé should be replaced by the related form asté. The spellings of the occurrences retained should be modernized and regularized (vuea[s], vueo[s], vuetra[s], vuestro[s] — vuesa[s], vueso[s], vuestra[s], vuestro[s]), and all abbreviations should be resolved according to the general rules outlined above. Sixteenth- and seventeenth-century typographical conventions should not be observed. The following excerpts give an idea of what modernization would entail:

Don Quixote		Don Quixote
First editions		Modernized edition
1.
EN FE Del buen acogimiento, y honra, que haze vueta Excelencia a todo uerte de libros, como Principe tan inclinado a fauorecer las buenas arte, . . . (I, ¶ 4, 8-14).		En fe del buen acogimiento y honra de que hace vuestra excelencia a toda suerte de libros, como príncipe tan inclinado a favorecer las buenas artes, . . . .
SONETO.		Soneto
2.
MAguer eñor Quixote, que andezes		Maguer, señor Quijote, que sandeces
Vos tengan el cerbelo derrumbado,		vos tengan el cerbelo derrumbado,
Nunca ereys de alguno reprochado,		nunca seréis de alguno reprochado
Por home de obras viles, y oezes.		por home de obras viles y soeces.
Seran vueas fazañas los joezes.
Pues tuertos desfaziendo aueys andado,		Serán vuesas fazañas los joeces,
Siendo vegadas mil apaleado,		pues tuertos desfaciendo habéis andado,
Por follones cautiuos, y rahezes.		siendo vegadas mil apaleado
Y i la vuea linda Dulzinea,		por follones cautivos y raheces.
Deaguiado contra vos comete,
Ni a vueas cuytas muetra, buen talante.		Y si la vuesa linda Dulcinea
En tal deman vueo conorte ea,		desaguisado contra vos comete,
Que Sancho Pança fue mal alcaguete,		ni a vuesas cuitas muestra buen talante,
Necio el, dura ella, y vos no amante.
(I, ¶¶ 8-¶¶ 8v, 26-28, 1-12)		en tal desmán vueso conorte sea
		que Sancho Panza fue mal alcagüete,
		necio él, dura ella, y vos no amante.
3.
Età bien quanto vuetra merced dize, dixo Sancho. Pero querria yo aber (por i a caso no llegae el tiempo de las mercedes, y fuee neceario acudir al de los alarios) quanto ganaua vn ecudero, de vn cauallero andãte en aquellos tiempos? y i e concertauan por mees, o por dias, como peones de albañir? (I. M5, 10-16).		—Está bien cuanto usted dice— dijo Sancho —pero querría yo saber (por si acaso no llegase el tiempo de las mercedes y fuese necesario acudir al de los salarios), cuánto ganaba un escudero de un caballero andante en aquellos tiempos, y si se concertaban por meses, o por días, como peones de albañir.
4.
Señor comiario, dixo entõces el galeote, vayae poco a poco, y no andemos aora a delindar nombres, y obrenombres. Gines me llamo, y no Gineillo, y Paamonte es mi alcurnia, y no Parapilla, como boace dize, . . .(I, N8, 8-13).		—Señor comisario— dijo entonces el galeote —váyase poco a poco, y no andemos ahora a deslindar nombres y sobrenombres, Ginés me llamo, y no Ginesillo, y Pasamonte es mi alcurnia, y no Parapilla, como asté dice, . . . .

5.
Dexeme V. m. eñor Licciado, no es razon yo etè a cauallo, y vna tan reuerda perona como V. m. etè a pie. Eo no conentiré yo en ningú modo, dixo el Cura, etee la vuetra grãdeza a cauallo, pues etando a cauallo acaba las mayores fazañas, y auenturas en nuestra edad e han vito, que a mi aunque indigno acerdote, batarame ubir en las ancas de vna detas mulas detos eñores con V. m. caminan, ino lo han por enojo: . . . (I, X4, 6-14).		—Déjeme usted, señor licenciado, que no es razón que yo esté a caballo y una tan reverenda persona como usted esté a pie.     —Eso no consentiré yo en ningún modo— dijo el cura. Estése la vuesa grandeza a caballo, pues estando a caballo acaba las mayores fazañas y aventuras que en nuestra edad se han visto; que a mí, aunque indigno sacerdote, bastaráme subir en las ancas de una de estas mulas de estos señores que con vuestra merced caminan, si no lo han por enojo.
6.
. . . vos hermano, ydos a er gouierno, o inulo, y entonaos a vuetro guto, . . .  ydos con vuetro dõ Quixote a vuetras auenturas, y dexadnos a nootras con nuetras malas vturas Dios nos las mejorarâ, . . . (II, C2, 23-30).		. . . Vos, hermano, idos a ser gobierno o ínsulo, y entonaos a vuestro gusto, . . . Idos con vuestro don Quijote a vuestras aventuras, y dejadnos a nosotras con nuestras malas venturas, que Dios nos las mejorará . . . .
7.
. . . quãdo yo vi ee ol de la eñora Dulcinea del Toboo, que no etaua tan claro, pudiee echar de i rayos algunos, y deuio de er, que como u merced etaua ahechãdo aquel trigo, . . . (II, D3, 8-11).		. . . cuando yo vi ese sol de la señora Dulcinea del Toboso, que no estaba tan claro que pudiese echar de sí rayos algunos, y debió de ser que como ella estaba ahechando aquel trigo . . . .
8.
MVcho contento me dio, Señora mia, la carta vuea Grandeza me ecriuiô, en verdad la tenia bi deeada: la arta de corales es muy buena, y el vetido de caça de mi marido no le va en zaga: de V. S. aya hecho Gouernador â Sancho mi conorte ha recebido mucho guto todo ete lugar, . . . (II, Bb8, 17-22).		Mucho contento me dio, señora mía, la carta que vuesa grandeza me escribió, que en verdad que la tenía bien deseada. La sarta de corales es muy buena, y el vestido de caza de mi marido no le va en zaga. De que vuesa señoría haya hecho gobernador a Sancho, mi consorte, ha recebido mucho gusto todo este lugar, . . . .
9.
. . . allâ os auenid, eñoras, con vuetros deeos, que la que es Reyna de los mios la in par Dulcinea del Toboo no coniente, que ningunos otros que los uyos me auaallen, y rindan, . . . (II, Gg7v, 12-15).		. . . Allá os avenid, señores. con vuestros deseos, que la que es reina de los míos, la sin par Dulcinea del Toboso, no consiente que ningunos otros que los suyos me avasallen y rindan.

10.
Y Sanon le dixo, aora eñor dõ Quixote, que tenemos nueua, que etâ desencantada la eñora Dulcinea, ale v. m. con eo, y agora que etamos tan a pique de er patores, para paar cantando la vida como vnos Principes, quiere vuea merced hazere ermitaño? (II, Mm5v, 15-20).		. . . Y Sansón le dijo:     —Agora, señor don Quijote, que tenemos nueva que está desencantada la señora Dulcinea, ¿sale vuesa merced con eso? Y, agora que estamos tan a pique de ser pastores, para pasar cantando la vida como unos príncipes, ¿quiere vuesa merced hacerse ermitaño?

     I have chosen the reading vuestra (merced) as the object of this study because it is an excellent example of the many complex but (at least partially) solvable typographic, grammatical, and stylistic puzzles that we find constantly in the first editions of Cervantes' works, and above all, because it is one of the few instances in which a substantial number of non-Cervantine readings (usted[es], asté) have to be introduced into the text. There are also several other words which never occur in their modern spellings in the first editions of Cervantes' works (the reading voces, for instance, whose standard spellings were vozes or voçes), but their modernization would not represent any radical departure from the original form, nor are the numbers of occurrences considerable. As far as changes relating to vuestra merced are concerned, even though some of those shown in the ten excerpts quoted above might seem at first glance too drastic, it should be kept in mind that they are exceptional. Moreover, excerpts 1, 2, 6, and 9 are verbally the same in the modernized edition all in the first editions. In excerpts 8 and 10, 1 have merely developed the abbreviations “V.” and “v.” to the improper written form vuesa, which in all probability is the form Cervantes used in his manuscript (see “vuea Grandeza” in excerpt 8, and “vuea merced” in excerpt 10). And in excerpts 3, 4, and 7, I have replaced only one construction in each quotation (“vuetra merced” - usted, “boace” —> asté, “u merced” —> ella). Only in excerpt 5 were extensive changes necessary, replacing three abbreviations (“V. m.”) and one occurrence of the vuestra form with two readings of usted and two vuesa's. When Don Quixote addresses the priest he does not use any archaism, hence the usted form, but in his answer the priest mocks the affected speech that the Knight uses at times, requiring the old-fashioned title in its vuesa form (a similar thing happens in excerpt 10). Furthermore, although the readings usted(es) and asté cannot be found in the first editions of Cervantes' works, both words were probably already in use during

Cervantes' lifetime, though apparently they had not as yet found their way into print (for usted see Pla Cárceles, pp. 265 ff.). So in effect these changes do not represent a break with the language of the period, and the text has been effectively modernized without sacrificing the archaic speech of Don Quixote (and of those who imitate him) or the incorrect written forms used by non-Castilian characters and peasants. It should be emphasized also that, with the exception of boacé, all the other relevant Cervantine spellings have been retained.
     These changes are not only necessary for and consistent with out-and-out modernization, they also stress and make evident to present-day readers some of the linguistic nuances of Cervantes' style, which had been lost through time and the ill-advised standardization carried out by the compositors of the first editions of Parts I and II of Don Quixote. Despite the radical departure from other twentieth-century editions, however, these changes do not destroy or falsify the tone or sense of the text, especially because the original verb forms already agree with the non-Cervantine readings used (“vuetra merced dize” —> usted dice, excerpt 3). And if we take account of the fact that most editors of Don Quixote have silently introduced three hundred and six “incorrect” readings by resolving all V. m. abbreviations as vuestra(s) merced(es) when it is now clear that the appropriate form used in dialogues was vuessa(s) merced(es), the number of changes that should be made in a modernized edition is not unreasonable.
     What is true of the reading vuesa and its related forms is also true of the scores of other differing spellings which have been customarily modernized or regularized without warning readers of what has been done with the text and without looking for textual evidence that might support these changes. Countless words have lost their original pronunciation and typographical appearance without the editors having pointed out beforehand to their readers these changes. Such changes run into the thousands, for example, “nuue” to nube, “ay” or “aí” to either allí or ahí, “truxo” to trajo, “età” to está, “veya” to veía, “ombro” to hombro, “dixo” to dijo, “aldeguela” to aldegüela, “vno” to uno, “Sol” to sol, “cõ” to con, “çurron” to zurrón, “vuo” to hubo, “este” to éste, “della” to de ella, “Dulzinea” to Dulcinea, “acauallo” to a caballo, “paar” to pasar. In a scholarly, modernized edition these types of changes should be dealt with in the general introduction (see below).
     Major editorial or bibliographical problems should be discussed by the editor in special sections of his introduction, because it would be impractical to enter indispensable but necessarily lengthy editorial

explanations (or the hundreds of spelling variants) in footnotes. Moreover, this information, essential to any scholarly edition of Cervantes' works, obviously does not belong in footnotes, because its fundamental importance would not be fully appreciated, nor understood, if it is given piecemeal. I have shown here with my observations concerning the reading vuestra, and in some of my previous monographs, that even though orthographic variants, typographical evidence, and textual content are all interrelated aspects of the printed book, they yield their total information only when each of them is studied separately and when the various elements that make up each aspect can be seen together and analyzed as a whole. It serves no practical or scholarly purpose to isolate from each other the many pieces of the different puzzles (orthographic, creative, typographical, etc.) and mix them in with the pieces of the other puzzles in what must necessarily be incomplete, repetitious, and uninformative footnotes entered at random, that is, as the text imposes them, rather than following a logical order determined by the subject matter. For these same reasons, sixteenth- and seventeenth-century conventions regarding punctuation and division (or lack of division) into paragraphs should be studied together in the general introduction, and corrections of obvious typographical errors should be listed separately (following the order of the text) by editors as sections of their introductions. There should be no footnotes of any editorial, textual, bibliographical, or typographical character within the pages of the text proper.
     To conclude this section on footnotes, I would like to emphasize that there will always be a need for explanatory footnotes to clarify the text. This sort of footnote does belong within the pages of the text, but because there are several recent annotated editions of the separate works of Cervantes now on the market (and more are forthcoming), which fulfill these needs splendidly, a scholarly, modernized edition should not include explanatory or interpretative footnotes. In short, such an edition would present readers with a thoroughly modernized, scholarly edition of Cervantes' text, giving specialists all the essential information they might need concerning editorial, textual, typographical, and orthographic matters, but without footnotes.¹⁰

     ¹⁰ There is little new or revolutionary about some of the general editorial policies I am proposing in this article. Scholars have given a great deal of thought to the many problems involved in editing. For a [p. 85] comprehensive survey of the controversies concerning editing which have arisen among specialists sea G. T. Tanselle, “Recent Editorial Discussion and the Central Questions of Editing,” Studies in Bibliography, 34 (1981), 23-65. Tanselle's definition of the “historical” and “ahistorical” approaches and of “scholarly,” “unscholarly,” “critical,” and “noncritical” editions are somewhat narrow. He states that in a critical edition one uses “editorial judgment to determine when, and whether, emendations are to be made in the text,” whereas in a noncritical edition one reproduces “exactly one particular text, without alteration” (p. 60). Out to reproduce a text without alterations is almost impossible. In fact, a true “noncritical” edition would have to be a one-hundred-per-cent faithful facsimile reproduction of the [p. 86] only extant manuscript or printed copy of any given work. In other words, the editor of a noncritical edition would have to have no alternative as to his copy-text, because any decision involving a choice between different manuscript copies, between manuscript or printed copies, between different copies of the same edition, or between copies of different editions already constitutes an editorial judgment, as Tanselle acknowledges on page 62 of his article. “Regularizing and modernizing,” Tanselle states, “(their aims may be different, but they amount to the same thing) are ahistorical in orientation and therefore have no place in the historical approach to texts —which is to say, in scholarly editions” (p. 61). First, regularization and modernization may in some instances “amount to the same thing,” but this is not necessarily true in all cases, and, second, an editor may, and can, take the “historical approach” and, at the same time, produce a modernized edition (see above, footnote 1). Tanselle concludes that “a text prepared for scholars will also be the appropriate one to present to students and to the general public” (p. 61). Indeed, but leaving aside the fact that each work and the problems it presents to its editors are unique, I would go a step further: a text prepared by scholars for the general public, i.e. a scholarly, critical, modernized edition, will also be (in most circumstances) an appropriate text to present to scholars and to students.

     In theory, then, a scholarly, modernized edition of Cervantes' works should consist of an introductory volume (which would in fact be the last volume of the series to appear) and six volumes containing Cervantes' texts: I. Introducción a la edición modernizada de las obras completas de Cervantes, II. La Galatea, III. Don Quijote de la Mancha, IV. Novelas ejemplares, V. Viaje del Parnaso y poesías sueltas, VI. Comedias y entremeses, VII. Los trabajos de Persiles y Sigismunda.
     The introductory volume would then consist of:

1. An introduction setting out the general editorial policies.
2. A separate introduction to each of the six volumes of text, consisting of:
   1. a bibliographical or philological introduction to the work(s) in the volume.
   2. a short bibliography of the most important editions of the work(s).
   3. a select bibliography of studies of the work(s).
   4. the rationale of important editorial changes (this section may be part of item a).
   5. a list of typographical errors in the original and the correct readings.

     Summing up, we now know that the compositors who met the first editions of Don Quixote and of the other works of Cervantes altered the orthography of their manuscript copies in accordance with general conventions or personal preferences, and with idiosyncratic and irregular variations from the norm. In the process, among many inconsequential changes, they destroyed some of the grammatical, linguistic, and stylistic characteristics of Cervantes' writings. it is also clear that this mayhap well meant, but remorseless distortion of the authorial orthography needs to be reversed. These facts entail a complete reassessment of previous editorial practices.

For future editors to accept the texts of the first editions uncritically would mean giving undue authority to the various and differing orthographies and to the typographical vagaries of the compositors who set these works. To regularize the texts without first having a clear knowledge of exactly what happened during the setting and printing of these works would be to compound compositorial inconsistencies and preferences with editorial complacency, and, in effect, to obliterate the meager but tell-tale evidence of Cervantes' orthography which escaped the misplaced attention of the compositors. To avoid these pitfalls, future editors should part from tradition. First, they should recover Cervantes' orthography where possible, correct the authorial and compositorial errors and inconsistencies we have in the first editions, and prepare “ideal copies” of his works. Then, they should produce either old-spelling, regularized, or thoroughly updated editions. These solutions, especially the old-spelling and modernized edition, represent complete breaks with well-established but dated editorial policies, which, though acceptable in previous editions, can no longer satisfy present-day scholarly standards.
     In a scholarly critical, modernized edition, sixteenth- and seventeenth-century spellings, punctuation, and typographical conventions should be modernized. Abbreviations should be resolved throughout. Changes in substantive readings and lists of typographical errors, and of major editorial changes should be given in the introductory

volume, to avoid crowding the text with necessarily lengthy apparatus. And because of the availability of some excellent annotated editions which have appeared in the last few years, what is now needed is a scholarly edition which will give readers Cervantes' texts without any sort of footnotes. It must be emphasized, however, that all these departures from previous editorial policies would be accomplished without in the least sacrificing the stylistic and linguistic characteristics of the original or withholding any relevant editorial or bibliographical information from the reader. True, such an edition would be, in several respects, unlike any other edition published previously; but it has not been conceived as a substitute for recent annotated editions, nor, of course, as a definitive edition of Cervantes' works, which in this case is a contradiction in terms. No edition will ever serve all the needs of specialists, who will always have to consult more than one edition and go back time and again to the first editions, regardless of how many modernized, regularized, and old-spelling editions appear in the future. The sort of edition I am propounding here would not be an annotated edition. It would be, however, a more rigorous and more scholarly edition aimed at the general public than any other such edition published to date. Its goals would be to bring back and make evident to the non-specialist the various linguistic levels so masterly handled by Cervantes.
     Cervantes was the greatest Spanish writer of all time, but he was not perfect. The compositors who set the first editions of his works had their own spelling and setting preferences and imposed them in their stints. The task of future editors of Cervantes is, in my opinion, to correct and improve both the presentation and orthography of his texts rather than perpetuate and give authority to questionable readings and obvious inconsistencies and errors, or to attempt to justify them with circuitous, tortuous, or implausible exegeses. Every editor must have respect for the work being edited: for this respect to be extended to minor authorial lapses and contradictions, compositorial errors, inconsistencies, and orthographic preferences when they are demonstrably incorrect and detract from the work is excessive.