University of Virginia Library

Search this document 


expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
collapse section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 
expand section 

expand section 



Each of these four works presents editors with completely different job patterns, and this distribution of labour amongst the various compositors produced texts which retain little of Cervantes's orthography, reflecting instead a complex web made up with the differing spelling and punctuation preferences and setting habits of the compositors who set them. See my monographs The Compositors of the First and Second Madrid Editions of "Don Quixote", Part I, The Modern Humanities Research Association, London, 1975, "El caso del epígrafe desaparecido: Capítulo 43 de la edición príncipe del Quijote", Nueva Revista de Filología Hispánica, 28 (1980), 352-360, "The Compositors of the First Edition of Don Quixote, Part II", Journal of Hispanic Philology, 6 (1981), 3-44, and "The Need for a Scholarly, Modernized Edition of Cervantes' Works", Cervantes, 2 (1982), 69-87. I am now at work on three articles dealing with the setting and printing of the first editions of Galatea (Juan Gracián, Alcala, 1585), Viaje del Parnaso (widow of Alonso Martín, Madrid, 1614), and Persiles y Sigismunda. I wish to thank Professors J. B. Avalle-Arce and L. A. Murillo for their friendship and for the continuous support they have given me for the past several years.


My reference copy is copy G. 10181, The British Library. Formula: 4° in eights, ¶4 ¶¶8 A-Ll8 Mm2, fols. 1-12, 1-274 (572 pp.)


When the compositors of Don Quixote found the form vuesa merced in their copy texts they set it either as vuessa merced (two hundred and eighteen occurrences) or vuestra merced (five hundred and fifty-three occurrences), or used an abbreviation (three hundred and seventy-six occurrences), each compositor having his own preferred form (these numbers of occurrences include plural forms and a few occurrences of related forms of the word vuestra which are not an element of the form of address vuestra merced); see my article "The Need for a Scholarly, Modernized Edition", pp. 72-77.


It is no coincidence, of course, that one third (forty-three) of the total number of occurrences of the form of address vuesa merced appear in the short story Rinconete y Cortadillo, where the principal characters are two pícaro-like urchins, who are surrounded by tramps, beggars, corrupt bumbailiffs, prostitutes, pimps, bullies, and murderers, none of whom should have been addressed as vuesa merced. Cervantes's humorous darts aimed at his countrymen's all-too-frequent abuse and misuse of this formal form of address are glaringly evident in this short-story.


The Cuesta compositors set mainly spaces before punctuation marks, but some of them set also quads. Here I use the term "space" to refer to the blank spaces left on the printed page, and the term "quad" to designate the pieces of type, regardless of width. When one plots the data entered in Table 3 according to the numbers of pages having the same number of quads set before punctuation marks per page, the resulting mean line is what I call a compositorial quadmark; see Diagram 1. In my article "The Compositors of the First Edition of Don Quixote, Part II" I showed that each one of the compositors who set copy for the first editions of Parts I and II of Don Quixote had his own characteristic quadmark, and I was able to corroborate the boundaries of their stints with the help of this device.


Most pages of the first edition of Novelas ejemplares have thirty-four lines of text, but some pages are irregular in this respect and have either thirty-one (A2v), thirty-two (P6), thirty-three (¶¶1, A5, A5v, E4-E5v, G1, G2, G2v, G5-G6, K3, K3v, L2, O5, O5v, P2, S1, S1v, V5, V5v), or thirty-five lines (C4). Thus, some pages have more or fewer quads set before punctuation marks than others simply because they have more or fewer lines of text than the average page. These possible differences in the numbers of quads set before punctuation marks cannot be estimated, but, in order to avoid further discrepancies, in Diagram 1, I considered only full pages of type irrespective of the total number of lines per page, and pages with paragraph divisions which amount to less than two full lines of blank spaces per page. The pages excluded are: ¶1-¶4v, ¶¶1v, ¶¶2, ¶¶3, ¶¶3v, ¶¶5-¶¶8v, A1, A2, A2v, A3v-A5v, A6v-A7v, A8v-B1v, C1-C2, D5v-D6v, E6, G5, I1v, I2, L2, L2v, L4, L4v, L6v, L7, O7, P6v, Q5v, Q6, S1v, T7v-T8v, V6, V6v, X5, Y3v-Y6v, Y8, Y8v, Aa5v, Dd3v, Dd4, Ff8v, Gg1, Gg6v, Gg8v, Kk5v, Mm2, and Mm2v.


For example, in Rinconete y Cortadillo the uneducated pimp Repolido says: "no fe toquen estorias paffadas" (L2v, lines 3 and 4; my italics here and below), appositely using the 'incorrect" form estoria, but his mistress the prostitute Juliana la Cariharta, incongruently uses the "correct", and phonetically different form historia, "de la qual verdadera historia fon buenos teftigos eftos cardenales" (K7, lines 21 to 23). Cervantes, with his finely tuned ear to this sort of linguistic nuance, in all probability used the form estoria in both excerpts, because at no other time in the story does he make any differentiation between the vocabulary of Repolido and that of his mistress, as he pointedly and repeatedly does when he wants to emphasise the differences between the vocabulary of Rinconete, who was able to read, and those of most of the other characters of the story, who were illiterate.


What firm conclusions could we draw, for instance, from the spelling variant camuça/camuza when we have only one occurrence of each form (pages I2v, line 2, and Ii6, line 5, respectively), except that the word gamuza (soft leather made from the skin of chamois, sheep, deer, etc.) could be spelled in at least two different ways?


One example of this tendency to switch spellings in midstream is the clear-cut change that compositor E made in the first edition of Don Quixote, Part I from the form sospirar to suspirar (sospirar, gatherings I [one occurrence], M[1], N[1], and Q[1]; suspirar, gatherings R[2], V[7], Bb[1], Kk[1], and Nn[1]). In this instance the dominant compositorial spelling is suspirar, not only because there are more occurrences of this form (12) than of the form sospirar (4), but also because the compositorial spelling appears in the last gatherings set by compositor E, when he was already used to Cervantes's handwriting and it was easier for him to impose his own spellings over and above those of his printer's copy.


The spelling Precetor is in fact a double variant (capitalization and spelling, Precetor/preceptor) which could not have been incurred accidentally; hence I have included it in Table 4 even though it shows only one occurrence.


The quadmarks of some compositors change according to whether the compositors were setting seriatim or from cast-off copy (see "The Compositors", Diagrams 5 to 8).


8(pages) x 34(lines per page) + 1(extra line in page C4) = 273(lines). 273(lines) x 4(millimetres added to each line) = 1092(millimetres added). 1092(millimetres) ÷ 98 (length in millimetres of an average full line of type) = 11(average lines). 11(average lines) + 1(extra line in page C4) = 12(extra lines of type absorbed).


I am considering here and throughout the remaining pages of this study only the following contractions: ã for "am" or "an", etilde for "em" or "en", õ for "om" or "on", ũ for "um" or "un", ð for "de", and qtilde for "que".


The main typeface used was St Augustin Roman of Garamond, 95 x 1.9:2.9. Occasionally it is difficult to tell exactly which sort or sorts were scarce at a given time, especially because the same contractions might reflect different shortages at different times (an "ã" could be used to make up for shortages of a's, m's, or n's), but compositor K and apprentice Y usually made up for shortages of a's by setting the preposition a with accentuated types (á, à, or â). Hence, we can often point out the nature of the shortages without entering into exhaustive studies of the types. For example: page N4 has only four contractions and two accentuated prepositions (out of a total of eleven occurrences of the preposition a); pages N4v and N5 have, in contrast, one hundred and thirty-three contractions and thirty-one accentuated prepositions (out of a total of thirty-seven occurrences of the preposition a); and page N5v has only two contractions but ten accentuated prepositions (out of a total of fourteen occurrences of the preposition a). It is crystal clear that the compositor had sufficient types of all sorts when he set page N4, that he was short of all the most frequently used sorts when he set pages N4v and N5, and that he was short only of a's when he set page N5v.


One shortage of type could not be solved by using contractions. When compositor K began setting page N2v he ran out of upper-case Y's because he had set a considerable number of occurrences of the word Yſabela in the previous pages. To make up for this shortage compositor K borrowed about twenty upper-case Y's from another case (the extraneous Y's have a longer body than the Y's of the main fount used). Compositor K used ten of the borrowed Y's in page N2v, but just as he was finishing this page his case was replenished with type (the last seven lines of page N2v have no contractions and the last upper-case Y used in this page [line 34] is not one of the borrowed types). (The clean type came from the inner forme of the outer sheet of gathering M—M1v, M2, M7v, and M8—the broken E that appears on pages M8, line 14, and N4v, line 32, is the same type). When the clean type was emptied into the case the compositor's own Y's fell on top of the borrowed types he had not yet used. Apprentice Y (page N3, thirty nine quads set before punctuation marks) and compositor K set their own Y's in pages N3 to N6, but as page N6v was being finished they began using the borrowed Y's that had been left at the bottom of the box, mixing them with the original Y's (one long Y appears on page N6v, line 33; three on page N7v, lines 25, 29, and 34; five on page N8, lines 8, 21, 26, 32, and 34; etc. None of the long Y's used in pages N6v, N7v, and N8 could be any of the types used in page N2v because this page belongs to the last forme of gathering N—N1, N2v, N7, and N8v). The fact that compositor K ended with a mixed box of upper-case Y's is evident throughout the remaining gatherings of the book (including gatherings ¶ and ¶¶) where examples of both sorts of Y's appear side by side.


I refer only to compositor K here and below to simplify my explanation, but it should be kept in mind that he and apprentice Y were working together throughout the setting of this work.


In the following list the numbers of abbreviations of vuesa merced and other related forms and of contractions appear within parentheses: B7v(1, 10), D2(2, 61), E1(1, 33), E2(1, 38), H5(1, 58), I2v(3, 61), I3(10, 39), I3v(4, 14), I4(1, 4), I4v(1, 12), I5v(1, 8), I6(1, 6), I6v(1,6), I7(1, 5), I8v(1, 6), K2v(2, 14), L3(1, 21), L5(1, 6), M2v(1, 38), M3(7, 57), M3v(3, 6), N1(1, 42), N2(1, 19), N5(1, 95), R5(1, 7), V2(1, 10; the abbreviation and the contractions appear on the first eleven lines of the page, and it is worth noting also that page V1 has twenty-four contractions and page V1v has eighteen contractions), V3(1, 45), X1v(1, 32), Z1(1, 42), Z5(4, 62), Aa2(1, 41), Aa2v(4, 16), Bb3(2, 50), Cc1v(2, 50), Dd1(3, 50), Dd2v(3, 93), Ee5(2, 41), Ff1v(2, 72), Ff2(4, 57), Gg1v(4, 71), Gg2(1, 48), Gg2v(1, 36), Mm1(1, 16), and Mm2(1, 15). The abbreviations for the word nuestra (nrtildea.) and related forms show a similar distribution. Out of twenty-seven abbreviations only one appears on a page which has fewer than fifteen contractions (title-page). These facts are watertight proofs that at times the compositors were indeed short of some sorts of type, and that they used abbreviations and contractions to offset these temporary shortages.


This theory explains why some pages set seriatim have nearly as many contractions as do other pages set from cast-off copy (cf., for instance, pages C2v and C7 [seriatim] with pages C3 and C5v [from cast-off copy]; see Table 7). Also, if the variations on the numbers of contractions or abbreviations were the result of the compositors' not being able to fit their copy into their stints the increases in the numbers of contractions and abbreviations would under normal circumstances appear at the end rather than at the beginning or in the middle of gatherings (see, however, pp. 14 to 16 of The Compositors). An average page of the first edition of Novelas ejemplares has between zero and twelve contractions. Page ¶3 is the only page of gatherings ¶ and ¶¶ that has an irregular number of contractions (twenty-one), but this anomaly did not arise from any shortage of type (the compositors used four different typefaces in these gatherings: St Augustin Roman, St Augustin Italic, Parangon Roman, and Parangon Italic), but rather from the compositor's trying to fit three short official licences in one page. He was not able to do so, however, and ended by imposing the last four lines (less than three full lines of type) of the third licence in the next page (¶3v).


Running title m appears in similar conditions of wear on pages Ee4 (second printing) and Mm2. Running title p shows more damage on page Ee5v (second printing) than on page Ll2v; cf. especially the "p" and last "e" of exemplares. The swash uppercase "N" used in the running title of page Ee4v (second printing) is the same type used in running title n, but the running titles are not the same (cf. the types of the word exemplares). The "N" appears more damaged on page Ee4v (second printing) than on page Ll5v. The running title used on page Ee5 (second printing) is none of the original running titles.


The main reason why I consider that the number of sheets damaged must have been considerable is the fact that of the three copies I have examined, out of the few extant copies we have, two of them have the pages of the second printing (copy C.59.b.20, The British Library, and the copy of The Hispanic Society of America).


The changes are: one hundred and ten expansions of contractions, forty-two substitutions of one sort of accentuated type for another, eighty-three variations on line endings, thirty-six deletions or insertions of accents (five added, ten taken out), quads before punctuation marks (five added, nine taken out), and hyphens at the end of lines (five added, two taken out), four substitutions of "ſs" by "ſſ" (always -ſsetilde to -ſſen), two developments of abbreviations, three changes in punctuation marks (nõbre.>nõbre:—Ee5, 25 and Ee5, 26—quiſiere,>quiſiere—Ee5v, 19—and gente>gente,—Ee5v, 23), five corrections of textual errors (conocido,>conocido., Ee4, 24; Hiziet[on]lo>Hizier[on]lo, Ee4v, 17; Lorençò>Lorenço, Ee4v, 24; hermano>hermana, Ee5v, 2; and apatentes>aparentes, Ee5v, 10), introduction of two textual errors (pero qtilde no>pero no, Ee4v, 21; and lo qtilde quiere>lo quiere, Ee4v, 33), four spelling variants (por que>porque, Ee4, 19; ſi no>ſino, Ee4v, 1 and Ee5, 15; and Ygleſia>Igleſia, Ee5, 10), and five contractions.


The gatherings of the alphabetical sequence were set and printed between 14 July 1613 (date of the dedication) and 7 August 1613 (date of the errata). The book was probably on the market by mid August. Cervantes applied for the official licences for Aragon before the work was sent to press (the imprimatur states: "se deue imprimir"; and the copyright reads: "le desseays imprimir"), but the licences were granted when the book was about to be finished (imprimatur, 31 July 1613; copyright, 9 August 1613).