Early Modern Spain
[Close Print-Friendly Version]
It is a commonplace of literary studies that problems of text and interpretation are really the same thing. Is Hamlet's 'too, too solid flesh' really 'sullied'? Is Cressida 'the Trojans' trumpet' or 'the Trojan strumpet'? While the poetry of Garci Sánchez de Badajoz is hardly in Shakespeare's league, it nevertheless poses some interesting challenges both to the editor and the critic. In common with many poets of the late fifteenth and early sixteenth centuries, Garci Sánchez's poetry has not been transmitted in the form of neat compilations of 'collected works', but has to be assembled from numerous relatively scattered printed and manuscript cancioneros. 2 The nature of these sources creates complex problems for the production of a critical edition, and the purpose of this study is to re-examine one of the sources in particular, British Library Add. MS 10431 (Dutton LB1), 3 in order to reassess its authority as a source for Garci Sánchez, and to use it to draw some wider methodological and interpretative conclusions.
For a writer who was famed as a lunatic lover and spent at least part of his life as a court poet, Garci Sánchez's extant corpus of poems is fairly small. Altogether, some 66 poems have survived in 30 sources up to 1520 (Table 1). The headline points are fairly obvious: the sources which have the greatest number of poems by Garci Sánchez are the two printed collections 11CG and 14CG (the first and second editions of the Cancionero general, compiled by Hernando del Castillo and printed in Valencia in 1511 and 1514), and the British Library MS now known as Dutton LB1. 4 Understanding the relationship between these three early sources is therefore central to determining their relative authority as sources for Garci Sánchez and for many other poets represented in them.
[See Table 2 in pdf file (22KB)]
British Library Add. MS 10431, formerly known as the British Museum cancionero, is an important collection of some 470 pieces by 112 named poets from the reigns of John II, Henry IV and the Catholic Monarchs. 5 It is a quarto volume currently consisting of 120 folios written in a single hand, mostly in double columns. A second hand has added a small number of comments and annotations. The MS may once have been longer, since the rubric of fol 1r alludes to works by 'el famoso poeta' Pedro de Herriega, which do not appear in the volume. 6 The date of the MS is uncertain. Hugo Rennert estimated a date between the 1470s and 1500, 7 Carolina Michaëlis de Vasconcelos preferred 1500-1520, 8 Dutton estimates 'hacia 1500' (I,131), J. González Cuenca thought that it predated 11CG, 9 while Patrick Gallagher placed it between 11CG and 14CG (3- 5).
We have no record of ownership until the MS was sold by Sotheby's on 17 February 1836 from the library of Richard Heber to the British Museum for the sum of one shilling. Two pieces of evidence led R.O. Jones to conclude that the collection could have been the personal anthology of Juan del Encina: a marked similarity between certain aspects of the orthography of the MS and that of the 1496 edition of Encina's Cancionero, above all the use of intervocalic v rather than u; and the presence at the end of the MS of a group of poems attributed to the 'actor deste libro' among which there are two versions of poems from Encina's printed collection together with at least one other which is very reminiscent of Encina. Jones argued that the 'actor' ('compiler'?) was Encina himself. 10
Carlos Alvar has subsequently produced a helpful survey of the relationships between LB1 and several other MS and printed collections, and has been unable to escape the conclusion that LB1, 11CG and 14CG are closely related, both textually and chronologically: LB1 and 11CG have 202 poems in common; 17 of those dropped from 11CG are in common with LB1; and 19 of those added to 14CG are in common with LB1. 'LB1 presenta composiciones que se encuentran sólo en 11CG, y que no volvieron a ser publicadas en las ediciones posteriores, y también presenta otras piezas que se incluyeron por primera vez en 14CG: parece que es lógico pensar que LB1 es posterior a ambos Cancioneros impresos.' 11
While it is impossible to disagree with this conclusion as far as it goes, there is a case for questioning the assumption (which Carlos Alvar does not make, but which might easily be drawn from his conclusion) that chronology and authority are necessarily causally related. Put simply: is the authority of LB1 any weaker because it appears to postdate 1514? Let us turn some of the earlier figures around: of the 1033 poems in 11CG, 831 (80.4%) are not in LB1; if the compiler of LB1 used 11CG, how is it that 19 poems are included which appeared for the first time in 1514? And if he used 14CG, how is it that he includes 17 poems which were dropped after the first edition? Furthermore, how is it that of 191 poems in common with 11CG, and 189 in common with 14CG, there is only one common sequence of four poems, one of three poems, and six pairs? And is it possible to account for the presence of unique poems in the midst of common groups without assuming a much more complex relationship than that which is implied by Carlos Alvar's analysis?
These questions are important because the relationship between cancionero sources on the one hand and editions of the work of particular poets on the other is fundamentally an assymetrical one, and the editorial practices involved are very different in each case. Almost always, cancionero sources are compilations or anthologies, whether of one or several writers. They are a record of an individual's or group's interests, tastes and collecting habits, compiled at a particular place and time. 12 By their very nature, then, the cancionero anthologies impose uniformity on variety. The process of collection and selection is normative in cultural and artistic terms, while the material processes of preparing the anthology for presentation, private enjoyment or publication result in a 'fair copy' which appears to be consistent in look and feel. If we were able to look beneath the visual and editorial uniformity at the surface, we would almost always find a ragbag of sources drawn from different places at different times, displaying conflicting orthgraphical and typographical conventions, and most important, sources of widely divergent textual authority.
None of this is very different from the normal process of preparing any work for publication; behind the smoothly printed surface lies an agony of recensions, drafts, re-drafts, last-minute corrections and second thoughts at final proof stage. What makes this observation important in the case of cancionero sources is the warning it gives us that no statement about the authority of a particular text in a cancionero anthology can safely be applied to any other in that collection. The texts of Palgrave's Golden Treasury, for example, are just as likely to derive from an autograph presented to the compiler by the poet himself as from a unreliable printed source which perpetuates mis-readings from the first edition. The typography of the finished collection, and Palgrave's own editorial conventions as well as those of the publisher, would obscure those discrepancies from view.
The implications of this for editorial practice may be summarised in Figures 1 and 2. Figure 1 is a simple representation of what normally happens when one edits a text. There are one or more extant witnesses whose relationship has to be established before a series of procedures can be applied to the texts of the witnesses to attempt to establish the lost archetype. What we do when evaluating cancionero sources is almost exactly the reverse (Figure 2). We usually have one witness, which we know, suspect or have reason to believe has been compiled from several sources, all of which are lost. In order fully to evaluate the authority of the witness, we have to try to reconstruct all of the missing sources, or at least their relationship to each other and to the witness we have. A less conventional form of stemma is required to express these relationships, and an example is given in Figure 3.
[See Figures 1 (106KB), 2 (22KB) and 3 (35KB)]
How can we get behind the blank wall of uniformity presented by the cancionero text? A great deal of evidence about the process of compilation can be gleaned from a close examination of the physical characteristics of the MS. LB1 now consists of 120 leaves, written throughout in a single legible but non-professional hand, and in a single ink. The MS originally consisted of 124 folios, as is indicated by the original foliation, of which numbers 13 to 16 inclusive have been removed. They almost certainly contained the Liciones de Job of Garci Sánchez de Badajoz ('Pues amor quiere que muera', Dutton 1769), since the last poem on the verso of folio 12 is the beginning of a dedication of the Liciones to a friend. The poem was presumably removed in deference to a change of spiritual climate in the 1520s. This poem is also largely missing from a MS now preserved in the University of Salamanca (Dutton SA10b-95), from which four folios have also been removed, except that in this case the censor omitted to remove the first 17 lines. The opening rubric of LB1 ('Aqui comiençan las obras de garci sanches de vadajos con otras obras de algunos syngulares poetas') and the concluding 'finnis deo gratias' suggest that, apart from the four missing leaves, the MS is substantially complete.
The paper on which LB1 is written is unfortunately unwatermarked, but an irregularity in the spacing of the chain lines makes it clear that it was all made from the same pair of moulds and that presumably it was bought as a batch for this particular purpose. There is also evidence to suggest that the paper was made into a book before the MS was written. The original format is difficult to establish, since when the British Museum acquired the MS it was rebound and the leaves supplied with new hinges and quired in 15 lots of eight. It is unlikely that it will ever be possible to establish the original format with certainty, which is a pity, since quiring in 4s would suggest a ready-made book, whereas quiring in 2s need not. But it is noticeable that the final four leaves are significantly less closely written than the rest of the book and point to the copyist's possibly having some difficulty in filling his prescribed length, and might also account for the addition of the poems attributed to the mysterious 'actor'. Physical evidence, then, points to, but does not prove, that the MS was written 'at a sitting', into a loose or possibly bound batch of paper whose length had probably been estimated to fit the job in hand, and that it was written with the object of drawing together a disparate collection of material between two covers. The MS does not suggest a collection in the process of being formed or added to gradually over a long period of time.
More physical evidence of a slightly different sort is given by what we can deduce from the copyist's working habits. The MS is written for the most part in double columns: the only exceptions are some poems in arte mayor where the format will only allow single columns, and a couple of pages written in three columns where the verse is basically hexasyllabic. There are two other exceptions which will be mentioned later. For the most part, the copyist seems anxious to fill his page as tightly as possible, and throughout the MS one has the impression of never a wasted space. But from time to time the copyist has punctuated his progress through the book by leaving small areas of blank space, usually at the end of a page, but sometimes in the middle, so as to divide, in effect, the collection into 34 sections. Sometimes the division is reinforced by a change in the rubric; or by writing the rubric across both columns; or by using the formula 'de x cancion' where the 'de x' means that all the following poems until the first attribution are by x; by the use of some formula such as 'comiençan las obras de x'; or by the use of what passes for a 'florid' capital (eg. fol 104r). The 34 sections of the MS are summarised in Table 2.
[See Table 2 in pdf file (22KB)]
Two alternatives are suggested by this unconsious ordering of the MS. Does it represent an attempt to classify the material, and if so what system of classification was used? Or do the divisions represent the points at which the copyist changed his exemplar? In the first alternative - the compiler or the copyist if they were not the same person - might have taken his shuffled collection of odds and ends and dealt them into piles corresponding roughly to a division by authorship. This is quite possible, since changes of section never come in the middle of a group attributed to one author, though some authors, for example Pinar and his sister Florencia, are common to more than one section. It is likely that the longest and most substantial sections were put together in this way, since within the section one can detect changes in the type of source, although these could well have been present in the immediately preceding exemplar.
On the other hand, some divisions correspond more obviously to particular sources, especially where a number of poets are grouped around a particular theme or subject or are related geographically or by period. Besides, the copyist sometimes indicates a change of author by a florid capital within a section [90v-91r], and some of the sections consist of a main author plus a fill-up, which would suggest that the source might have been a MS or printed pliego with the characteristic make-weight to fill out the extra space left by a longer work. In another section the familiar tone of the attributions, the preponderance of first names in the rubrics, suggest a group of domestic pieces originating in one place at one time and recorded by one of their number, or a secretary who knew them well and expected his readers to share that familiarity. Some sections are little more than scraps, and this impression becomes stronger in the final groups where the copyist seems hard pressed to fill the remaining pages [119v-120r] and where the attribution of all the pieces to the 'actor' is not entirely clear.
Although the realisation that the MS is organised in some not very obvious way represents a considerable step forward in our study of the collection, it is at the next stage that the major unknown quantities enter the argument. If some at least of the divisions do represent changes of source, it should be possible to be more precise about the nature and status of these sources, provided we can be reasonably certain that the copyist has given us an accurate record of what he saw. To that extent, then, are any differences discernible from section to section an accurate reflection of the state of the sources? To answer this question would involve demonstrating that the copyist was or was not an accurate worker, a task which is strictly speaking not possible unless one has the original from which he worked. But it should be possible to gain some indications without straying too far in the direction of guesswork.
We need to know as much as possible about the scribe's orthographical preferences, the relative strengths of his preferences for some forms rather than others, his general attitude towards his exemplar, his willingness to leave difficulties or irregularities unsolved or alternatively his readiness to intervene with his own suggestions and solutions: in short, some guide to the literalness of his mind. The first source of such information is where a text appears twice in the MS and where it can only have been copied from the same exemplar. For all practical purposes this means places where the copyist has corrected himself, such as Figure 4, where three lines of the poem 'O mi dios y giador [sic]' (Dutton 0706) were copied at the wrong point in the poem, have been crossed through and repeated at the correct position, 11 lines later. The differences in orthography between the two appearances of these three lines are trivial (que/que; alunbra/alunbra; y/e; virgo/virgo) while the similarities are telling (the long final s in 'nos'; the spelling of 'gia'; the contraction of 'lacrimarum'). Furthermore, the rhyme scheme of the poem indicates that in each instance there is a line missing.
| [Image: Figure 4 (fol 10v, column b). The first three lines have been cancelled and are repeated in their correct place in the following stanza. The rhyme scheme shows that in each case there is a line missing between ‘yn hac lacrimarum vale’ and ‘o clemens virgo maria’.] |
Unfortunately, such relatively long examples are very rare. More common is the opposite case where a poem is repeated but we can be fairly sure that two different exemplars are involved. Figure 5 shows two versions of the poem 'La vida seria perdella' (Núñez; Dutton 0843) which appear on fols 43v and 44r. The orthography is virtually identical until the last line, where the variant reading almost certainly indicates a different source. Note the contracted final e on 'muerte', which is a strong preference of this copyist. There are seven such repeated pairs in LB1 and they can be quite helpful in gaining an insight into the copyist's habits.
| [Image: fol 43v, column b] [Image: fol 43v, column a] | [Image: fol 44r, column b] | |
A further guide is provided by cases where a contrast in spellings, especially names, indicates a respect for both readings. Figure 6 illustrates a section of the text and commentary of Garci Sánchez's poem 'El dia infelix noturno' (Dutton 0731). Note how the lemma from the text preserves the spelling 'tis', while the commentary gives the more correct form 'Stix', and Figure 7 shows several examples of contrasting pairs of spellings from the same poem: 'menalanpo/melampo'; 'pamphago/panphago'; 'dorçeo/doçeo'. Note too the clue given by the cancelled 'fueron' just before the poetic text resumes. These are clear cases where spellings seem deliberately contrasted rather than being the result of variation or carelessness: if the copyist knows that 'tis' should be spelled 'Stix', then he could have changed it in the text of the lemma. All the lemmata in this commentary show the same kind of literal attitude to the text.
| [Image: Figure 6 (fol 15v, columns a and b). The text (column a, sixth line) and the lemma (column b, first line) both read ‘hasta que de tis laguna’, while the gloss reads ‘Stix’.] |
A further source of information about the nature of the sources is provided by those occasions where the copyist abandons his normal working pattern, such as Figure 8 ('Caminando por mis males'; Dutton 0693), where the double-column format is abandoned in favour of the common convention of writing the romance across the page. The mise en page makes it clear that he prepared for this change of format by making sure that the preceding poem ('No pido triste amador') finished in the top half of the page, so that he could set off the cantar which is included in the romance by reverting to double-column layout at the bottom of the page. On the verso of this leaf he has the rubric 'torna a romance' and resumes writing across the full width of the page. Only four poems are written in this way, all are romances and all are in the sections dedicated to Garci Sánchez de Badajoz, and they give a strong indication that the copyist was trying to preserve the format of his exemplar.
| [Image: Figure 8 (fol 7r, whole page). The copyist made sure that the poem ‘No pido triste amador’ finished in column b, so that the ensuing romance could be written across the page and the embedded cantar ‘Son en canpo desperança’ could be set off by reverting to two-column format.] |
I have already suggested that the abandonment of preferred forms, especially strongly preferred forms, and the choice between forms in free variation might also provide evidence of the original source. This approach is suggested by the implications of textual critics working in the field of the printed book, such as D.F. McKenzie's work on the composition of The Merchant of Venice. 13 McKenzie demonstrates that when Jaggard's compositor B abandons his preferred spelling he rarely does so when his copy coincides with his preference, and when that preference is very strong he never sets the alternative unless it is in the copy. Applied over a wide range of examples, this study supports our intuition that when a compositor has no preference he will tend to follow the copy, and this observation might be expected to apply even more strongly to a scribe than a compositor, since the scribe has fewer restrictions caused by justification and the constraints of his fount, and is less likely than the compositor to alter spelling to assist justification.
Now although it is interesting to know the scribe's preferences, (and in the case of LB1 we can say with some certainty that they are: hard gi- and ge-; final -e; final -es; and final long s, in descending order of strength) it is unlikely that these alone will tell us much about the state of the exemplar, which is really what we are interested in. But one feature which can tell us a good deal is the type and degree of contraction present in a text, since there are considerable differences in the use of contraction and ligatures by scribes and compositors. For the scribe, the primary function of the contraction is ease and speed of writing, though it could be and often was used to assist in fitting a text into a given space. For the compositor, the function of contraction is almost solely that of justifying a line of type to the measure, and the range of its use is constrained by the characteristics of the fount: he can only set the ligatures which he has in his type cases. Inevitably, then, a compositor's contractions can easily be distinguished from those of a scribe, which tend to show a greater variety and density, and where we find significant deviations from the copyist's norm in respect of contractions, it is reasonable to look for a different kind of exemplar. Figure 9 contrasts two columns of LB1. The left-hand column (fol 1va) has 42 contractions, the second (fol 10ra) has 22. Note also the different treatment of final -es/-es. The right-hand column is part of a series of four poems which share the same characteristics (see Table 3, below); the left-hand column is nearer the norm for this copyist.
| [Image: fol 1v, column a] | [Image: fol 10r, column a] |
The final test of a copyist's attitude to his exemplar is the surviving state of the texts themselves. At this point in the argument it is convenient to concentrate on the section of the MS devoted to the poems of Garci Sánchez de Badajoz, in order to bring together the two threads of the argument: that it is dangerous to generalise about the reliability of texts within an anthology; and that a close scrutiny of the physical features of the collection, especially the accidentals, can give some idea of the type of exemplar used and the extent to which the copyist accurately reproduced what he saw before him. This section of LB1 is also a good test because it raises in microcosm all of the larger questions mentioned earlier about the relationship of LB1 to the first and second editions of the Cancionero general.
[See Table 3 in pdf file (35KB)]
Table 3 shows the order of poems by Garci Sánchez in LB1, 11CG and 14CG. It should be noted that six poems added in 14CG do not appear in LB1 (seven, if the Liciones de Job, Dutton 1769, are included) and that the only significant consecutive grouping common to LB1 and 14CG is 17, (17a=9), 18, 19. These four poems are the same group of four which showed a markedly lower density of contraction than the norm for LB1 (Figure 9), and which might therefore be supposed to have derived from a printed source. 14 17a is an extra stanza which has been added to 17 in the MS, but which is in fact the last stanza of the poem which occurs between 17 and 18 in 14CG and which appears in full as no 9 in LB1. Since this rogue stanza is headed 'fin' it seems likely that a massive case of haplography has elided the two poems into one. As the evidence of contraction suggests a printed exemplar, it is probable that these four poems circulated in a printed pliego as a group and that they found their way independently into LB1 and 14CG. It is unlikely that they were taken from 14CG since poem no 17 and the rogue stanza do not appear on the same page of the Cancionero general until the 1517 edition, and in any case there is a major variant in the penultimate line.
Although the occurrence, or co-occurrence, of poetic texts in various collections can be used to establish textual and chronological relationships, there is a great deal more to be learned from detailed study of the variants between these texts. One of the reasons which led Patrick Gallagher to prefer CG readings to LB1 was the fact that, although neither tradition, in his view, had more authority, the LB1 tradition was 'badly damaged in this sole surviving copy by inaccurate and fragmentary transcription' (48). Carlos Alvar's subsequent conclusion that LB1 postdates 1514 could be said to cast further doubt on the reliability of the LB1 texts: if LB1 derives in some way from CG, why are so many of the texts so different from each other? But the implicaton that LB1 is a poor copy of CG is open to challenge when we come to consider differentially the degrees of variation between LB1 and CG from poem to poem. This raises an interesting question about rates of decay in the transmission of texts, and could provide the editor with another tool for investigating and elucidating the nature of a cancionero anthology.
It is commonly accepted that every act of copying introduces changes, some voluntary and some accidental. We might call this 'the law of textual entropy'. But textual entropy is not constant; not all acts of copying will introduce the same amount of change, so that there is no such thing as a textual 'half life', a constant rate at which texts will decay through successive copying. But it might be reasonable to expect a single copyist who copies several texts under similar circumstances to cause a much more constant rate of decay, even allowing for variations in mood, tiredness, hunger and other working conditions. If MS A shows a 10% divergence from its supposed exemplar B, whereas another MS C shows a 30% divergence from the same supposed exemplar B, we might question whether C did indeed derive from B, or at least whether it derived via the same route.
In the case of LB1 and CG, we find a range of divergence from poem to poem of between 10% and 50%, 15 and it might be reasonable to ask whether this wide range of divergence called into question whether all the MS texts derived from printed exemplars, or indeed whether any of them did: 10% of lines with variants is in any case a high figure, as the rate of variation between 11CG and 14CG, for example, which we know to be directly related, is less than 1%. The importance of this line of argument can be seen in Figure 10, where the 14CG text is longer than that of 11CG and where the rate of divergence between LB1 and 14CG is different in different parts of the poem:
| ll. 1-34, 49-94 | 43/79 lines different (54%) |
| ll. 35-48 | 8/14 lines different (57%) |
| ll. 95-129 | 10/35 lines different (28.5%) |
[See Figure 10 in pdf file (46KB)]
It would be reasonable to conclude from this evidence that it is very unlikely that whoever copied this poem into LB1 ever saw Castillo's text, since the only possible hypothesis - that he already had the short version and added the extra lines from Castillo, just as Castillo had simply added extra lines to the 11CG version - is belied by the markedly different rate of divergence between the two sections common to LB1 and 14CG. Since there are no variants between 11CG and 14CG in the sections they have in common, we might assume that, rather than reprint the new, longer text in its entirety, Castillo simply added the new lines at the appropriate points in the 1511 text. This could affect any conclusions which might be drawn from the above analysis of differential rates of variation.
We should be wary, in any case, of imposing modern ideas and expectations about textual transmission onto those of another period, especially when we have no clear idea of what Castillo meant by 'ordenado y corregido por la mejor manera y diligencia que pude', and when we note that he himself was aware of the lack of authority for some of his texts: 'no fue en mi mano aver todas las obras que aqui van de los verdaderos originales o de cierta relacion de los auctores que las hizieron'. Yet Castillo's finished product is (like Palgrave's?) visually very impressive and, by implication, authoritative. His texts seem to be the longest, the tidiest and metrically the most consistent, but are they hyper-correct by contrast with other sources?
Furthermore, it will often not be possible to locate the 'verdaderos originales' for the simple reason that the author never intended to produce a finished product, but rather an open-ended scheme, which like the music of the period and that of today, would take as many forms as it had performances. 16 Such a scheme is that of Garci Sánchez's Infierno de amores (Dutton 0662), which consists of a sequence of stanzas each telling of the sufferings of the most famous poets of the period, and ending with an injunction to any gentleman who feels excluded to write himself a stanza. As this poem exists in at least five sources it is possible to get an idea of the way in which it grew in response to Garci Sánchez's invitation. In Table 4, 13*BI is a rare early pliego suelto preserved in the Bibliothèque Nationale and assigned by Norton to Cromberger, Seville, 1511-1515, and SA10b is a late 15th-century anthology now in the University of Salamanca. Note that SA10b is complete in that, although it is the shortest, it has the stanza with the invitation; that the series A-H is unlikely to be a response to the publication of the invitation in 11CG since parts of it appear in two other sources; that the group 10,14,24,29 is absent from all versions except Castillo's; and the pairing 28+30 is common to all versions except CG, since Castillo makes the Don Sancho of stanza 30 the brother of Antonio de Velasco in stanza 29, whereas all the others make him the brother of Diego de Castilla. The common sequences between LB1 and 13*BI are very impressive and strongly indicate without recourse to variants that LB1 does not in this case derive from CG.
[See Table 4 in pdf file (38KB)]
Figure 11 shows just how cavalier the anthologist of 1500 could be towards his texts. The Figures contrasts four stanzas from Garci Sánchez's poem 'A la ora en que mi fe' (Dutton 0687) as recorded in LB1 and 11CG. Very few stanzas from the LB1 text of this poem retain anything like the correct prosodic structure of the poem, and yet there are very few places where the text does not make sense. The reason seems to be that this is an example of memorial transmission, since what has been lost is in almost all cases the structure of the stanza, particularly the second half of each. Most stanzas preserve something like the tight formal arrangement of the first six lines - abc, abc - but the second part with its two quebrados has been elided into a straight prose account of the sense. This is almost certainly because of the rarity of the form (Le Gentil notes only six examples) which seems to have been enough to defeat someone who had not seen it written down. It is most unlikely that the text could have been reduced to this condition had it been copied from a written, still less printed, source. Yet the loss of the prosodic structure does not in itself render the text useless, since there is nothing to prevent the compiler having captured the sense more accurately than Castillo's neater version; that kind of evaluation can only be made by looking at the variants themselves, and there are a great many to be taken into account. 17
[See Figure 11 in pdf file (20KB)]
It would be reasonable to draw a number of general and particular conclusions from this discussion:
Finally, Table 5 summarises my own best estimates of the likely sources of the Garci Sánchez poems in LB1.
[See Table 5 in pdf file (26KB)]
[See all tables in downloadable Excel file (77KB)]
| Footnotes | ||
|---|---|---|
|
||
| 1. | |
Earlier versions of this paper were given at a meeting of the Seminar in Editing and Textual Studies at King's College London on 17 February 1976, and at a two-day conference on 'Spanish Cancioneros: Materials and Methods' held at the Institute of Romance Studies and Queen Mary and Westfield College, 27-28 June 1997. |
| 2. | |
Patrick Gallagher, The Life and Works of Garci Sánchez de Badajoz (London: Tamesis Books), 1968. See also J. Castillo, Cancionero de Garci Sánchez de Badajoz (Madrid: Editora Nacional), 1980. A further poem attributed to Garci Sánchez was subsequently identified in Madrid, Nacional MS 5602 by N.F. Marino, 'An early attribution to Garci Sánchez de Badajoz', Anuario medieval 2, 132-40. |
| 3. | |
Brian Dutton, Catálogo-Indice de la poesía cancioneril del siglo XV, 2 vols in 1 (Madison: Hispanic Seminary of Medieval Studies), 1982; Brian Dutton and Jineen Krogstad, El cancionero del siglo XV, 7 vols (Salamanca: Universidad de Salamanca), 1990. LB1 appears on pp 131-275 of vol 1. |
| 4. | |
MN14 is a 19th-century MS in the Biblioteca Nacional Madrid (MS 3777) whose importance for Garci Sánchez studies has yet to be clearly established. |
| 5. | |
Dutton's numbering is from 1-470, but numbers 83 and 318 are not used, while three poems are numbered 81b, 319b and 380b. |
| 6. | |
Since we have no record of such a poet, and since the attribution has been added in a later hand, I am inclined to suspect festivity on the part of a previous owner of the MS. |
| 7. | |
H.A. Rennert, 'Der spanische Cancionero des Brit. Museums (Mss. Add. 10431)', Romanische Forschungen 10 (1895), 1-176. Rennert only printed the poems which were unique to this MS. |
| 8. | |
In a review of Rennert's edition in Literaturblatt für germanische und romanische Philologie XVIII-4 (1897), 127- 43, she pointed out that, from internal evidence, the MS cannot have been compiled in its present form before 1498. |
| 9. | |
'Cancioneros manuscritos del Pre-renacimiento', Revista de literatura (1978), 177-215. |
| 10. | |
R.O.Jones, 'Encina y el Cancionero del British Museum', Hispanófila 11 (1961), 1-21. |
| 11. | |
Carlos Alvar, 'LB1 y otros cancioneros castellanos' in Madeleine Tyssens (ed.), Lyrique romane médiévale: la tradition des chansonniers. Actes du Colloque de Liège, 1989. Bibliothèque de la Faculté de Philosophie et Lettres de l'Université de Liège, fascicule CCLVIII, (Paris: Droz), 1991, 469-97. The quotation is on page 495. The relationship between LB1 and CG is even closer than is apparent in Alvar's summary. 15 of the poems added to 14CG and common to LB1 were by one author, Garci Sánchez de Badajoz. In addition, at least four poems were lengthened in 14CG and each of these exists in LB1 in its longer version. There is also a striking similarity between LB1 and CG in terms of the poets omitted from both collections, especially those poets whose work was extensively published in the 1490s and might reasonably be expected to have been known to both Castillo and to the compiler of LB1: Mendoza, Montesinos and Encina. See Jane Whetnall, 'El Cancionero general de 1511: textos únicos y textos omitidos' in Juan Paredes (ed.), Medioevo y literatura. Actas del V Congreso de la Asociación Hispánica de Literatura Medieval. Granada 1993. 4 vols. Granada, 1995. IV, 505-15. |
| 12. | |
V. Beltrán Pepió, 'Tipología y génesis de los cancioneros', Cultura neolatina 55, 233-65. |
| 13. | |
Studies in Bibliography 12 (1959), 75-90. Harold Jenkins's magisterial analysis of the Q1, Q2 and Folio texts of Hamlet (London: Methuen), 1982, pp. 18-74 was equally helpful in this regard. |
| 14. | |
There are also three pairs of poems common to LB1 and 14CG. Note that poems 10 and 35 in LB1 (498 and 515 in 14CG) are each a single poem, with a slightly different treatment of self-quotation in each case. To that extent, they help to differentiate LB1 and 14CG even further. |
| 15. | |
In this case 'divergence' is measured in terms of the number of lines which contain a variant other than an orthographic variant. Another way of measuring divergence might be to compare the number of bytes which differ between texts. Neither is ideal, because variants, whether orthographic or substantive, can differ greatly in their significance. However, the method chosen may at least serve as an approximate indicator of the order of magnitude of difference which exists between two texts. |
| 16. | |
Simon Gaunt, 'Orality and writing: the text of the troubadour poem' in Simon Gaunt and Sarah Kay (eds.), The Troubadours. An Introduction (Cambridge: Cambridge University Press), 1999, 228-45. |
| 17. | |
In line 88, for example, LB1 appears to have the better reading. |
| 18. | |
Gaunt, 'Orality and writing', 229. |