From: Cervantes: Bulletin of the Cervantes Society of America 14.1 (1994): 108-09.
Copyright © 1994, The Cervantes Society of America

To the Editor:
     The trouble with what Professor Parr sees as a problem is, bluntly, that it has problems. I have nothing to say about his personal genre uncertainties: to be sure, if one were to read Don Quijote as, say, a poem in disguise, or a prose poem, or a translation from Sanskrit, varying readings would inevitably be produced. I read Don Quijote as a book, written by one Miguel de Cervantes, and take my cues —all my cues— from the text rather than from any critical stance. I too am a professor of literary studies, and have for my sins written my share of criticism and given the world the immense benefits of my share of scholarship. But a literary translator must, to the extent humanly possible, work from his text rather than from any and all preconceptions.
     Indeed, this is precisely what I have tried to do, and Professor Parr's chosen example, the dialogue between the Duchess and Sancho, II, 32, is I think an excellent example of his and my ineluctably differing approaches. I have a copy of Murillo, and have duly noted what is in that edition; I have a copy of Covarrubias and have carefully noted, too, what that invaluable source has to say. I have the O.E.D., and I used it to decipher Smollett's “bucking.” But I also have what Cervantes rather than any and all of these authorities wrote, and among other things the Cervantian text makes me ask myself, and ought to make Professor Parr at least ponder, is why on earth, if this passage is designed to exhibit, inter alia, the Duchess's “cruel humor,” Cervantes has her go on to say, as she does, “Mirad, maestresala . . . to que el buen Sancho pide, y cumplidle su voluntad al pie de la tetra”? Are we to imagine, if Professor Parr has correctly glossed these lines, that she is carefully instructing her maestresala to (in his words) “douse [Sancho] in the linen colander, if need be”? Would she be speaking of Sancho's “voluntad,”


108 BURTON RAFFEL Cervantes

much less of observing it “al pie,” if that were the case? Sancho —who after all was there, as neither Professor Parr nor I were— has just observed that “un lavatorio de éstos antes es gusto que trabajo” (emphasis added). If both the Duchess and Sancho were thinking in the terms Professor Parr stipulates, would “gusto” be an appropriate word? We do indeed know that Sancho knows the phrase “meter en la colada”: quite so. But we also know that he is distinctly lecherous; that young and nubile females are doing the washing; and we know that “innuendo” is precisely what, under similar circumstances in this book (whatever its genre) the Duchess has employed. And will, in this same book, employ again. Must we, then, in the face of such contextual evidence, insist on absolute literalness of lexicon? That is not Cervantes's way. And if we refrain from imposing textbook-flat literalness on the lines in question, can we not remind ourselves that Cervantes frequently thinks, like most people, in metaphorical terms? Might the Duchess thus be saying, as I have made her say, something more like “if you like, we'll go even further than that”? Could that not be what the reference to putting someone or something in the “colada” really means? And, finally, isn't “meaning” what, as a translator, I am supposed to be after —not genre theory, not lexical knowledge which distracts us away from, rather than further into, the text being translated?
     I am sorry that Professor Parr thinks I am not aware (I teach in a department of English and comparative literature) of Smollett's dates. I am also aware, however, that my essay was appearing in a journal read primarily by Hispanists, and for their benefit I highlighted the now unusual spelling, in English, of “dutchess.” I was not thereby signalling: Look what an idiot Smollett is, not knowing how we spell this word, but simply: dear readers, this is indeed the spelling employed by Smollett. That seems to me no more than scholarly good manners, which is, alas, more than I can say for Professor Parr's frequently irony-laden comments on my essay. My “puff-piece,” as he calls it, will appear in Spring 1994, in not very different form, as chapter six of mv Penn State UP study, The Art of Translating Prose, an immensely serious and scholarly volume with, I think, some moderately significant linguistic and literary things to say. Would he have me, or any translator, pretend not to prefer our own translations to those of others? Why indeed would we bother doing the translations in the first place? or have (in his

14.1 (1994) Forum 109

words) “the temerity to launch a new version of the Quijote”? (There are of course those who fear any substantial, not to say any significant project. But ought we thus to penalize those who are not so afraid?) I am in truth saddened by any and all such displays. If it did not seem to me important to set the record straight, believe me, I would not have bothered (as I usually do not bother) to reply.


Digitized with the help of Contessa Marion
Fred Jehle Publications of the CSA HCervantes

From: Cervantes: Bulletin of the Cervantes Society of America 14.1 (1994): 109-16.
Copyright © 1994, The Cervantes Society of America

Lighten up, Geoffrey!

     Almost ten years ago (Fall 1984 issue) Cervantes printed a rather unfriendly “review article” by Prof. Geoffrey Stagg on my 1982 Tamesis volume, Cervantes: Pioneer and Plagiarist, in which I argue against Cervantes's authorship of the Porras MS. versions of Rinconete y Cortadillo (R/C) and El zeloso extremeno (ZE). Upon reading Prof. Stagg's less-than-complimentary comments about me and my non-canonical conclusions, my initial reaction was to wonder why the editor of the time, John J. Allen, had neither apprised me in advance of Prof. Stagg's forthcoming negative review nor afforded me the opportunity to rebut his charges against me, either in the same issue or in a subsequent one. After a brief consideration of the matter, I dismissed Stagg's carping as simply one more annoyance that “comes with the territory” of scholarly publication these days, merely a trifle unworthy of more than a moment of my attention.
     Having returned now to the subject of the Porras vs. the 1613 Cuesta versions of R/C and ZE for a lengthy study of the Novelas ejemplares that I am preparing for publication, I have had occasion to reread Prof. Stagg's attack on me and my scholarship, and I would now ask the new editor of Cervantes to “do the right thing” and offer me the chance to rebut that his predecessor denied me. Ten years is a long time to wait for justice, but I believe that what I have to say —especially since I have had a calm and peaceful decade to compose my thoughts— will be of interest to virtually all Cervantists.

     The fundamental problem with Stagg's “review” is that he totally misses the point of my argument. What I was attempting to do in Pioneer and Plagiarist was to present the case —as might a prosecuting attorney— against Cervantes's original authorship of

110 E. T. AYLWARD Cervantes

R/C and ZE (as well as of La tia fingida, but that's another matter). I never said I could prove beyond the shadow of any doubt that Cervantes “borrowed” these stories from another source. Given that the primary piece of evidence in the case, the Porras Manuscript, disappeared in 1823, a definitive verdict in either direction is now impossible. In other words, there is no “smoking gun” here. What I most certainly did attempt to do in my 1982 volume was to lay out all the circumstantial evidence —and there is plenty of it, as I demonstrated in my book— that can be used to argue against Cervantes's claim to full authorship.
     As one might easily imagine, in a case where there are so many gaping lacunae in the documentation, it is necessary to supplement the hard evidence with a good deal of hypothetical supposition to flesh out the argument and create a complete picture. I was simply hypothesizing, which is precisely what countless Cervantine scholars —most notably Manuel Criado de Val, whose research served as a major documentation source but who had reached a different conclusion from mine— had been doing for centuries in this matter (with nary a word of complaint from Stagg). But now Stagg has decided to censure me for duplicating their methodology. Let's take a closer look.
     He writes: “Why is Aylward's explanation one whit more likely than Criado de Val's?” (143). The answer: it doesn't have to be. My goal was only to show that a negative conclusion regarding Cervantes's supposed authorship of these two stories was equally valid, given the fragmentary evidence at hand. Stagg immediately proceeds to offer his own “traditional” hypothesis as to how and why the two stories found in the Porras MS. could have been copies of some early Cervantine drafts. I may not believe such a scenario, but I have no complaint with Stagg's offering it. One of the major differences between Stagg and me is that I do not have a problem with bunched shorts when I read an opinion that happens to be contrary to my own.
     As for Stagg's charge that I don't even try to make a case for the existence of the anonymous primitive draft I suggest as a source for both the Porras and Cuesta versions (143), again, there is no need to do so. It's simply a hypothesis. No one has ever reported seeing such a draft; I was just pointing out yet another possible explanation for the existence of totally different versions of these two stories.
     On pages 144-45 Stagg makes what he thinks is a big deal about the date of composition of R/C, when, in fact, there is absolutely no indication anywhere as to when it was written. The

14.1 (1994) Forum 111

Porras title page states that the events of the story took place in the year 1589, but nothing is said as to the actual composition date. Stagg can conjecture all he wishes about when Francisco Porras de la Cámara actually finished his compilation for the new Archbishop of Seville —he concludes that it was “not earlier than the summer of 1605” (145)— but there is no testimony at all about when Cervantes completed his own version of R/C. We know only that the writer refers to that story in Chapter 47 of the 1605 Quijote. All else is pure speculation. I'll leave it to the reader to decide which hypothesis —Stagg's or mine— is more verisimilar.
     To be fair, Stagg does give me credit for one “revealing insight” (146). He credits me with having seen the full implication of what Criado de Val had only partially glimpsed about the two Porras texts when I wrote: “the Porras and 1613 versions of R/C and ZE were not written in the same style or by the same author” (Pioneer and Plagiarist 36). But he immediately jumps on me for reaching a seriously doubtful conclusion from that observation.
     Stagg's next section is dedicated to a recreation of the authorship argument according to his own theoretical construct. Fair enough. Unfortunately, when he attempts to represent his version graphically on page 147, he produces a scheme in which options (a) and (c) are identical. This is what he shows:

(a) O (i.e., the original —} Porras —} Cuesta;
(b) O —} Cuesta —} Porras;
(c) O —} Porras —} Cuesta.

     Curiously enough, a similar error is repeated in a slightly different context on page 151, but here he posits ‘O’ as a Cervantine borrador; as we see, the duplication is found in (A) and (B):

(A) Cervantes original —} Cuesta —} Porras;
(B) Cervantes original —} Cuesta —} Porras;
(C) Cuesta —} Porras.

     Clearly, in order to represent the Porras and Cuesta versions emerging in parallel lines of transmission from a common source, it would be necessary to represent the process as follows:

(c)         —} Porras


             —} Cuesta

112 E. T. AYLWARD Cervantes

and then:

(B)                            —} Cuesta

Cervantes original— |

                                 —} Porras

     For the record, I do not expect that at any time soon Prof. Stagg will be sending me a note of thanks for clarifying this obvious proofreading error on his part.
     On pages 148-49 Stagg offers four carefully selected examples of scribal miscopying in which the wording of the 1613 Cuesta version is clearly preferable to what is found in Porras. Stagg's methodology here is, on the surface, quite legitimate. What he conveniently fails to mention is that in my book, especially in Chapter IV, I cite a variety of cases in which the wording of one or the other version seems more appropriate to the situation; in some cases I even declare that both versions seem equally apt. What I find unacceptable is Stagg's subsequent conclusion regarding what, exactly, these four samples are supposed to prove: “These few but telling examples suffice to demonstrate that the Porras MS. was copied from a Cervantes original. That that original was Cervantes's own autograph manuscript is also probable” (149).
     Oh, really? Is there no other plausible explanation to be offered? For instance, is it not possible that Cervantes simply copied more accurately than Porras? Or that perhaps Cervantes was working from a cleaner document?
     As support for his theory —and that's all it is, after all— that the “original” document was Cervantes's own, Stagg introduces the comments of M. Romera-Navarro to the effect that Cervantes, among six major Golden Age writers whose penmanship was studied, was known to have sloppiest, most irregular lettering: “el menos uniforme, repito, el más irregular, aun dentro de una misma página, aun firmando su nombre mismo, es Cervantes” (2).
     Well, there we have it! The Porras MS. was clearly copied from Cervantes's own borrador because the great writer was known to have had terrible handwriting! Case closed. Stagg sums it up for us in the following words: “It was easy for the copyist to go wrong” (150).
     The list of dubious conclusions goes on. For some reason, the notion that Cervantes may have been working from a fair copy of the original while Porras was dealing with a corrupt one

14.1 (1994) Forum 113

is dismissed by Stagg as “improbable” (150), because neither copy has come down to us. On the other hand, the existence of a Cervantes borrador —which, by the way, has also not survived— is not only posited, but presented by Stagg as the only possibility that can ever be considered. Let me back this up with Stagg's own words on page 150:

The only combination of the Set that can even be considered is “Cervantes original —} Porras —}Cuesta.”
     . . . if the source [of both versions] were an unknown, it is again unlikely that Cervantes would transcribe more accurately in all the examples given [by Stagg, of course] than Porras.
     The only possibility . . . is . . . that Cervantes and Porras used a Cervantine original.

     And Stagg censures me for drawing doubtful conclusions! This is clearly a case of the pot calling the kettle black. If Stagg believes the existence of a now-lost Cervantine original to be the most likely explanation of the case, I'll allow him his opinion. But to state that this is the only possibility is 1) illogical, 2) hyperbolic and 3) shoddy critical judgment.

Nonetheless, Stagg has managed to convince at least one noted Cervantist of his hypothesis. Daniel Eisenberg, in a foot note to a recent article on Cervantine attributions, opines that Stagg, in his review article about my book, “establece que los textos del ms. Porras se derivan del manuscrito aprovechado para la edición princeps de Juan de la Cuesta” (477, n. 1; my emphasis). Eisenberg, like Stagg before him, has simply overstated the case. I believe Eisenberg would have been more accurate if he had used the verb “propone” instead of “establece” to describe Stagg's theory. There is no fact in what Stagg proposes; it is all supposition/hypothesis —which is precisely what I was doing in my 1982 book. At least I was open about my conjecturing.
     Here comes another startling bit of “news” from Stagg: he maintains that all the scholars who have written on the subject from 1788 to 1983 were guilty of “false logic” in assuming that the Porras version antedated Cervantes's (151). On the basis of the four carefully selected samples where the Cuesta version is preferred over the Porras, Stagg says it is “highly possible” the Cuesta version not only antedated the Porras text, but that also served as Porras's source document (151-52). The logical conclusion of this theory —and Stagg actually uses the term at the close

114 E. T. AYLWARD Cervantes

of his article (153)— is that the Porras text is reduced to the status of a literary red herring.
     In the end I believe most Cervantes specialists will admit that Stagg's theoretical model is certainly no more “highly possible” than my hypothesis, and I gladly give Stagg credit for his imaginative musings here. It's pleasant to note that at least one of us can receive an opposing idea without becoming dyspeptic. Critics like Stagg need to learn not to take themselves too seriously. We're only dealing with literature here, not nuclear physics or a cure for cancer or AIDS. Let us pause for a second to consider, all hypotheses and hyperbole aside, the facts of the dispute between Prof. Stagg and me.

1- We both agree on one key point: the Porras and Cuesta versions of RIC and ZE, respectively, are neither by the same hand nor in the same writing style.

2- Both of us have accepted the possibility of there being a third document in the equation (which Stagg calls O, for “original”), one which may have predated and served as the source document for both the Porras and Cuesta versions of these two stories. Unfortunately, no one has ever seen or heard tell about such a document, so Stagg and I have hypothesized, in opposite directions, about such a document and what it may have contained.

3- In the matter of the surviving texts of RIC and ZE, no one actually knows for certain which version came first. Most critics, myself included, have opted for the Porras as the older text. Stagg believes the opposite to be true.

4- Stagg has attacked me personally and my conclusions in a “review article” published in 1984. He refuses to accept any part of my hypothesis that Cervantes did not create the original versions of RIC and ZE.

5- I have rebutted Stagg's criticism in the present document, but I do not attack his hypothesis that a lost Cervantes autograph manuscript served as the source of the Porras document. I accept his theory as one more that should be considered, even if it happens to take a position 180 degrees from my own.

     The case of Porras and Cervantes is very similar to that of the JFK assassination. All those studying the case, regardless of their personal bias, must recognize that there is simply not enough hard evidence available to reach any firm, incontrovertible

14.1 (1994) Forum 115

conclusion about the origins of R/C and ZE. The trail has been cold for more than a century-and-a-half. Consequently, conjecture abounds. Prof. Stagg and I happen to have conjured up conflicting hypotheses to fill in the lacunae. I urge all interested scholars to review Prof. Stagg's article and my monograph on the subject before deciding.
     However, I also caution my colleagues to remember that both of us are presenting, purely in a speculative mode, the strongest case we can to support our respective points of view. This is very much like a criminal case being tried by a jury. The objective “truth” of the matter is not at hand; all the jurors can try to do at this point is to determine which hypothesis they find more verisimilar.
     And I'll bet that Cervantes, wherever he may be, is having a hearty laugh at our expense.


University of South Carolina


Aylward, E. T. Cervantes: Pioneer and Plagiarist. London: Tamesis, 1982.

Eisenberg, Daniel. “Repaso critico de las atribuciones cervanti nas.” NRFH 38.2 (1990): 477-92.

Romera-Navarro, M. Autógrafos cervantinos. Estudio. Austin, TX: University of Texas UP, 1954.

Stagg, Geoffrey. “The Refracted Image: Porras and Cervantes.” Cervantes 4.2 (Fall 1984): 139-53.

Digitized with the help of Contessa Marion
Fred Jehle Publications of the CSA HCervantes