From: Cervantes: Bulletin of the Cervantes Society of America 13.2 (1993): 61-91.
Copyright © 1993, The Cervantes Society of America

EGINNING with Don Quixote's¹ earliest adventures —the knighting in the inn and the battle with the giants/windmills— he repeatedly interprets his experiences according to his reading of the sixteenth-century romances of chivalry, the libros de caballerías.² In turn, most common types of chivalric incidents are echoed in the mad knight's adventures. Thus, in a few well-known examples, the knight aids the army of a Christian king against a Moorish one (the rebaños, or armies of sheep); he challenges evil giants (the windmills); an enamoured lady pays him a nocturnal visit (Maritornes); he rescues a lady from her kidnapper (the vizcaíno or Basque); he avenges a slain knight (the funeral procession); he challenges a wild beast (the lion); he

     ¹ Throughout this essay I have used the spelling of “Quixote” for the character and Quijote for the work.
     ² I am referring here to those sixteenth-century (and late fifteenth) romances of chivalry written in Spanish. As we know, many other sources for Cervantine chivalric parodies have been suggested, including Italian sixteenth-century romances, Arthurian and Carolingian romances, the Spanish ballads (romances), and others.

competes in jousts and tournaments (his plans for the St. George's Day tournament in Zaragoza); he defends his lady's beauty against all comers (the Toledo merchants); he is whisked away on quests by mysterious means (the enchanted boat); and he changes chivalric epithets according to circumstances (“El Caballero de la Triste Figura,” “El Caballero de los Leones”). The willful deceivers of Don Quixote, like the priest and the barber, Dorotea, Sansón Carrasco, the duke and duchess, and Altisidora, are also readers of the books of chivalry and thus know very well how to invent chivalric plots that fit Don Quixote's expectations.
     Although we cannot know for certain how Cervantes finally felt about the Spanish books of chivalry, it is obvious that he knew them well.³ It would also seem clear that the full parodic effect of Don Quijote depends upon readers who will immediately recognize the chivalric material. Thus the fictional world of Don Quijote, full of readers of chivalric romances, was created for a public made up of readers of the same romances, by an author who was also a reader of the romances. Nowadays, however, with a few notable exceptions, out of the fifty-some Spanish chivalric romances, even the most dedicated Cervantistas seldom go beyond Amadís de Gaula.⁴ Since Diego Clemencín's monumental edition of the 1830's, with its extensive footnotes identifying possible chivalric sources for incidents in Don Quijote, there have been few scholarly attempts to match particular chivalric material

     ³ See Daniel Eisenberg (1987, 3-44) on this point as well as the possibility that Cervantes wrote his own romance of chivalry. Instinct would suggest a couple of possibilities: first, that like the “second author” of Don Quijote of I, 8, Cervantes read everything that came his way, even scraps of paper, and second, that at some point in his life —like Santa Teresa, Ignatius of Loyola, Juan de Valdés, Gonzalo Fernández de Oviedo, the Emperor Charles V as well as many lesser personages— he was fond of these endless tales of fabulous and magical adventures, despite his harsh criticism of their defects, including their possibly pernicious effect on the unsophisticated. In any case, still remaining to be written is a detailed study of the connection between the romances of chivalry and Don Quijote.
     ⁴ Few scholars these days are familiar with sixteenth-century Spanish romances beyond Amadís and the Spanish translations of Breton and Carolingian material. (The Catalan Tirant, in contrast, seems to be undergoing a revival.) Exceptions are Daniel Eisenberg, one of the foremost authorities on the Spanish romances, Martín de Riquer, Sydney Cravens and Marie Cort Daniels on Feliciano de Silva, and a handful who have studied or edited individual romances. For details see Eisenberg, Bibliography, and the last few years of JHP.

with particular incidents in Cervantes's work.⁵ Moreover, those who do write about the romances often may find it difficult to refrain from establishing their felt superiority to their subject matter, Henry Thomas's standard 1920 study being the most obvious example: “All that need be said of this romance . . . is that the author ends by threatening a second part which a merciful providence prevented from being written, or at any rate from surviving” (101).⁶ Clearly these lengthy tales of action and adventure, with their chivalric ethos and pseudo-medieval settings, hold little appeal for fans of the romances' various modern descendants —what Fredric Jameson calls “the half-life of the various paperback lines: gothics, mysteries, bestsellers and the like” (136). Moreover, the majority of the sixteenth-century Spanish romances of chivalry lie buried in rare books collections at the Biblioteca Nacional, the British Library, and a few other locations, making them inaccessible practically as well as aesthetically to all but the most determined prospective reader.⁷
     Clearly, modern readers can enjoy Don Quijote without knowing any more about the romances than what we learn from the novel. However, as P. E. Russell (1985) says, reading it “will scarcely suffice to enable us to understand the kind of impact it had on Cervantes's earlier readers” (29). Francisco Márquez Villanueva has recommended the study of sources as the first step in understanding an author and his work,⁸ and Daniel Eisenberg has specifically spoken in favor of reading the Spanish romances for a greater understanding of Don Quijote, but the

     ⁵ See Clemencín's notes to his edition and also Daniel Eisenberg's comments on the significance of Clemencín's contribution (Romances, 132-36). Jeanne Ellis (372-79) argues for reading the romances in conjunction with the Quijote.
     ⁶ Another of Thomas's typical statements indicates his determination to remain at a safely ironic distance from his subject: “Satisfied with his previous performance, Montalvo announced his intention of serving up more of the same dish” (68). One must wonder in passing what can have persuaded him to take on a project for which he had such little enthusiasm.
     ⁷ For locations of the originally Spanish romances of chivalry see Daniel Eisenberg's indispensable bibliography (1979). I have not listed separately each of the many that I mention in passing in this essay.
     ⁸ See his defense of identifying of topoi in a text, on similar grounds: “El interés del topos no se halla así en el contenido, sino en su aspecto funcional, que es diverso en cada caso y, por to tanto, nada tópico (12). He warns as well of the dangers of “positivist” cataloguing tendencies: “la recaída en áridas y dogmáticas actitudes catalogadoras” (18). Also see Michael McGaha's perceptive summary of the current situation in source studies, as [p. 64] opposed to intertextuality (149). Rafael Altamira defended reading chivalric romances on the grounds of better understanding the epoch in which they were popular, but he was speaking mostly of the Breton and Carolingian material.

continued neglect of the texts indicates that these are minority opinions.⁹ Some critics even seem to argue against reading the romances. For example, E. C. Riley (1954), citing Émile Gebhart's identification of about sixty Amadisean episodes in Don Quijote, asserts that “the many parodic reminiscences of Amadís may stand for all” (37).¹⁰ Luis Murillo (1988) states that it is unnecessary actually to read the Spanish romances of chivalry to find out what they are like because “any reader of Don Quixote can find out for himself: no one has given a more vivid idea of them than Cervantes” (13). Perhaps, but it strikes me as rather like trying to reconstruct a typical Western by watching “Blazing Saddles”.
     Most important, however, it is simply not true that we can know the romances of chivalry solely by reading Don Quijote. For example, after one reads most of the romances, it appears that at least two types of chivalric adventures are experienced by most knightly heroes but not by Don Quixote. First, Cervantes's hero never engages in the typical enterprise of a Christian knight: the forcible conversion of a “pagan” (i.e., Moslem or idol-worshipping) knight, usually after defeating him in battle.¹¹ Likewise, Don Quixote never undergoes that classic chivalric episode in which an enamoured maga (enchantress) works her

     ⁹ See Daniel Eisenberg's criticism (1975) of Armando Durán's Estructura y técnicas and its limitations which for Eisenberg derive partially from reading only unrepresentative works, including Amadís. A number of modern scholars have read Montalvo's work, of course, but few have read the other romances.
     ¹⁰ Interestingly, Riley, like Murillo, then proceeds to cite examples from other romances which are quite unlike Amadís (38-44). Perhaps the most egregious example of mistaken generalizations about the Spanish romances is Roger Walker's 1964 article. Eduardo Urbina's excellent article on Chrétien de Troyes and Don Quijote might at first glance be construed as recommending against reading the Spanish romances of chivalry, but in reality he simply cautions against what he calls “la miópica consideración de fuentes en los libros de caballerías;” in favor of considering the place of Don Quijote within the larger medieval chivalric tradition (139). Arthur Efron defends the Don Quijote-as-imitator point of view but does not advocate reading the romances, while Howard Mancing indirectly recommends reading them (Ch. 1).
     ¹¹ Cf. Whitenack (1988).

erotic magic on the hero so that he remains out of action and in thrall to her —often for years.¹² This is especially intriguing when we consider the constant presence of magic, enchantment, and the erotic in the novel. These two major omissions raise the question of how Cervantes decided on which chivalric material to use or not to use in Don Quijote and more specifically, whether we can come to any conclusion concerning why these particular incidents might have been omitted.¹³ We might assume that his choice to include certain material was governed by each incident's comic/parodic possibilities —dependent first upon instant reader recognition and then upon what could be done with it in a parody. But it is much more difficult to understand why Cervantes omitted certain material, particularly the kinds of incidents just mentioned, which would also have been immediately recognizable to readers of Don Quijote.
     Cervantes surely was aware of chivalric conversion episodes; there occurs, after all, at least one in the majority of books mentioned or referred to in his novel: Tirant, Esplandián, Lepolemo (El Caballero de la Cruz), Belianís, Cirongilio de Tracia, Clarián de Landanís, Florambel de Lucea, and others.¹⁴ Even Amadís de Gaula, not always typical of the Spanish romances, includes a kind of conversion episode, which we will outline here since Amadís appears to have been Don Quixote's favorite. The hero (Amadís) threatens the defeated enemy (the giant Madarque) with death if he does not agree to convert to Christianity:

—Madarque . . . si quieres tomar mi consejo, hazerte he bivir, y si no, la muerte es contigo . . . .  Pues lo que yo de ti quiero es que seas christiano y mantengas tú y todos los tuyos esta ley, faziendo en este señorío iglesias y monesterios, y que sueltes todos los presos que tienes, y de aquí adelante que no mantengas esta mala costumbre que fasta aquí tuviste (II, 979).

Other elements typical of such episodes include the promise that the future Christian will convert all of his relatives and that he will perform good Christian deeds while desisting from what

     ¹² I am presently engaged in a book-length study of the enchantress figure in the Spanish romances of chivalry and elsewhere.
     ¹³ There are other omissions: for example, see Avalle-Arce (63) and Parr (1988, p. 86), among many others, for comments on the knight's missing childhood scenes.
     ¹⁴ For more detail on chivalric conversion episodes (through 1524), see Whitenack (1988).

ever “mala costumbre” has attracted the hero's attention —in this case the practice of not allowing anyone, “caballero, ni dueña ni doncella,” to enter his domain without killing or imprisoning them. But unlike equivalent figures in most other romances, this giant apparently never intended to keep his promise:

El gigante, que ál tenía en el coraçón, dixo con miedo de la muerte:
     —Todo lo haré como lo mandáis, que bien veo, según mis fuerças y de los míos con las de vosotros, que, si por mis pecados no, por otra cosa no pudiera ser vencido, especialmente por un golpe solo como lo fui.

Since Amadís has saved his brother Galaor and King Cildadán from the giant and freed 100 captives, all of whom he sends to his lady Oriana, this adventure is an overall success. However, even though Madarque promises again, he is obviously not going to convert, since we also hear that his savage sister Andandona not only tries to kill Amadís and company as they depart but will eventually turn her brother into as great an enemy of Christians as she is:

Era an enemiga de los cristianos y hazíales mucho mal, y mucho más lo fue allí adelante, y lo fizo ser a su hermano Madarque, fasta que en la batalla que el rey Lisuarte ovo con el rey Arávigo y los otros seis Reyes lo mató el rey Perión, assí como adelante se dirá” (II, 981-82).¹⁵

Madarque never appears again in the narrative (despite the “como adelante se dirá”), but this information, along with the earlier “ál tenía en el coraçón,” tells us that Amadís has failed in his only attempt at enforced conversion.¹⁶

     ¹⁵ In Amadís, the few times Andandona is mentioned, it is invariably in association with the devil, emphasizing that she is the sworn enemy of all Christians: “era muy fea de rostro, que no semejava sino diablo” (II, 980-81); “aquella endiablada giganta” (II, 981); “algún diablo era” (II, 981). The scene in which Amadís's squire Gandalín finally kills her (a knight cannot kill a woman), abounds in references to the devil (II, 1025-27).
     ¹⁶ While many pagans appearing in the romances are “endurecidos” and refuse to convert, it is interesting that the only conversion episode mentioned in Amadís is an unsuccessful one. Perhaps this phenomenon is another indication of the work's probable medieval origins. Cf. Whitenack (1988). We will also remember that in conversions of infidels, as in so many other ways, author Montalvo has Amadís's son Esplandián outdo his father in the sequel.

     Because there are so many of these conversion episodes in the romances, it is neither necessary nor possible to establish that Cervantes was familiar with this particular one, although there is some evidence that would suggest that he was.¹⁷ For example, Madarque is not mentioned in Cervantes's novel, but Sancho Panza praises his evil sister Andandona (who only appears in this section of Amadís) in a comically inappropriate comparison with Teresa Panza: “a no ser celosa [Teresa], no la trocara yo por la giganta Andandona, que según mi señor, fue una mujer muy cabal y muy de pro”.¹⁸ Also, when Don Quixote is certain that Juan Haldudo will keep his promise regarding Andrés (I, 6), it is reminiscent of Amadís's strangely naive acceptance of the giant's promise.¹⁹ The “conversion or death” threat which Amadís issues to Madarque is of course a variation on the classic chivalric demand, echoed in several “surrender or die” moments in Don Quijote, for example at the defeat of the vizcaíno: “le dijo que se rindiese; si no, que le cortaría la cabeza” (I, 9, 146), or of the bachiller Alonso López: “llegándose a él, le puso la punta del lanzón en el rostro, diciéndole que se rindiese; si no, que le mataría” (I, 19, 232). In each of the two battles between Don Quixote and the disguised Sansón, we again see the demands of the victor over the vanquished: Don Quixote to Sansón: “—Muerto sois, caballero, si no confesáis que la sin par Dulcinea del Toboso se aventaja en belleza a vuestra Casildea de Vandalia” (II, 14, 144); and Sansón to Don Quixote: “—Vencido sois, caballero, y aun muerto, si no confesáis las condiciones de nuestro desafío” (II, 64, 534). Like the knights of his books, Don Quixote issues peremptory demands on several occasions: for example, that the Toledo merchants acknowledge the supremacy of Dulcinea's beauty: “Todo el mundo se tenga si todo el mundo no confiesa que no hay en el mundo todo doncella más hermosa . . . .” (I, 4, 100) or that no one follow Marcela: “—Ninguna persona, de cualquier estado que sea, se atreva a seguir a la hermosa Marcela . . . .” (I, 14, 188).

     ¹⁷ Avalle-Arce, for example, says that Cervantes knew Amadís de Gaula “al dedillo” (65).
     ¹⁸ Don Quijote, II, 25, 236. These and all subsequent references will be to the Luis A. Murillo edition.
     ¹⁹ When Amadís orders all of the freed captives to present themselves to Oriana it also reminds us of the galeotes episode in Don Quijote (I, 22), even though there are countless other such episodes in Amadís as well as the other romances.

     Given all of these demands, one might wonder why, then, Don Quixote never issues an ultimatum of “convert or die.”²⁰ He even demonstrates his awareness of chivalric conversion scenes in the rebaños episode, where he describes the two combatants as a Christian king who is fighting his “pagano” counterpart (in this case a Moslem) in order to force him to convert to Christianity before marrying his daughter (I, 18). While this scene is clearly of chivalric inspiration, it contains a humorous twist: a more usual sequence of events in the romances would be for a pagan to request baptism in order to marry someone, rather than for a Christian father to go to war in order to force conversion upon a prospective future son-in-law.²¹ Before converting anyone, Don Quixote would first have had to imagine meeting a non-Christian knight, as in Fernández de Avellaneda's apocryphal Quijote, where the inconsistent protagonist first mistakes a nobleman in Madrid for Perianeo, a pagan prince from Belianís de Grecia (ch. 29), and then imagines that they are in pagan territory (ch. 30). Within Don Quixote's variety of madness, however, the scene is almost always rooted in seventeenth-century Spain, so that even considering the way that he takes the prostitutes for doncellas or the innkeeper for the master of a castle, it would have required another kind of logical leap to take the vizcaíno or Alonso López for a pagan knight.²²
     Perhaps it is a question of the permissible bounds of humor: it is comical and even plausible when the vizcaíno is insulted at Don Quixote's remark that he is no gentleman or when the bachiller, angry about a sprained ankle, excommunicates Don Quixote by citing Tridentine dicta. However, in such a sensitive society, where even to imply that someone was not of Old Christian ancestry was both insulting and dangerous, one wonders how easy it would have been to make “convert or die” humorous.²³ In Avellaneda's Quijote, the actors force the cowardly

     ²⁰ I am indebted to Amy Williamsen for the interesting suggestion that the scenes of chivalric demands in Don Quijote could be intentional parodies of the typical “convert or die” demands of the romances.
     ²¹ It is odd that Efron, who insists that Don Quijote is a follower of chivalric patterns rather than an “individual,” calls this a “stock scene” without mentioning this comic inversion (30).
     ²² An exception could be Don Quixote's destruction of Maese Pedro's puppet show (II, 25), during which he is temporarily taken over by the illusion and transports himself back to the Middle Ages.
     ²³ Note that in the Avellaneda Quijote, while Sancho freely throws around the insults of “paganos” and “luteranos” (actors' episode, Ch. 26), [p. 69] and the false Don Quijote fights with the gentleman's page over whether he really is a pagan or not, he never tries to convert anyone.

Sancho to become a Moslem upon threat of death (ch. 27), and the protagonist concocts a chivalric fantasy in which he will persuade an enamoured pagan lady to convert and marry him (ch. 36), but such farcical scenes do not have quite the resonance of a threat with sword point to neck. Then again, the entire problem could be avoided by having Don Quixote defeat and convert some giants —frequent targets of proselytizing in the romances —as long as they were not windmills (or wineskins, as in I, 35).
     There are other points to consider: for instance, despite the low regard which contemporary churchmen had for the entire chivalric genre, a given author's inclusion of conversion scenes may have come from genuinely pious motives, so that parody would have been unwise. Making the conversion battle comic might even have verged on the dangerous. In Don Quixote's relatively few opportunities to issue demands, at sword point or otherwise, perhaps it would have been inconsistent of him to alternate between sending some people to render homage to Dulcinea and demanding conversions of the others. Cervantes also may have decided to avoid the entire problem by not involving his hero directly with conversion scenes. This is certainly true of the conversions of moros in the rest of the novel, where no one sets out to convert anyone: for example, in the story of the returned morisco Ricote and his family, and in the tale of Zoraida, the Muslim daughter who deserts her father because of her devotion to the Virgin Mary and desire to live in Christian Spain.²⁴ Similarly, in “El amante liberal,” as we will recall, the good turco Mahamut wishes from the beginning to become a Christian. One might also account for the absence of episodes in which the knight is the agent of conversion by applying Marie Cort Daniels's conclusions on Feliciano de Silva, i.e., that the distinct lack of enthusiasm for conversion in his Amadís de Grecia and others of his chivalric romances might be explained by his converso heritage (257-71).²⁵ Certainly it would be difficult to discern

     ²⁴ On the interpretation of the Zoraida tale, see Márquez Villanueva's analysis in Personajes, 115-46. Also see Alison Weber's interesting article on the subject.
     ²⁵ That is, the many who believe that Cervantes was a converso and that Don Quijote is noticeably unenthusiastic on the subject of Christianity might also point to the lack of conversion episodes. On Cervantes's religious [p. 70] views and their possible reflection in his literary work, see Daniel Eisenberg's extended commentary and references in his book on the Quijote (1987), 13-15, n. 40. In the same book he also cites the major evidence for Cervantes's converso heritage (148-149). Also see Canavaggio, Ch. 1.

any stand taken in the novel on the theological debate still raging over the question of enforced conversions. In the rebaños episode, for instance, are we to approve or laugh at the Christian king's effort to effect a conversion by force, or at the idea of two armies fighting an entire battle for such a reason? Or is this irrelevant to the comic center of the episode: Don Quixote's ill-advised assault on the sheep?
     Another key issue left open in the novel also applies to all conversions (particularly the enforced kind): the question of expediency versus sincerity. In the morisco conversions just mentioned, sincerity never seems to be in question. No one appears to doubt the sincerity of Zoraida's prospective conversion from Islam (I, 39-42), for example, perhaps because of her extreme fervor, or perhaps because she is such a standard literary type: the woman who flees with her father's captive.²⁶ As many modern readers have noted with dismay, the morisco Ricote defends the recent expulsion of his people, forcibly converted generations before, even though he also cites the presence of a few sincere Christians among them: “no porque todos fuésemos culpados, que algunos había cristianos firmes y verdaderos; pero eran tan pocos, que no se podían oponer a los que no lo eran, y no era bien criar la sierpe en el seno, teniendo los enemigos dentro de casa” (II, 54, 450-51).²⁷ Ricote's condemnation of the majority might even be a way of separating himself from his compatriots, lending credence thereby to the claims of his daughter Ana Félix that she and her parents are exceptions: “sin que me aprovechase decir que era cristiana, como, en efecto, lo soy, y no de las fingidas ni aparentes, sino de las verdaderas y católicas . . . . tuve una madre cristiana y un padre discreto y

     ²⁶ On the other hand, there seems to be evidence that Zoraida was based upon a real person, the daughter of Agi Morato. Cf. Eisenberg (1987), 102, n. 87. Also see Márquez Villanueva (141-42) on the tradition of the enamoured saracena.
     ²⁷ In support of the possibility that Cervantes had a “moderate” attitude toward this minority, Márquez Villanueva (1975) remarks that no one in the Quijote uses the traditional epithets against moriscos (although there are various slurs against moros). Ellen Anderson has been doing a great deal of interesting work in this area lately.

cristiano” (II, 63, 527).²⁸ This episode resembles the Zoraida tale, both because it unites a converted Muslim woman with a cristiano viejo and because both women's futures are left unresolved at the end (like that of Doña Rodríguez's daughter [II, 56]).²⁹ What Cervantes meant by this indeterminacy is uncertain: it could even be caution on his part, to avoid putting himself in the position of defending the sincerity of a morisca's Christianity. It is also interesting in the case of the renegade that despite earlier doubts by the admiral and the viceroy regarding his trustworthiness (II, 63, 531), once he has decided to rejoin the fold, no one seems to question his decision to reembrace Christianity: “Reincorporóse y redújose el renegado con la Iglesia, y de miembro podrido, volvió limpio y sano con la penitencia y el arrepentimiento” (II, 65, 539).³⁰ But finally, considering all of the possible complications outlined here, the most likely conclusion is that for Cervantes, no advantage in including a conversion episode would have outweighed the inherent disadvantages and even dangers.³¹
     As puzzling as the absence of conversion episodes might seem, even more puzzling is the fact that despite the ubiquitous presence of enchantment in the novel, no enamoured maga ever enchants Don Quixote. The first observation to be made in this context is that the only woman of continuing importance to the protagonist throughout the novel is the lady Dulcinea, who

     ²⁸ When Zoraida (I, 40, 490) denounces all moros as “marfuces,” in contrast, it is in the context of one about to leave the group (and join that of Ricote, incidentally), rather than one seeking to be considered an exception to the majority.
     ²⁹ We will remember that Ana Félix will remain with her father in Barcelona, while Don Antonio tries to arrange permission for her to stay in Spain and her lover goes off to visit his parents (II, 65, 540). Márquez Villanueva also reminds us that Cervantes in his old age was not fond of haply endings (335).
     ³⁰ Perhaps the distinction between the cases of the renegade and of the moriscos may be found in the reference to the former as a “miembro podrido” who became “limpio y sano” through “penitencia y arrepentimiento” (II, 65, 539), while Ricote's apparent use of the same metaphor calls the entire cuerpo rotten: “todo el cuerpo de nuestra nación está contaminado y podrido” (II, 65, 540). In contrast, in Guzmán de Alfarache we will remember the vast doubt Guzmán casts upon the sincerity of his renegade father, who converted easily from Christianity to Islam and back again as it suited his financial advantage. See Whitenack (1985), Ch. 2.
     ³¹ For more on Cervantes vis à vis marranos as well as moriscos, see Canavaggio, Eisenberg (1987), and Márquez Villanueva.

never appears directly.³² However significant this lack of women might be for those trying to analyze the psyche of Don Quixote (or that of his creator), we should not forget the chivalric model which shapes the novel.³³ As one might expect, in the chivalric tales of action and adventure written primarily by men, women do not play a major role.³⁴ As suggested in recent studies by Edward Friedman and Anne J. Cruz on the male-authored picaresque novels with female protagonists, the pícaras are very much male creations. Clearly it is difficult for authors to keep their own stereotypes, prejudices and fantasies from being reflected in characters different from themselves. If this is true of novels which concentrate on female protagonists like the pícaras, it would seem even more likely when women are only shadowy, insignificant characters like those characteristically found in chivalric romances.³⁵ Most women in the romances are incidental to the plot, whether appearing only once or more frequently, for example, mothers, other older women and relatives, wives and fiancées of other knights, servants, and so forth. The most important woman, the hero's lady, is typically an inaccessible beauty from a higher position on the social scale than the hero, so that to win her represents upward mobility for him.³⁶ Often she is, for example, the sole heiress to a kingdom, like the Queen of Bohemia in Floriseo, or to an empire like Constantinople or Niquea, as in Tirant, Esplandián, various of the Amadís series, and many others. As important as the idea of this lady is to the hero, she intervenes little in the action. Her main functions, other than to provide inspiration for the hero, is to remain aloof during most of the narration before finally accepting his proposals,

     ³² Good places to start for commentaries on Dulcinea are Edwin Williamson, 128-31, 171-5, and Arthur Efron.
     ³³ Important recent studies on the problem of women and Don Quijote (and Cervantes) are those of Combet, Johnson, Rossi, and El Saffar (1989 and 1984). Also consult Efron, especially Ch. 2.
     ³⁴ Much could be said about the role of women in Beatriz Bernal's Cristalián de España (1545). See Sidney Park's introduction.
     ³⁵ Several chivalric romances feature Amazon or “woman warrior” types who despite their physical strength and military prowess are rarely important in more than one or two episodes in the plot. An exception would be the Minerva character in Cristalián, about which see Sidney Park's introduction to his edition.
     ³⁶ Ruth El Saffar (1989) sees this pattern (in which “the male aspirant is given power through the woman”) in several chivalric episodes recounted in the Quijote but does not mention the connection with the chivalric romances.

to drive him crazy with irrational jealousy at least once, and possibly to be rescued by him from capture or enchantments.
     Because the lady often maintains her inaccessible stance until almost the end of the narration, and because the hero spends much his time away from her, he is subject to the temptations represented by what we would now call “groupies” —the many women who fall in love with him and seek him out.³⁷ One might have fallen in love with him “de oídas,'' as in the case of a queen of some distant (often pagan) land. She might be the daughter of the highest authority of wherever he spends the night —ranging from the modest castle of an ordinary knight to the palace of a king or emperor— or she might be the proverbial damsel in distress whom he has rescued or to whom he has restored a usurped kingdom. Frequently such a lady —most often still a virgin (doncella)— comes to the hero's bed, often when the knight is resting in a castle, recovering from battle wounds —wounds which have often been tended recently by the very same lady.³⁸ In some romances one has the impression that every unattached young woman is potentially ready to offer herself to the knight, and many, like Amadís's brother Galaor or Amadís de Grecia, cheerfully indulge in one or more sexual indiscretions with these available ladies.³⁹ However, there are also various models of knightly fidelity like Amadís, Esplandián, Arderique, or Lepolemo, and others who use their fidelity as a metaphorical shield to protect them. Some of the extremes to which a faithful knight will go in order to avoid any contact with another woman are almost comical: consider the prolonged agonies of guilt suffered by Clarián de Landanís (I, 1) simply because he once allowed

     ³⁷ Less typical are several heroes, notably Belianís, who spend a large part of the narration trying to decide between two ladies. Even rarer is a type like Florindo, a confirmed woman-hater who rejects all idea of love and marriage until the end, when he is virtually forced to choose a bride for dynastic reasons. See Río Nogueras on Florindo.
     ³⁸ Let us recall, for instance, Queen Briolanja in Amadís, who in the lost medieval version might have succeeded in persuading Amadís to sleep with her —an incident quite unlike Montalvo's more reserved fifteenth-century version. Other examples are the Amazon queen Calafia in the Sergas de Esplandián and several women in Lepolemo, Floriseo, Felixmarte, Tirant, the Clarián series, and many others.
     ³⁹ We will remember one of the barber's arguments for Galaor over Amadís: “tenía muy acomodada condición para todo; que no era caballero melindroso” (I, 1, 73); also that the comic epitafio suggests that Don Quijote did not admire Amadís's brother: “el que . . . en muy poquito a Galaores tuvo” (I, 52, 605).

a lady to kiss him briefly. When one of these ladies is rejected, she is left with few options: she can kill herself, she can waste away from sorrow (as Altisidora pretends to do in II, 70), she can give up and marry someone else (rare), or she can resort to magic.⁴⁰ For example, in Palmerín de Olivia and in Clarián I, 1 we find rejected women who commit suicide, and one who even leaves orders that after her death, her eyes and heart are to be cut out and sent to Clarián in a box —a rather dreadful chivalric motif echoed in the Cave of Montesinos episode.⁴¹
     Among the various rejected ladies in each libro de caballerías there is often a maga who, determined to have the hero, casts an erotic enchantment over him, i.e., philocaptio.⁴² This maga is almost invariably a non-Christian: often a mora or turca, or a generic pagana, like the Queen of India in Floriseo. Not only erotic enchantments but all kinds of magic spells are built-in hazards of the knight's life, as seen by the way so many knights arm themselves with a magical object or charm, often a small piece of jewelry (for example, a ring in Cristalián, a bracelet in Lepolemo), or a magic sword or shield, which must be removed if the enchantment is to work. No metaphorical shield of fidelity is sufficient against the enchantments in the Spanish romances, but often the maga cannot enchant the knight because she is unable to divest the knight of his protection. We will remember that after Don Quixote claims that enchantment has prevented his rescuing Sancho from his blanketing at the inn, he proclaims his

     ⁴⁰ For examples of dying for love, see Platir, Clarián, I, 1; Polismán; and Amadís VII, among many others. Needless to say, those who do commit suicide are almost invariably non-Christians.
     ⁴¹ If the hero rejects the lady, he might try to marry her off to another knight, but if he has accepted her attentions for a night or two and perhaps engendered a child with her, he typically either endows her with a castle or other property or simply abandons her. If no child is involved, she might, again, waste away from sorrow, commit suicide, or devote her life to good works, but she rarely marries anyone else. Once the lady has borne the hero's child, she behaves almost like an eternally bereaved widow —devoted solely to raising her child. Worthy of future examination is this apparent authorial reluctance to have any of the hero's ex-mistresses become involved with other knights. For the “Oliva” vs. “Olivia” problem, see Eisenberg, Bibliography.
     ⁴² On philocaptio, consult P. E. Russell's classic Celestina article (1978) and M. E. Perry. Some of the best on magic and enchantment: Caro Baroja, Garrosa Resina, Maravall, Pavia, Thorndike, and Winkler. A few examples of enchantment episodes: Arderique; Clarián I, 2; II; III; Polismán; Belianís; Floriseo; Polindo; and Florambel.

determination to carry in the future a sword which will protect him against enchantments: “que al que la trujere consigo no le puedan hacer ningún género de encantamentos” (I, 18, 217).⁴³
     The magas use various devices to enchant the hero: often the same kind of item which protects against enchantments, or sometimes a love potion, a candle, music, a magic cloud, or a trick of disguise or metamorphosis, usually to make the knight think that the maga is his lady love.⁴⁴ Like Calypso in Homer's Odyssey, with whom Odysseus spent eight years, or Circe, with whom he spent one year, a maga might keep the hero enchanted for a long time —although sometimes only the nine months required

     ⁴³ Other such objects include the magic swords of Amadís, Reymundo de Grecia, Primaleón, and many others; Lisuarte's scabbard in Amadís VIII; and Platir's shield. In Clarián I, 1 the knight has a magic ring which cures all wounds as well as keeping him from enchantments, and it is not until Clarián III (by a different author) that Daborea succeeds in removing it. Lepolemo has a magic brazalete which protects him from all of the ladies in love with him, and no one ever succeeds in removing it. A maga might also use a similar object to keep him enchanted, in which case the hero's rescuers must remove it to disenchant him. See Giamatti for an analysis of the role of the enchantress in the Italian and English verse epics of the time. We will also remember that in Alcina's enchantment of Ruggiero in Ariosto's Orlando furioso (Canto VI), the knight foolishly does not use his magic shield to protect himself (also, a magic ring is required for his disenchantment).
     ⁴⁴ In Arderique, for example, Blancheflor keeps him in thrall with a ring on each of his little fingers and a stone on a silken cord around his neck. It might be music (Floriseo), or a love potion (Polismán), or another liquid, like the “agua de desacuerdo” in Florambel de Lucea or the sleeping draught (Palmerín), a magic mirror (Polindo), a ring (Florindo), a book (Félix Magno), a magic cloud or fog (also Polismán). Sometimes it is simply generic artes mágicas (Beliflor in Clarián IV), while sometimes these artes are used to deceive the knight, most often to make him think (as in Lancelot-Elaine episodes), that the maga is his lady (Belianís IV) or that he is only dreaming of the encounter. Interestingly enough, although magas and magos (and often their books) are sometimes burned for heresy in the romances (Floriseo, Clarián II and III, and Amadís VIII, for example), the magas who succeed in enchanting the hero (and therefore having sexual relations with him) are rarely punished —Arderique is a major exception. Rather, their fate often resembles that of the rejected ladies already discussed: suicide, wasting away from sorrow, etc. If they bear the hero's son, they are treated especially well, the most exaggerated example being in Floriseo, where Queen Lacivia (!) of In dia bears Floriseo a son and later becomes a great friend to him and his wife, the Queen of Bohemia. See Nunemaker (55) on Alfonso el Sabio's Lapidario for descriptions of the magical properties (erotic and otherwise) of various objects.

for her to conceive and give birth to his child.⁴⁵ And of course the knights enchanted by magas are typically the most faithful ones, who usually are wracked with guilt afterward until invariably, either their own rationalizations or their lady's express forgiveness releases them from guilt. It is important to observe that in the Spanish romances, enchantment by a maga is viewed as an unassailable excuse for infidelity, because the unprotected knight cannot resist the magic spell, however strong his desire to remain faithful. As perhaps an indication of the medieval character of the Spanish romances, we should note the contrast with contemporaneous Italian and English verse epics, which condemn those who allow themselves to be taken in by erotic enchantments.⁴⁶
     The maga initially would seem to be the most powerful figure in the love triangle she forms with the knight and his lady, because she can have any man she wants. However, she is the ultimate loser, first because of the obvious emotional limitations of love induced by magic,⁴⁷ and second, because sooner or later the spell is broken and the knight returns to his heroic career and to

     ⁴⁵ The chivalric author's apparent attitude toward the enamoured enchantress varies considerably. Some are painted as quite sympathetic figures, like Homer's Calypso or some of the non-enchantresses who also fall in love with the hero, like Esplandián's eternal handmaiden Carmela, or the most famous “other woman,” Scott's unfortunate Rebecca —the “dark lady” who loves the hero and cannot have him. Various magas, described as being motivated solely by lust, are condemned by the author, particularly if they are older women in youthful disguise. This litter type resembles the maga Cenotia in Persiles (II, 9).
     ⁴⁶ See Giamatti on the moral implications of erotic enchantment in the works of Ariosto, Tasso, Spencer, and others. We should also note that while in the pseudo-medieval Spanish romances of the sixteenth century, the Christian conversions are distinctly different from their medieval predecessors, the enchantments retain their medieval quality. Notable as well is the reversal of the conventional enchantment in Zayas's “La inocencia castigada,” where the authorities absolve Inés, but her husband does not.
     ⁴⁷ Johann Weyer, in his 1583 De praestigiis daemonum, quotes Plutarch on this subject: “Those who seek sexual pleasures from their spouses by means of amatory arts and love charms, spend their lives in company with persons who are stupefied, demented, and totally ruined” (278). Although Weyer is quoting Plutarch in the context of denying (with multiple classical references) the efficacy of occult amatory arts, the Plutarch passage is ambiguous: does it prove that these arts and charms are ineffective, or that they induce stupor?

his true love, leaving the maga alone, often to raise his child.⁴⁸ In these male-authored texts, the ubiquitous presence of so many beautiful and willing women could well represent some form of masculine wish fulfillment, particularly the fantasy of irresistibility to the opposite sex. However, although the enchantment episodes would seem to offer illicit sexual interludes with no blame attached, they also have nightmarish aspects for the male, not the least of which is that he is captured, imprisoned, dominated, and kept away from his heroic mission by a woman. The fear of women's powers, particularly insofar as they might neutralize the hero's effectiveness, seems to be universal in heroic literature going back to Homer.⁴⁹ Mihoko Suzuki goes so far as to say that all of the female characters in heroic literature “exemplify the personal impulse that opposes public imperatives” (144). In this sense erotic enchantment may also be seen as a metaphor for overwhelming and much-feared feminine seductive powers. Richard Predmore actually makes the logical connection between literary enchantment and “real world” romantic love, but without mentioning the chivalric magas (45-46). Odysseus's dalliance with Calypso, that of Aeneas with Dido, or even the “Joie de la Court” episode in Chrétien de Troyes' Erec et Enide are fine examples of the male perception of the dangers of ceding to a woman's love, with or without magic spells. For example, even if the knight had intended to be faithful to his lady, he can be made to betray her simply by the casting of a magic spell. Classical heroes like Odysseus were little concerned with sexual fidelity to one woman and simply viewed women as part of the spoils of war, but by the Christian Middle Ages, to force a knight to be unfaithful is significant, as we have already indicated. We will remember, for example, Lancelot's absolute fidelity to Guinevere and his despair at the enchantment which causes him to betray her and spend the night with Elaine, the future mother of Galahad.⁵⁰ Especially horrifying for many

     ⁴⁸ Pierre Ullman has reminded me that a maga might well wish to bear the hero's child as an end in itself —something that will be well worth investigating in the course of further research book on magas.
     ⁴⁹ See Lederer, for example.
     ⁵⁰ Malory's might be the most famous rendering of this version of the tale. El Saffar (1989) also finds significance in Don Quijote's use of Lancelot as a model, but more in the context of his being a knight more faithful to his lady than to his lord.

Christian heroes is the idea of sex with a non-Christian, so that some rejections of enamoured ladies (magas and others) have dimensions beyond the wish to remain loyal to one lady. This kind of horror finds a famous echo in the dismay of Sir Walter Scott's Ivanhoe when he learns that the beautiful Rebecca is Jewish.⁵¹ Also in Cervantes's early play El trato de Argel, Aurelio cites Christian doctrine while repelling the sexual advances of the Moor Zahara: “En mi ley no se recibe / hacer yo lo que me ordenas; / antes con muy graves penas / y amenazas se prohíbe” (I, 12).
     Returning to the absence of magas from Don Quijote, let us first consider the peculiarities of Don Quixote's reactions to the women he meets in the course of his travels. If we consider the expectations he derives from his romances of chivalry, we can see that as a knight without any special magical protection he will be subject continually to enchantments of all kinds. And as any reader of the novel knows, Don Quixote uses enchantment as an excuse for almost everything inexplicable or unacceptable, especially personal failures.⁵² He can also expect to be assailed at every turn by erotic temptations which he must overcome if he is to remain faithful to Dulcinea. And among knights he is, as he says, of the faithful variety: “no soy de los enamorados viciosos, sino de los platónicos continentes” (II, 32, 283). As James Parr has said (1988) about Don Quixote's attitude toward women, “they are a combination of Eve and Pandora, not to be trusted, intent on seducing him” (163).⁵³ Within the context of chivalric

     ⁵¹ It is worth recalling that Rebecca's considerable medical skills were regarded as sorcery. Ackerlind is also instructive (14-32) on the connections between women's healing skills and sorcery or magic. On sex with an infidel, there were also chivalric heroes like Palmerín, who had few scruples about making love to various ladies but was nonetheless horrified at the thought of sexual contact with a mora. On the other hand, these dark ladies with obvious sexual desires not only fulfill male various fantasies, but they provide a contrast to the pure future wife of the hero, who rarely admits to sexual feelings and indeed, often is described as having been overcome by him against her will.
     ⁵² Richard Predmore remains possibly the best place to start on Don Quijote and enchantment (36-52). Also see n. 41, above.
     ⁵³ To present even a small selection of critical observations on Don Quijote's wariness/reticence in the presence of women would require many pages. Weiger comments extensively on his “sexual timidity” (34-40), while Efron devotes most of a chapter to it (22-64). The theories of Carroll Johnson regarding the importance of Don Quijote's teenaged niece, and the homosexual interpretations of Combet and Rossi are well-known. For an [p. 79] excellent summary, see Ruth El Saffar (1979). Also see Frye on “chastity and magic” in the romances (153).

expectations, for example, the presence of the two prostitutes at the first castle/inn almost certainly suggests to him an erotic opportunity like those in the romances. That this chivalric scenario might have been on his mind is also suggested by the lines he quotes from the romance (ballad) of Lancelot —a knight renowned for his fidelity to Guinevere, but also, in many versions of the tale, for having conceived Galahad while under an erotic enchantment. Moreover, as many have noted, in the line immediately following the ones partially quoted in Don Quijote (“donzellas curaban dél / princesas del su rocino”), Guinevere takes him to bed with her: “la linda reina Ginebra / se lo acostaba consigo” (Romancero General, I, 198). Don Quixote's prediction of a future time when he will come to the rescue of these “doncellas” also has erotic undertones because of the kind of gratitude so often shown by rescued damsels: “tiempo vendrá en que las vuestras señorías me manden y yo obedezca, y el valor de mi brazo descubra el deseo que tengo de serviros” (I, 2, 86). We should also note the word curar, which with its dual meaning of caring for and curing appears frequently in these chivalric scenes of arrival at a castle.
     It is at the second castle/inn where his erotic fantasies become even more obvious. When the innkeeper's wife and pretty daughter, aided by the grotesque Maritornes, are the ones who attend to (curar) his injuries, Don Quixote's immediate response, with the appropriate rhetoric of regret at a previous romantic attachment, sounds almost as if the daughter had already offered herself to him, as in his books: “pluguiera a los altos cielos que el amor no me tuviera tan rendido y tan sujeto a sus leyes” (I, 16, 200). When later that night he is lying awake, he makes the not illogical assumption that the pretty lady who had helped to cure him would soon arrive at his bed. And it is with the greatest enthusiasm that he seizes the “incomparable fermosura” (Maritornes), in no hurry to release her from his arms until he has explained at length his loyalty to Dulcinea.⁵⁴ Of course part of the humor of the scene derives from the contrast between Don Quixote's self-designation as one of the faithful knights and his evident enjoyment of the rejection scene. Although Don Quixote does not refer here to magic spells the spurned lady might

     ⁵⁴ See Efron on this and other scenes of Don Quijote's rejection of various ladies (52-4).

later cast upon him, he clearly connects magic with erotic rejection in the other Maritornes episode, where, left hanging by one hand (the only part of his body he would allow the eager damsel), he wishes for Amadís's magic sword, “contra quien no tenía fuerza de encantamento alguno” (I, 43, 529).
     When he later rejects Altisidora, he clearly associates both episodes with the magas of chivalric romance:

Mirad, caterva enamorada, que para sola Dulcinea soy de masa y de alfeñique, y para todas las demás soy de pedernal; para ella soy miel, y para vosotras acíbar . . . para ser yo suyo, y no de otra alguna, me arrojó la naturaleza al mundo. Llore, o cante, Altisidora; desespérese Madama por quien me aporrearon en el castillo del moro encantado, que yo tengo de ser de Dulcinea, cocido o asado, limpio, bien criado, y honesto, a pesar de todas las potestades hechiceras de la tierra (II, 44, 374).

He mentions specifically both her imagined despair at his rejection (“desespérese Madama”) and the palace of the “encantado moro” where it all took place. And even more significantly, his final statement proclaims his defiance of the powers of enchantment: “a pesar de todas las potestades hechiceras de la tierra,” which further suggests that he had the chivalric maga episodes in mind. Don Quixote's assertion of moral strength is not simply directed against the “caterva” of lascivious ladies, for whom he is “pedernal,” but also against magic spells. That Altisidora also connects magic with rejected ladies is demonstrated when at the end of his disastrous encounter with the maddened cat (which Don Quixote sees as an evil spirit) she proclaims that all of the magical harm that has come to him is because of his “dureza y pertinacia” as an “empedernido caballero” (II, 46, 386). We might also just note the resemblance of “empedernido” to the earlier “pedernal.”
     The most specific connection between lovelorn damsels and magas occurs shortly afterward, when Don Quixote is surprised by Doña Rodríguez's late-night appearance in his bedroom, and we see once again where his mind inclines. He imagines that Altisidora, ignoring his previous rejection of her, is about to make another assault on his honestidad: “sintió que con una llave abrían la puerta de su aposento, y luego imaginó que la enamorada doncella venía para sobresaltar su honestidad y ponerle en condición de faltar a la fee que guardar debía a su señora Dulcinea del Toboso” (II, 48, 395-96). Then, when he sees the

mysterious form of Doña Rodríguez, swathed in white from head to toe, his immediate reaction relates to magic spells and magas: “Miróla Don Quixote desde su atalaya, y cuando vio su adeliño y notó su silencio, pensó que alguna bruja o maga venía en aquel traje a hacer en él alguna mala fechuría, y comenzó a santiguarse con mucha priesa” (II, 48, 396).
     Why, with all of the other kinds of enchantment in Don Quijote, does the knight never speak of being enchanted by an enamoured maga? As we can see by the events just discussed, as well as the presence of all those magas in the romances of chivalry, it is not because the author is unaware of magas. On the contrary, the world of magic and enchantment, as Américo Castro observed long ago, seems to have fascinated Cervantes, which might partially explain his attraction to the libros de caballerías.⁵⁵ Several of his other works also contain various references to magic and the supernatural.⁵⁶ Naturally in those days of Tridentine strictures on imaginative literature and the dangers of Inquisitional investigation to suspected practitioners of magic, an author could hardly allow the presence of efficacious magic in his works, and Cervantes therefore always weakens or questions its effects within his plots. The best example is Don Quijote itself, where ubiquitous invisible enchanters are safely located in the mind of a deranged knight. But also, after Berganza's lengthy description of the powers of the witch Camacha in “Coloquio de los perros,” Cipión immediately calls the account a “grandísimo disparate” (309). It is also very interesting in this context what James Parr (1988) says about Don Quixote's lifelong chastity and his determination to remain faithful: the author “never puts him to any real test” (86). Or, in the context of this discussion, no maga casts an irresistible spell on him. In person he has only to resist Maritornes (confused in his mind with the innkeeper's daughter). In the middle of the first and most erotically dangerous scene she is rescued by the “encantado moro” and in the second, he completes the rejection and suffers enchantment, which although not erotic, certainly immobilizes him). Although he must reject both marriage to the Princess Micomicoma and the advances of Altisidora, in neither case is he in as close physical proximity to the lady as he is to Maritornes. With Doña Rodríguez,

     ⁵⁵ See Castro, El pensamiento de Cervantes, 94-104.
     ⁵⁶ For further discussion of enchantment in Cervantes, see Avalle-Arce (1974), Castro (1925), Garrote Pérez, Harrison, Maravall, Predmore, and Stackhouse, for instance.

he evidently observes —after the initial based upon the resemblance between the nighttime visit and countless similar chivalric episodes— he evidently recognizes that she represents no erotic threat.
     Moreover, when it comes to Don Quixote's own possible enchantment by a lascivious maga, it almost seems that the knight was aware of a theological conflict. Catholic doctrine defended the preeminence of Christian faith over diabolical (“black”) magic and of free will over enchantment, and this was echoed in seventeenth-century literature. For instance, we will recall that the devil in Calderón's El mágico prodigioso cannot deliver Justina to Cipriano as he had promised, because she is a Christian endowed with free will. Similarly, in María de Zayas's miracle tale, “La perseguida triunfante,” the magician's powers over Beatriz are limited because she is under the Virgin's special protection, like a character in a Marian miracle tale from the Middle Ages. Evidence from emblem collections as well as poetry from the period also consistently defends free will over the powers of erotic magic, and we see the same in several of Cervantes's works. In Persiles, for example, it is because of the free will of Periandro/Persiles that Hipólita cannot attract him through magic spells and must content herself with employing an hechicera to make Auristela/Segismunda deathly ill: “no que mudase la voluntad de Periandro, pues ya sabía que era imposible” (450).⁵⁷ Even the maga Cenotia (in the same work) defends the preeminence of the will over magic: “Puesto que en mudar las voluntades, sacarlas de su quicio, como esto es ir contra el libre albedrío, no hay ciencia que lo pueda, ni virtud de yerbas que lo alcancen” (202).
     Similarly in El trato de Argel, when the Moorish slave Fátima tries on behalf of her mistress Zahara to cast an erotic enchantment on Aurelio, a devil actually arrives from hell to tell her that she is wasting her time because Aurelio is a Christian and not subject to these kinds of spells (II, 40). And in “El licenciado Vidriera” there is a moment which resembles all of those chivalric rejection scenes: a woman spurned by Tomás takes the advice of a morisca and gives him a Toledan quince with a magic potion in it, which only serves to make him ill for six months and then to derange him —but not to attract his love. The narrator, commenting on such hechizos, makes fun of the idea that they could be effective against free will: “como si hubiese en el mundo yerbas,

     ⁵⁷ All Persiles citations come from the Avalle-Arce edition.

encantos ni palabras suficientes a forzar el libre albedrío” (33-34). And importantly, within Don Quijote itself the knight repeats the same idea quite plainly to the old alcahuete in the galeotes episode: “Aunque bien sé que no hay hechizos en el mundo que puedan mover y forzar la voluntad, como algunos simples piensan; que es libre nuestro albedrío, y no hay yerba ni encanto que le fuerce” (I, 22, 269). He also defends free will in the Golden Age speech, when referring to young ladies whose “perdición nacía de su gusto y propria voluntad” (I, 11, 157). The underlying implication of these assertions is that someone who is enchanted is in some way susceptible, consciously or not, and is therefore morally responsible, as is the moral stance in the earlier English and Italian verse epics, as mentioned above. A Roman Inquisitional manual of the time states that the Devil cannot coerce the will, although he can “stir human fantasy, either by way of charms, or by inflaming the blood and the humours . . . .”⁵⁸
     The idea that an erotic enchantment cannot work without human weakness or complicity, as we see in the Italian and English verse epics of the sixteenth century, is far from new. When Santa Teresa recounts in her Vida (Ch. 5) the tale of a priest enchanted by his mistress, she considers him deluded but also morally responsible. However, in the creative literature of the Middle Ages, reflected in the later, pseudo-medieval Spanish romances of chivalry, the prevailing convention is that knights are powerless against erotic enchantments, unless protected by their own magic. We should also note, however, that in the sixteenth and seventeenth centuries, the victims of other, non-erotic magic spells and enchantments (curses on their crops or livestock, casting of the “evil eye,” etc.), might be considered deluded, but hardly deficient in will power. In most seventeenth-century literature the cases of magic and enchantment —so much to the taste of both author and public— are tempered in some way. In Zayas's “La inocencia castigada,” for example, there is the strong suggestion that Inés —despite the title— was susceptible to erotic enchantment because of her husband's neglect of her sexual needs, and in “El desengaño amando” the supposed victim of enchantment is shown to be a money-hungry philanderer whom the maga might have attracted easily enough without resorting

     ⁵⁸ Desiderio Scaglia (d. 1639), Prattica per procedere nelle cause del Sant' Officio. Quoted in Tedeschi, 233.

to magic.⁵⁹ So also with Castillo Solórzano's “La fantasma de Valencia,” which plays upon the rage for ghost stories but avoids any possible accusations of giving credence to magic by revealing that the fantasma is only an ordinary human being in disguise. It is also worth recalling in the “Escrutinio” chapter of Don Quijote the priest's much-cited condemnation of Montemayor's magical solution to problems of love in La Diana: “soy de parecer que no se queme, sino que se le quite todo aquello que trata de la sabia Felicia y de la agua encantada” (I, 6, 118). Avalle-Arce has noted (1974) that Cervantes's rejection of supernatural solutions is also implicit in the episodes of Juan Palomeque's inn in Part I, a place where as in Felicia's palace, many problems of love are resolved, but without resorting to the supernatural.⁶⁰ And with regard to the authorial distance maintained from the magic elements in Persiles, Stephen Harrison has recently suggested that Cervantes went back and added rationalizations to all of the magical elements after completing the novel. Once again, when magic is discounted, erotic enchantment may be seen as a metaphor for seductive power, in which the complicity of the victim is more likely.
     To return to Don Quixote's case, then, for him to assert the presence of evil and envious enchanters represents no theological or moral problem within his chivalric fantasy world. On the other hand, Inquisitional censors were alert to possible heresies in Don Quijote as well as in other works, as we see in the censors' well-known demand that Cervantes modify the Duchess's remark to Sancho on faith versus good works in II, 36.⁶¹ Then again, it is possible that both Don Quixote's remarks to the alcahuete and his pattern of behavior with what he perceives as predatory women mean that the omission of an enchantment by a

     ⁵⁹ Kenneth Stackhouse talks about the “attenuation” of magic in several of Zayas's tales but curiously does not mention the erotic enchantment that is most intricate and difficult, the one found in “La inocencia castigada.” Cf. Whitenack's Zayas article, forthcoming in the Williamsen and Whitenack volume.
     ⁶⁰ As Avalle-Arce remarks in this context, “Se hace obvio que la solución ofrecida por Montemayor no es tal en opinión de Cervantes” (89). Avalle-Arce also refers to what Américo Castro said on this subject in El pensamiento de Cervantes: “Como explicó hace años Américo Castro, el amor, fuerza vital, no puede ser desviado por medios sobrenaturales” (89-90). I am indebted to Michael McGaha for reminding me of Avalle-Arce's remarks.
     ⁶¹ On this point see Américo Castro, Hacia Cervantes.

lascivious maga has less to do with authorial concerns with Inquisitional censors than with consistency in Don Quixote's own moral code. Despite the many irresistible enchantments of faithful lovers depicted in the romances of chivalry, Don Quixote surely knows, even when mad, that he cannot be thus enchanted because he cannot be susceptible, or else his whole self-created identity as a faithful lover is destroyed. Given the prevailing view on erotic enchantment, to allow himself to be enchanted by a lecherous lady would reflect upon his virtue by implying that he is not truly committed to sexual fidelity. Of course his resistance actually makes him superior to all of his model knights, since almost none of them could resist enchantment without magical help. Hence the significance of his defiance of spells in the Altisidora episode: he is usually only too willing to cede all power to enchanters and enchantments (particularly when it is to his advantage to do so), but he cannot damage his identity as the faithful lover without also damaging the essence of what makes him Don Quixote.
     In conclusion, as we look at two episodes not parodied by Cervantes —the knight's conversion of paganos and his enchantment by magas— we can conclude that there is nevertheless some trace of them in the novel. Don Quixote (and thereby Cervantes) is demonstrably aware of both types of episode. The questions raised by the transformation of major episodes into shadowy references are in themselves a justification for reading the romances with care. Christian conversion and the “purity of blood” issues were so explosive at the time that having the mad knight try to convert someone to Christianity might have been seen as denigrating the Christian obligation to spread the faith and convert all unbelievers (even if one doubted their sincerity later). Even Avellaneda, we will recall, makes his “conversion or death” scene into a burlesque —just one more trick on his doltish Sancho, and while his erratic Don Quixote creates a pagan knight out of a Castilian nobleman, he never mentions converting him.⁶² Perhaps the most attractive thought is that Cervantes himself rejected enforced conversion as a modus operandi, but as a reasonable, humane man, rather than as one of converso

     ⁶² Of course in “El retablo de las maravillas” Cervantes parodies people's concerns about their ancestry, but in the spirit of farce. He stops far short of criticizing current Purity of Blood laws, although we might wish that we could find evidence of such criticism.

heritage.⁶³ Certainly there is no evidence that he approved of conversions for convenience, and at the very least, it seems as if he did not wish to have his knight responsible for trying to turn a moro into a morisco, with the attendant complications.⁶⁴ Cervantes may well be offering a vision of the moriscos which, however unsatisfactory to modern tastes, was quite open-minded for his time. As Márquez Villanueva notes, just the presence of one sympathetic morisco is in itself original in the period.⁶⁵ In the case of the enamoured maga, as we have said, one should first recall current beliefs on erotic enchantments, since Don Quixote tends not to go explicitly against any Catholic doctrines, no matter what his state of mind. Even more significant is the question of the ethical consistency of Don Quixote as a character. One of the most disturbing aspects of Avellaneda's protagonist is his renunciation of Dulcinea: it is hard to imagine Cervantes's Don Quixote doing the same —the only possible interpretation of his falling victim to an erotic enchantment. As in the case of his attitude toward moriscos, Cervantes appears here to be approaching a modern idea: enchantment as a metaphor for seduction. And finally, this kind of study raises possibilities of alternative readings —possibilities which cannot be made certainties but still remain intriguing. If nothing else, a glimpse into the mind and creative methods of the all but inaccessible Cervantes would seem to be sufficient reward for reading all of those chivalric romances.

     ⁶³ On this point see Allen, II, 102-3.
     ⁶⁴ Márquez Villanueva comments that neither Zoraida nor Ana Félix is shown living happily ever after with her Christian husband. Does this represent Cervantes' doubt on the sincerity of moriscos, or is it simply realism, in terms of possible happiness for such couples, or does it reflect his decreasing fondness for happy endings?
     ⁶⁵ See Personajes, 229-335, on the Ricote episode and the morisco problem and Ackerlind (33-56) on the moriscos and other marginal social groups in the period. We might also recall Cervantes's apparent parody (following I, 52) of the “libros plúmbeos” hoax perpetuated by Granada moriscos in the late sixteenth century. See Eisenberg (1987), 72, n. 80, for information and references. Also see Castro (1964), 22ff.

Ackerlind, Sheila R. Patterns of Conflict: The Individual and Society in Spanish Literature to 1700. New York: Peter Lang, 1989.

Allen, John J. Don Quixote: Hero or Fool? Part I. Gainesville: U of Florida P, 1971. Part II. Gainesville: U of Florida P, 1981.

Altamira, Rafael. Arte y sociedad. Barcelona: Cervantes, 1921.

Avalle-Arce, Juan Bautista. Don Quijote como forma de vida. Madrid: Castalia, 1975.

——. La novela pastoril española. Madrid: Istmo, 1974.

Bernal, Beatriz. Cristalián de España. Ed. Sidney S. Park. Diss. Temple, 1980.

Canavaggio, Jean. Cervantès. Paris: Mazarine, 1986. Trans. J. R. Jones. New York: Norton, 1990.

Caro Baroja, Julio. “La magia en Castilla durante los siglos XVI y XVII.” Algunos mitos españoles. Madrid: Ediciones del Centro, 1974. 185-295.

Castro, Américo. Hacia Cervantes. 1930. Madrid: Taurus, 1967.

——. El pensamiento de Cervantes. 1925. Barcelona: Noguer, 1972.

——. El Quijote, taller de existencialidad.” Revista de Occidente, Series 2, 18 (1964): 1-33.

Cervantes Saavedra, Miguel de. El ingenioso hidalgo Don Quijote de la Mancha. Ed. Luis Andrés Murillo. 2 vols. Madrid: Castalia, 1987.

——. Novelas ejemplares. Ed. Francisco Rodríguez Marín. Madrid: Espasa-Calpe, 1975.

——. Teatro completo. Ed. Agustín Blánquez. Barcelona: Iberia, 1966.

——. Los trabajos de Persiles y Sigismunda. Ed. Juan Bautista Avalle-Arce. Madrid: Castalia, 1986.

Combet, Louis. Cervantès ou les incertitudes du désir. Lyons: Presses Universitaires de Lyons, 1981.

Craig, Cynthia C. “Enchantment and Disenchantment: A Study of Magic in Orlando Furioso and Gerusalemme Liberata.” Comitatus 19 (1988): 2-45.

Cravens, Sydney Paul. “Feliciano de Silva and His Romances of Chivalry in Don Quijote.” Inti 7 (1978): 28-34.

Cruz, Anne J. “Sexual Enclosure, Textual Escape: The Pícara as Prostitute in the Spanish Female Picaresque Novels.” Seeking the Woman in Medieval and Renaissance Texts, ed. Sheila Fisher and Janet Halley. Knoxville: U of Tennessee P, 1991. 135-59.

Daniels, Marie Cort. “The Function of Humor in the Spanish Romances of Chivalry.” Diss. Harvard, 1976.

Durán, Armando. Estructura y técnicas de la novela sentimental y caballeresca. Madrid: Gredos, 1973.

Efron, Arthur. Don Quijote and the Dulcineated World. Austin: U of Texas P, 1971.

Eisenberg, Daniel. Castilian Romances of Chivalry in the Sixteenth Century: A Bibliography. London: Grant and Cutler, 1979.

—— . Review of Armando Durán, Estructura y técnica de la novela sentimental y caballeresca (Madrid: Gredos, 1973). HR 43 (1975): 425-29.

——. Romances of Chivalry in the Spanish Golden Age. Newark, DE: Juan de la Cuesta, 1982.

——. A Study of Don Quixote. Newark, DE: Juan de la Cuesta, 1987.

El Saffar, Ruth. Beyond Fiction: The Recovery of the Feminine in the Novels of Cervantes. Berkeley: U of Calif. P, 1984.

—— . “Sex and the Single Hidalgo: Reflexions on Eros in Don Quijote.” Studies in Honor of Elias Rivers, ed. Bruno M. Damiani and Ruth El Saffar. Potomac, MD: Scripta Humanistica, 1989. 76-93.

Ellis, Jeanne Louise Barnes. “The Saved and the Damned: Cervantes, the libro de caballerías, and the Novel.” Diss. Cornell, 1984.

Fernández de Avellaneda, Alonso. Don Quijote de la Mancha. Ed. Fernando García Salinero. Madrid: Castalia, 1972.

Friedman, Edward H. The Antiheroine's Voice: Narrative Discourse and Transformations of the Picaresque. Columbia: U of Missouri P, 1987.

Frye, Northrup. The Secular Scripture: A Study of the Structure of Romance. Cambridge: Harvard UP, 1976.

Garrosa Resina, Antonio. Magia y superstición en la literatura castellana medieval. Valladolid: U de Valladolid, 1987.

Garrote Pérez, Francisco. “Universo supersticioso cervantino: su materialización y función poética.” Cervantes. Su obra y su mundo. Actas del I Congreso Internacional sobre Cervantes, ed. Manuel Criado de Val. Madrid: Edi-6, 1981. 59-74.

Giamatti, A. Bartlett. The Earthly Paradise and the Renaissance Epic. 1966. New York: Norton, 1989.

Harrison, Stephen. “Magic in the Spanish Golden Age: Cervantes's Second Thoughts.” Renaissance and Reformation 4 (1980): 47-64.

Jameson, Fredric. “Magical Narratives: Romance as Genre.” New Literary History 7 (1975): 135-63.

Johnson, Carroll B. Madness and Lust: A Psychoanalytical Approach to Don Quixote. Berkeley: U of California P, 1983.

Lederer, Wolfgang. The Fear of Women. New York: Harcourt, 1968.

Mancing, Howard. The Chivalric World of Don Quijote: Style, Structure, and Narrative Technique. Columbia and London: U of Missouri P, 1982.

Maravall, José Antonio. Utopía y contrautopía en el Quijote. Santiago de Compostela: Pico Sacro, 1976.

Márquez Villanueva, Enrique. Personajes y temas del Quijote. Madrid: Taurus, 1975.

McGaha, Michael. “Intertextuality as a Guide to the Interpretation of the Battle of the Sheep.” On Cervantes: Essays for L. A. Murillo, ed. James A. Parr. Newark, DE: Juan de la Cuesta, 1991. 149-61.

Murillo, Luis A. A Critical Introduction to Don Quijote. New York: Peter Lang, 1988.

Nunemaker, J. Horace. “An Additional Chapter on Magic in Mediaeval Spanish Literature.” Speculum 7 (1932): 556-63.

Parr, James A. Don Quixote: An Anatomy of Subversive Discourse. Newark, DE: Juan de la Cuesta, 1988.

Pavia, Mario. Drama of the Siglo de Oro: A Study of Magic, Witchcraft and Other Occult Beliefs. New York: Hispanic Institute, 1939.

Perry, Mary Elizabeth. Gender and Disorder in Early Modern Seville. Princeton: Princeton UP, 1990.

Predmore, Richard L. The World of Don Quixote. Cambridge: Harvard UP, 1967.

Riley, Edward C. Don Quixote. London: Allen and Unwin, 1986.

Río Nogueras, Alberto. “Una trayectoria caballeresca singular: el Don Florindo de Fernando Basurto.” JHP 12 (1988): 191-206.

Riquer, Martín de. “Cervantes and the Romances of Chivalry.” Trans. Joseph R. Jones. Don Quixote. The Ormsby Translation. Ed. Joseph R. Jones and Kenneth Douglas. New York: Norton, 1981. 895-913.

Rodríguez de Montalvo, Garci. Amadís de Gaula. Ed. Juan Manuel Cacho Blecua. 2 vols. Madrid: Cátedra, 1987.

Romancero General. Ed. Agustín Durán. BAE 10. Madrid: Rivadeneyra, 1859.

Rossi, Rosa. Escuchar a Cervantes: Un ensayo biográfico. Valladolid: Ámbito, 1987.

Russell, Peter E. Cervantes. Oxford: Oxford UP, 1985.

——. “La magia como tema integral de La Celestina.” Temas de La Celestina y otros estudios del Cid al Quijote. Barcelona: Ariel, 1978. 241-56.

Stackhouse, Kenneth. “Verisimilitude, Magic and the Supernatural in the Novelas of María de Zayas y Sotomayor.” Hispanófila 62 (1978): 65-76.

Suzuki, Mihoko. Metamorphoses of Helen: Authority, Difference and the Epic. Ithaca: Cornell UP, 1989.

Tedeschi, John. The Prosecution of Heresy. Collected Studies on the Inquisition in Early Modern Italy. Binghamton, NY: MRTS, 1991.

Thomas, Henry. Spanish and Portuguese Romances of Chivalry. Cambridge: Cambridge UP, 1920.

Thorndike, Lynn. A History of Magic and Experimental Science. 7 vols. New York: Columbia UP, 1923.

Ullman, Pierre L. “Romanticism and Irony in Don Quixote: A Continuing Controversy.” Rev. article of Anthony Close. The Romantic Approach to Don Quixote. Papers in Language and Literature 17 (1981): 320-23.

Urbina, Eduardo. “Chrétien de Troyes y Cervantes: más allá de los libros de caballerías.” Anales Cervantinos 24 (1986): 137-47.

Walker, Roger. “Don Quixote and the Romances of Chivalry.” New Vida Hispánica 12 (1964): 13-14, 23.

Weber, Alison. “Padres e hijas: una lectura intertextual de La historia del cautivo.” Actas del segundo Coloquio Internacional de la Asociación de Cervantes: Alcalá de Henares, 1989. Barcelona: Antropos, 1991. 425-31.

Weyer, Johann. De praestigiis daemonum. 1583. Binghamton, NY: MRTS, 1991.

Weiger, John F. The Individuated Self: Cervantes and the Emergence of the Individual. Athens, Ohio UP, 1979.

Whitenack, Judith A. “Conversion in the Spanish Romance of Chivalry, 1490-1524.” JHP 13 (1988): 13-39.

——. The Impenitent Confession of Guzmán de Alfarache. Madison: Hispanic Seminary, 1985.

——. “‘Lo que ha menester’: Erotic Enchantment in ‘La inocencia castigada’ of María de Zayas y Sotomayor.” María de Zayas: The Dynamics of Discourse, ed. Amy R. Williamsen and Judith A. Whitenack. Teaneck, NJ: Fairleigh Dickonson UP. In press.

Williamson, Edwin. The Half-Way House of Fiction: “Don Quixote” and Arthurian Romance. Oxford: Clarendon, 1984.

Winkler, John J. “The Constraints of Desire: Erotic Magical Spells.” The Constraints of Desire: The Anthropology of Sex and Gender in Ancient Greece. New York and London: Routledge, 1990. 71-98.

Zayas y Sotomayor, María de. Desengaños amorosos, ed. Alicia Yllera. Madrid: Cátedra, 1983.