The three parts of La Araucana were first published in a single volume by Pedro de Madrigal in Madrid in 1590 -just four years prior to Cotarelo y Valledor's dating of Cervantes's play (261-66)- in an intellectual milieu in which Neo-Aristotelian precepts were just beginning to circulate and be hotly debated. It is thus not inconceivable that Cervantes first read Ercilla's epic in the Madrid 1590 edition, recognized its links to Neo-Aristotelian issues, and was in part moved to write his play in response to it, signalling his play's connection to La Araucana and to the problem of historical verisimilitude in epic through his allusive title. When a shortened version of this paper was read at the Modern Language Association's Annual Convention in 1993, Professor Mary Gaylord called my attention to another possible conspicuous literary echo recalling the same issues, that of Fernando de Herrera's «El osado español», which exhibits a self-awareness of the alternation of history and poetry in epic discourse. One of Herrera's sonnets, «Esconde tardo Bárgada en tu seno», is as well dedicated to the North African exploits of Don Alvaro de Bazán.

The assault on the fortress of Tucapel in La Araucana is also shot through with echoes of the siege of Troy, including a Trojan-horse-like ruse used by the Indians to penetrate the stronghold. Cervantes's preoccupation with the tension between historical and poetic discourses in epic texts may thus have been conflated imaginatively in terms of both Homer and Ercilla-in terms of its manifestation in both ancients and moderns.

Cervantes knew the argument well since he cites the passage from Horace's Poetics in Don Quijote II, 3 when addressing the perceived narrative lapses of Part I (Ed. Riquer, II, 564). On the Horatian topoi and their accommodation with Aristotle's Poetics in relation to ancient and renaissance literature, see Weinberg, I, 71-714, but especially I, 106-10. In part two of his study, Weinberg goes on to survey the texts whose reception provoked the most heated debate amongst humanists (810 ff.). The incongruities posed by all the Homeric poems, from Proclus's Chrestomatheia through the scholiasts and Renaissance glossators up to modern times, continue to be one of the enduring themes of classical philology. The best introduction remains John Adams Scott's Sather Classical Lectures, collected under the title The Unity of Homer. See especially Chapter 5, titled The Contradictions» (137-71).

On cross cultural cross dressing in Cervantes's plays, see now Ellen Anderson, who assigns an ethical, psychological, and ideological significance to the phenomenon.

Wardropper sees the canónigo de Toledo as a parodic figure, and points to his faulty rhetoric and fallacious logic as proof of Cervantes's desire to distance himself from Neo-Aristotelian precepts («Comedias», 155-56). In his classic study, «Don Quijote: Story or History», Wardropper also insists that Cervantes's art continuously reflects upon the «dilemma posed by the uncertain frontier separating story and history» (87); how «the truth, far from being simple, is complex and ultimately unascertainable in all its complexity» (89). Although Riley perceives a steady Cervantine allegiance to Neo-Aristotelian precepts, he concedes that Cervantes «was often capable of exploiting their often mutually exclusive character». (10)

Although the major Spanish treatise on Neo-Aristotelian aesthetics, López Pinciano's Philosophía antigua poética, was not published until 1596, El gallardo doubtless reflects the theoretical discussions taking place in Spain just prior to the publication of el Pinciano's work. Cervantes also need not have read el Pinciano to be familiar with the issues besetting Neo-Aristotelian theorists, as Riley has suggested (6-12). On the question of the chronology of Cervantes's familiarity with Neo-Aristotelian precepts, and Tasso in particular, see Eisenberg.

John Jay Allen's edition is the source of citations; all not identified by part and chapter as here, are from I, 51 (bracketed all-Spanish texts within quotations are Allen's; see pp. 37-38).

Williamson agrees with Immerwahr and Herrero that this story comes very close to overt parody: «It would appear that for this final tale in Part I Cervantes has assembled most of the conventional elements which exist in different permutations in the other tales and fashioned them into a composite story which verges on open parody and which is again likened to the romances of chivalry» (58) -avant la lettre and by Don Quijote, we must remember. Casalduero earlier found structural links to other intercalated tales and calls it «la reprise de las historias de amor» (198); Murillo modifies this characterization: «It is a reprise of the pastoral narrative on the low-brow level» (131). Perhaps our hero's mention of echoes of the libros de caballerías -so faint as to be difficult if not impossible to hear- is an initial clue to Cervantes' parodic intention.

A brief Spanish-language version of this article was read at the Cervantes conference at the University of Texas-Pan American in Edinburg on 16 April 1994: «Leandra y aquella pregunta palpitante». I thank my friend and colleague David Gitlitz for prompting me to address this topic.

Francisco Márquez Villanueva (139-40) and Stanislav Zimic (67-71) discuss sources, but these are not germane to this study, only the text which Cervantes left and which raises that Question.