One is reminded of Plutarch's «How the Young Man Should Study Poetry» (Moralia 15: 80) and of course the Horatian «enseñar deleitando». The theme is a recurrent one throughout late antiquity and Christian times to justify the reading of fiction. Covarrubias refers to books of chivalry as «ficciones gustosas y artificiosas de mucho entretenimiento y poco provecho», a characterization which seems to support both Riley and Allen.

Particularly with regard to the Quijote, whose Cuesta edition is a fairly reliable text, as Eisenberg, following Flores (1975) has demonstrated in the rest of his article (Eisenberg 1983). In general, conjectural emendation is a dangerous practice as well, because «Excessive subjectivity, an identification with the author leading to the assumption that the editor perfectly commanded his style, or a supersession of author by editor, were bound to discredit both conjectural emendation and by association, to some extent at least the whole practice of editing» (Kane 213). Kane does accept conjectural emendation under certain conditions, however (219). Cervantes' text has been peculiarly victimized by conjectural emendation. According to Casasayas (155): «Parece que esta teoría de la libertad del editor de 'enriquecer' (lo que es un supuesto muy discutible) el texto editado, y que Asensio proclamó abiertamente, es la que siguen los editores, incluso actuales, al socaire de la libertad que les confieren sus estudios e investigaciones. Cfr. los artículos de Flores, Allen y Eisenberg, que citaré luego, que defienden esta postura. Los resultados nefastos al hacer uso de esta autoatribuida libertad están a la vista. El investigador debe dirigir sus esfuerzos al esclarecimiento de la verdad, y, sobre ella, divagar como mejor le venga en gana; pero no le está permitido alterar los datos objetivos que, en sus investigaciones, se halle al paso».

Evidently 11 gross errors per C-gathering had to be corrected for the second Madrid edition (Flores 1975: 80). That does not prove, of course, that entendimiento should have been added to the corrections. The overwhelming % chance would still be in favor of its not being an error.

In so far as the autograph mss. presented by Romera-Navarro are in Cervantes' hand. Evidently not all of them are.

It was the common term used by the Inquisition, for instance (Bennassar 260), and as I stated above, was found rather frequently in the rest of the Quijote. Remember that Eisenberg rejected libros de entendimiento in part because it is unusual.

Evidently in his non-literary writings Cervantes never used a comma (Romera-Navarro 22). Of course it is impossible to know if the same was true of his literary works. it is also possible that the comma merely signaled a rhetorical rather than a logical pause (Eisenberg 1983: 11) or that it meant nothing at all.

Alternatively, one could view the richly nuanced entendimiento as a «portmanteau» word of the type described by Skelton: «now the word seems to operate as a unity of all its powers... now the word is a 'portmanteau'; it contains many personal, historical, and imaginative associations» (1), or one could simply choose not to choose or to delineate among primary and secondary meanings but, as Martin conjectures in interpreting a passage of Mallarmé: «Perhaps the intention is that, unable to give two mutually incompatible concepts any one interpretation, we should allow our minds to hover in a no-man's-land of uncertainty between them» (101). Charles W. Steele postulates that the term mudanza, a word which appears in the Marcela-Grisóstomo episode of the Quijote, has a somewhat related primary and secondary meaning, both of which the reader would be expected to grasp (12).

Beneficio de tercero, provecho de tercero and perjuicio de tercero were apparently commonly-used quasi-legal expressions in sixteenth-century Spain. Gabriel López, a beggar arrested by the Inquisition in 1570 for reciting a heretical prayer, for instance, tells the tribunal that «lo fundamental de sus recursos le viene de la mendicidad y de algunas monedas que le dan por recitar oraciones en provecho de terceros» (Bennassar 223; italics mine).

Los trabajos de Persiles y Sigismunda, Lib. III, cap. 14. Unless otherwise specified, all translations in this study are my own.

El ingenioso hidalgo Don Quijote de la Mancha, Nueva edición crítica de F. Rodríguez Marín (Madrid, 1948), VIII, p. 223.