¹ See Stephen Gilman, “Time and Tense in Spanish Epic Poetry,” Explorations, iv (1955), 72-81; “The Imperfect Tense in the Poema del Cid,” CL, viii (1956), 291–306; these are expanded into Tiempo y formas temporales en el “Poema del Cid” (Madrid, 1961). For the last mentioned theory, see M. Sandmann, “Narrative Tenses of the Past in the Cantar de mio Cid”, Studies in Romance Philology and French Literature Presented to John On (Manchester, 1953), pp. 258-281.

² Gilman (Tiempo, pp. 14 n., 49,108–109) rejects the notion that there may be some connection between the assonance a-a and the use of the imperfect in laisses having that rhyme, where one might expect the poet to have relied upon the convenience of the -ava ending of the first conjugation. His demonstrations on pp. 108-109, perhaps the weakest pages of his book, are unconvincing. For one thing, he studies only five of the laisses which contain -ava imperfects in series (omitting the other ten such laisses), with no proof of their typicalness, and declares that the remaining a-a laisses employ “la asonancia en ‘a-a’ con imperfectos de la primera conjugación sólo escasos y aislados,” which conveys no meaning in a context where the argument is statistical. A further objection is that he does not examine those laisses which have a different rhyme and thus cannot show that they do not differ substantially from the verses in a-a. Gilman may have effectively disposed of Spitzer's views, expressed originally in ZRP, xxxv (1911), 192–212, and subsequently modified, but his own treatment of the rhyme question, in an otherwise stimulating monograph, is totally inadequate. C. C. Smith (YW MLS, xxiii [1962 for 1961], 159) touches upon this very-omission in Gilman, but the matter is ignored by two other reviewers, Charles Aubrun (RR, liv [1963], 49–51) and R. Hamilton (BHS, xl [1963], 50–51). E. Staaf, “Quelques remarques concernant les assonances dans le ‘Poème du Cid’,” Homenaje a Menéndez Pidal (Madrid, 1925), ii, 417–429, studied the frequency of certain key nouns in conjunction with their appearance in the laisses with o-e rhyme but did not deal with the verb.

³ Both the paleographical and critical editions of the Cid by Menendez Pidal were used (Cantar de mio Cid, Madrid, 1954–56, Part iv), the former as a check upon the latter. The tables given are based on a study of the critical text, omitting only those verses which are wholly restored. Reference made to romances is based on the text of the Romancero General (Madrid, 1947), ed. A. González Palencia.

⁴ The theory of multiple authorship, current many years ago but rejected by Menendez Pidal, was revived by him in 1961 (Romania, lxxxii, 145–200). More recently there have been suggestive but undocumented applications of the oral theories of Parry and Lord (cf. A. B. Lord, The Singer of Tales, Cambridge, Mass., 1960) to the Spanish epic by L. P. Harvey on metrical irregularity (BHS, xl [1963], 137–143) and by A. D. Deyermond on the nature of oral composition (BBS, xlii [1965], 1–8). Harvey's views are contested by Robert A. Hall, Jr., (RPh, xlx [1965], 227), who dismisses them too abruptly. We still lack a rigorous study of the Cid based on Lord's methods, such as has been begun for Anglo-Saxon, Old French, and Biblical literature, although Ruth House Webber's Formulistic Diction in the Spanish Ballad (Berkeley, Los Angeles, 1951) succeeds in part for selected romances.

⁵ Cf. Harvey's remark: “If it [the Cid] is as corrupt as the metre would lead us to suppose, then our version is so far from the original as to make any attempt at a critical aesthetic appreciation ludicrous” (p. 137).

⁶ One should think, in this connection, of contemporary improvisational and aleatory music. The composer merely outlines the general trend of the composition, expecting that skilled performers will play their individual parts in proper relation to one another in order to produce the “realized” work. Or, if more exact notation is used, then the order of the bars, staves, or pages, may be altered at will by the performer.

⁷ This is probably what Harvey is speaking of when he says that a future critical text of the Cid would be frankly a “work of creation” (p. 141). He seems to view the MS as the first written text, which leads him to slight the “creativeness” of Menéndez Pidal's text. Deyermond (pp. 5, 6) corrects this view. There seems to be no reason not to consider the critical text, even with its added material and the arranged order of verses, a “realization” of acceptable validity, one that could have been sung on a certain occasion.

⁸ Staaf noted the irregular distribution but accepted Menéndez Pidal's view at that time of single authorship. The latter's change of opinion (Romania, lxxxii [1961], 145–200) likewise takes note of stylistic differences, but now in support of the dual authorship theory.

⁹ Gilman's data (given in tables in Tiempo, pp. 23, 25, 40–43,47, 56, 58–62, 69–71) are selected from verses in narration, omitting formulaic expressions, interjections by the poet, and certain other locutions. Sandmann deals only with narrative passages.

¹⁰ From Gilman's Table m (pp. 40-41) it would seem that in narration, at any rate, there are nearly twice as many verb forms from -er/-ir verbs as from -ar, although more than half of the former are from only four verbs: aver, dezir, ir, ser. That is to say that each of the three groups, -or, the four high-frequency verbs just cited, the other -er/-ir verbs, is represented by about the same number of finite forms. Is this typical of 12th-century Castilian?

¹¹ To carry this speculation further, consider the present subjunctive, which is a stem-rhymed form. In the a-e verses of the Cid, the number of futures in rhyme (i.e., an ending-rhyme), small as it was, was nevertheless nearly four times as great as the number of present subjunctives (of -ar verbs, as mande/salve, stem rhymes). Thus it is impossible to determine what correlation there is, if any, between that form and the a-e assonance. The total of such rhyming subjunctives was eight out of 174 rhyme-verbs, less than 5%. But in the romances studied, we found 34 of 77 (44%) rhyme-verbs in seven o-e poems, and 97 of 233 (42%) rhyme-verbs in twenty a-e romances in the present subjunctive. If only romances with these rhymes were studied, what strange psychological conclusions would be reached with such a surplus of subjunctives? What would an unsuspecting stylistician make of those clusters? The romances in question are nos. 5, 68, 116, 147, 155, 342, 344; and nos. 59, 66,69,173, 176,186,187,193, 231, 257, 303, 305, 359, 382, 387, 419, 441, 445, 484, 488, in tie edition cited.

¹² It is curious, but hardly significant, that even with the extremely few examples of -ra forms one may achieve comparable results. The figures are: 1.25% (-ra forms in a-a group), 0.43% (adjusted frequency for a-a group), 0.55% (average of other three groups). It should be recalled, however, that there are only six -ra forms in rhyme all told, and that of these five are restored forms. Perhaps similar findings could be made in the Cid with regard to the -re and -se forms in the a-e group, with scarcely greater statistical significance.

¹³ It is not possible to determine when a present participle, or indeed any non-finite form, may actually be “replacing” finite forms. The figures do go far to confirm a subjective impression that such passages as the following appear more often in a-o verses than elsewhere: “‘non se faze assí el mercado,/sinon primero prendiendo e después dando”‘(vv. 139-140).

¹⁴ We have not looked into the uses of the infinitive or past participle absolute and we have ignored whatever intereffect there may be among the future and its several periphrastic equivalents. Nor have we attempted to explain the divergencies among the rhyme-groups given in Table vii, that is, why the rate of appearance of finite verbs is not the same. However interesting these studies might be, we believe that they would unduly lengthen the documentation with little gain in strength of argument.

¹⁵ The following laisses, although not in a-a rhyme, have examples of the imperfect which might be difficult of explanation: in a-e, 20, 46, 63, 68, 72, 132; in a-o, 1, 95,99, 111, 119, 123, 130, 134, 147; in o-e, 3, 14, 102, 126, 128. Possibly other lists could be compiled of difficulties with tenses in non-associated rhymes.