¹ Hispanic Review, 35 (1967), 211–26.

² For the most authoritative declaration on the subject see Pedro Henriquez Urena, La poesia castellana de versos fluctuantes, the most recent edition of which is found in Estudios de versificación espahola (Buenos Aires: Universidad de Buenos Aires, 1961), pp. 24–27.

³ No attempt is made in this study to review existing bibliography on the Alex. See Ian Michael, “Estado actual de los estudios sobre ‘El Libro de Alexandre,‘ ” Anuario de estudios medievales, n (Barcelona: Instituto de Historia Medieval de España, 1965), 581–95.

⁴ All quotations from the Alex are identified according to the composite numeration used in the R. S. Willis edition, Elliott Monographs, No. 32 (Princeton, N. J.: Princeton Univ. Press, 1934).

⁶ Kelvin M. Parker, Vocabulario de la Crónica Troyana (Salamanca: Universidad de Salamanca, 1958).

⁶ M. Rodrigues Lapa, Vocabulario galego-português tirado da edição critica das Cantigas d'escarnho e de mal dizer (Coimbra: Galaxia, 1965).

⁷ Documentos lingiüsticos de España, I (Madrid: Centro de Estudios Historicos, 1919), 378. Recorded by Corominas, DCELC, ii, 906; see list of abbreviations and sources on p. 1035.

⁸ Lazrar with some frequency governed a direct object: 165d, 0–424b, P-2057d; Bsl 36d, 69c; Bmi 390a; Bsd 422d. The western dialectal character of the unsyncopated form lazerar is clearly recognized by Yakov Malkiel in his article “La familia léxica lazerar, laz(d)rar, lazeria: Estudios de paleontologia linguistica,” NRFH, 6(1952), 209–76; particularly noteworthy for our needs are pp. 229–51.

⁹ Latin ferre and derivatives, being irregular verbs, gave rise to analogical forms in -ferire. Thus we have this chain of development: offerre oferire ofrir ofrecer. The ofrir stage is documented in the Alex (P-1143c) and in Bsa 7b, 59c, 64c, 68d, 73b, 104a, 132b, 134d; Bdu 209c; Bsd 396a.

¹⁰ E. Alarcos Llorach, Investigaciones sobre “El Libro de Alexandre” (Madrid: Consejo de Investigaciones Cientificas, 1948). On pp. 106,184, he makes the proper choice for verses 364a and 773b respectively.

¹¹ DCELC, iv, 272. I have omitted bibliographical references in this quotation.

¹² The problem is aired in detail in Yakov Malkiel's memorial article dedicated to R. Menéndez Pidal, “Identification of Origin and Justification of Spread in Etymological Analysis,” RPh, 22 (1969), 259–80.

¹³ RFE, Anejo 18 (Madrid, 1934).

¹⁴ A large variety of western phonological traits are found in MS. O of the Alex. Need for brevity precludes giving more than a sampling. Instability of initial pl: progo (P-plugd) 2076d, prazio (P-p!azo) 2040b, laga (P-plagd) 546b, llanto (P-plantd) 889c, lanto (P-planto) 2260a, changer (P-planer) 1777a ; offl: lama ( = ‘flama’ ? 0–679d, which has opposite it P-traua; cf. Bdu 37d “trabónos con tal traba”); of interior bl: asembrado ( ='asemblado') 808a, diabror[ia] (P-diablerid) 2619a, obradas (P-obladas) 1635c, dobra (P-dobla) 2364d; of the interior group bd: dulda (P-dubda) 458d, enbelda (P-enbebdd) 2399d, muelda (P-muepta) 524c, recaldar (P-recabdar) 1572d. Palatalization of initial /: llodo (P-lodo) 2046b. Lack of palatalization of the Latin group cf. dito ( = 'dicho') 1626d. Dissimilation of initial m: nembraua (P-membraud) 299a, nembro (P-mienbro) 1056b. A strong tendency to assimilate adjacent n/l and s/l: bieno (P-bien lo) 2256a; todolos dias (P-todos los d.) 2066c. Metathesis of st: amizad (P-amistat) 1085b. Non-diphthongization of stressed vowels: colouro (P-culebrb) 10c, emendas (P-emiendas) 557a, uergonça (P-verguença) 1568b, longa (P-luenga) 806a, uostra (P-vuestra) 192c, neula (P-niebla) 2042a, elmo (P-yelmo) 572a. Diphthongization of stressed vowel before ch: nueche (P-noche) 1424a. Preference for an implosive syllable structure: uay ( = 'vd) 1630a, queymar (P-quemar) 1601c, cuydey (P-cuydavd) 2427d, proe (P-pro) 760b. Preference for e in unstressed functional words: nen (P-nin) 221 d, se (P-sy 'si') 84a. For more detailed treatment see Alarcos Llorach, Investigaciones, pp. 19–33; Dana A. Nelson, “El Libro de Alexandre: A Reorientation,” SP, 65 (1968), particularly pp. 726–35; and R. Menéndez Pidal's review of the Morel-Fatio edition of the poem, Cultura espahola, 6 (1907), 545–52b.

¹⁶ Origenes del espahol (Madrid: Espasa-Calpe, 1950), p. 91.

¹⁶ This diphthongized version of es was normal in 13th-century Aragonese (as well as in western dialects), although it existed in competition with ye and ys. See Manuel Alvar, El dialecto aragonés (Madrid: Gredos, 1953), pp. 227–28. It is no more an actual trait of the original poet's language than other markedly characteristic Aragonese traits sporadically present in P: juroron ‘juraron’ P-167c; entre ad ‘hasta’ P-2549c; lur ‘su’ P-228a; collyr ‘coger’ P-396d, estonda ‘rato’ P-2514d. See Alarcos Llorach, Investigaciones, pp. 34–45.

¹⁷ Sonbrosa in P-364a originates with a none too scrupulous copyist and can be omitted from this discussion. See the section on sabroso given earlier in the article.

¹⁸ In his “Identification of Origin and Justification of Spread in Etymological Analysis,” p. 272, Malkiel seems also to overlook the presence of sombra in 938b, and he refrains from expressing any opinion on which of the variants in the Alex are correct. If Corominas and Malkiel had noticed the presence of bisyllabic sombra in v. 938b, in an impeccable Alexandrine hemistich, they might have made some interesting comments on the chronology of the fusion of vowels—soombra>sombra—in the area of Castile. By the time of Berceo and his contemporaries, all traces of the form with hiatus had apparently disappeared from the north-central or Castilian zone of the peninsula. It is curious that Corominas, who ostensibly believes in the western origin of the Alex (DCELC, i, xxxiii), passes over the evidence given by sombra without making mention of it.

¹⁹ I do not believe we should give any dialectal significance to the adventitious / of vitio (any more than we would to the j present in vinjo ‘vino’ of P-1184d). The scribe has simply anticipated the combination ti present in the following word, cutiano. Vito/uicto is correctly preserved without epenthesis in P-962b, 1624c, and 0–1932a.

²⁰ Libro de Apolonio: Part 2 (Grammar, Notes, and Vocabulary), Elliott Monographs, xii (Baltimore, Md. : The Johns Hopkins Press, 1922).

²¹ Of course, the easiest critical stance to assume is to say that O-muy quebrantados and O-ya descoraçonados scan correctly and cannot, therefore, be discarded. But such a simplistic view stands at variance with our knowledge of the language of the period, which indicates that the syncopated form descoraznar predominated in the period of the poem's composition. O-muy quebrantados is a scribe's attempt to expurgate archaic mal coraznados. In O-158 5b we observe the scribes' tendency to maintain correct scansion on introducing changes: todos is replaced by ya to compensate for the additional syllable of descoraçonados, in itself a modern form. In connection with all three of these verses observe the highly unstable nature of an adjective like todo(s), one of those filler words that contribute considerably to the plethora of clerecia style. It is used several hundred times in the poem, and the scribes came to look on it lightly or even with disregard. They copy it out of order (77a, 125d, 154b, 315c, 566b, 653c, etc.); they introduce it where it is not (PGG‘-84d, P-86c, 0–121a, P-168c, P-216c, 0–296c, etc.) ; or they simply omit it where it originally was present (0–116d, P-199a, 0–257a, 0–335c, P-344c, O-505d, etc.). A clear solution is not always possible (13d, 155d, 171c, 173d, 193c, 205a, 242d, 346c, 380c, 387d, 407d, 427c, 459b, 610a, 615a, 646a, 700a, 726b, 1167c, 1609b, 1752b, etc.), but even where the solution is not clear, these ambiguous verses serve as testimony of the scribes’ careless handling of the word todo. While the poet had no prejudice whatsoever against accumulating a series of todo(s) in one strophe, the scribes of O in particular take measures to break up or shorten such series (cf. 872, 1109, 1541, 2547). The accumulation of this particular word is frequent in Berceo (Bsd 10, 742; Bso 126, 156; Bmi 35, 726, 876; Bsa 157; Bdu 194–5). One cannot avoid considering the deleterious effect which the alterations involving just this one word have had on the original state of the poem's meter.

²² “Notes on the Versification of El Libro de Alexandre,” Hispania, 19 (1936), 249.

²³ Rafael Lapesa, “La apocope de la vocal en castellano antiguo,” Estudios dedicados a Menéndez Pidal, ii (Madrid: Consejo Superior de Investigaciones Cientificas, 1951), 185–226. In a subsequent study I plan to concentrate on apocope and hiatus, since the truth or falseness of the poet's claim to metrical regularity rests largely on the interplay of these two opposing forces of poetic license. See my unpublished dissertation, “Toward a Definitive Edition of ‘El Libro de Alexandre,‘ ” Stanford 1964, pp. 43–81.

²⁴ Tomás Navarro Tomás, Documentas lingüisticos del Alto Aragón (Syracuse, N. Y. : Syracuse Univ. Press, 1957), p. 21.

²⁵ Navarro Tomás, pp. 18, 21.

²⁶ Manuel Alvar, El dialecto aragonés, p. 196.

²⁷ See conclusion of my earlier article, SP, 65, 747–52.

²⁸ Victor Fernandez Liera, Gramática y vocabulario del Fuero Juzgo (Madrid: Impr. Cásica Española, 1929), p. 218.

In the Fuero de Baeza, ed. Jean Roudil (The Hague: Van Goor Zonen, 1962), we find well documented the verb malfestar, the adverb malfiestamientre, and the adjective malfiesto ‘que ha confesado su delito o culpa.‘

²⁹ O-prigos should not be considered a syncopated form of V-pedricos (=predicos). The former derives from prëcare; the latter from praedicare. Harold G. Jones in his unpublished dissertation “Imperfect Rhyme in Medieval Spanish ‘Cuaderna Via’ Poetry,” Princeton 1968, pp. 121–22, rejects P-pedricos as spurious. The Alex poet often uses a proparoxytone before the caesura, but never at the end of a verse. The ri of prigos in O is the resolution of a scribal abbreviation by the editor, and it could and very likely does represent an original re (pregos) or even better priegos. Although Corominas (DCELC, in, 867) doubts the existence of pregar ‘rogar’ as a popular Castilian word, the documentation he gives is most significant: “Conozco un solo ej. en doc. riojano [!] de 1206 (M.P., DL, 83.5), donde quizá no es más que un romanceamiento ocasional de la voz latina.” In addition we find pregar documented in Bsm 186c, the Auto de los Reyes Magos 32, and in Alex P-2670c (where the scribe of O has substituted rogar). The evidence indicates that pregar had more vitality in the speech of Old Castile in the 12th and 13th centuries than has been supposed. It may have died out gradually because of interference from the homonymie relation between pregar (precare) and pregar ‘clavar, trabar’ (plicare; see Alex 738d and 2308c). I believe that Jones is correct in postulating the postverbal priegos and suggesting that the rime of strophe 705 be rendered as viejos:pri[e]gos:griegos:ciegos, an excellent example of acoustical equivalence used in place of perfect consonant rime. Compare priego to Catalan prec.

³⁰ The influence of legal parlance is to be seen in the phrase “todas sus heredades.” In documents pertaining to deeds and donations we quite frequently find expressions like “con todos e quiscunos dreitos e pertenencias,” “do e lixo a uos amas todas las ditas heredades de la dita capellania,” “ayan poder deprender e denparar e de tener todas e quiscunas las heredades,” “con todas e quiscunas heredades que uuey a e daqui adelante aura.” (See Navarro Tomás, Documentos lingiiisticos, pp. 14–16.) Or “Sabida cosa es quod Rodrigo Azanarez de Bilafauar metio toda la heredade en Sancto Domingo,” “uendemos a muertas a uos don Juan Sanchez . . . por xxx^a morauedis quanta heredat auemos en Madriz.” (In DL, pp. 122, 135.) The Juan Sanchez cited in the last passage is the same one whose name appears in P-2675 of the Alex, an interesting coincidence that I cannot resist pointing out. See Brian Dutton, “The Profession of Gonzalo de Berceo and the Paris Manuscript of the ‘Libro de Alexandre,‘ ” BHS, 37 (1960), 137–45. Still to be studied is the relationship of the Alex text to practices and formulae of the legal profession. Whether the poet himself was able to resist the influence of such legal phraseology and to use heredades without the enhancing adjective is a moot question. The poet may have written “la(s) su(s) heredad(es)” in 1592b and 2586a. Scribes, accustomed to copying legal documents, could then have introduced the enhancing adjective toda(s), which may have been present in the archetype, since we have it in both versions of these particular verses.

³¹ DCELC, in, 853. For a more general treatment of this

particular problem see Menéndez Pidal, Cantor de Mio Cid, i (Madrid: Impr. de Bailly-Baillière e Hijos, 1908), 178.

³² Müllier, Sprachliche und Textkritische untersuchungen zum altspanischen Libro de Alexandre (Strasbourg: J. H. E. Heitz, 1910), pp. 12, 21. Alarcos LI., Investigaciones, pp. 36–37.

³³ Alarcos Llorach fails to cite one lone occurrence of pies, in 1154d ; it falls at the caesura and in no way affects his data.

³⁴ “Notes on the Metre of the Poem of the Cid,” RR, 5 (1914), 24–25.

³⁵ See 47c, 57a, 67a, 84c, 194c, 230a, 342c, 602c, 659a, 710a, 745b, 754a, 766b, 887a, 929d, 958c, 973a, 1046d, 1158b, 1201c, 1297b, 1356d, 1396c, 1581b, 1584b, 1739a, 1746d, 1749d, 1943c, 1988d, 2005a, 2142b, 2178a, 2263a, 2470b, 2604a. Not included in the total cited are verses in which the form appears spurious or has an equally acceptable variant: 0–663d, P-1163c, O-1202c, O-1404c, 0–1574c, 0–2185c; 9c, 115c, 198b. Note particularly O-1202c and 0–2185c which are clearly scribal alterations; both times the copyist (not the poet!) has supplied the western form vioron.

³⁶ La “ Vida de San Millán de la Cogolla” de Gonzalo de Berceo, Estudio y Edition Critica (London: Támesis Books, 1967).

³⁷ If the poet of the Alex, like Berceo, used vidi ‘vi,‘ as 1 suspect he did, it has been completely lost: 149d, 817d, 896d, 1152c, 1156a. Vido, on the other hand, is present some nine times in P (never in O); 116b, 204d, 515a, 521a, 554b, 580c, 588b, 1195a, 2365c. Did Berceo write uidol Probably yes, but it has been largely obliterated. Vido is the usual form in Bso, but in his other works it occurs only as an occasional variant: Bsa 253b; Bsd 601c, 709a; Bsm 46a, 149a, 156a, 159a, 162c. Dutton has generalized vio in his critical edition of Bsm. He fails to make use of the very suggestive article by Yakov Malkiel, “Paradigmatic Resistance to Sound Change: The Old Spanish Preterite Forms vide, vido against the Background of the Recession of Primary -d-,” Language, 36 (1960), 281–346. Particularly helpful are pp. 284–90, in which Malkiel suggests that there is a relationship between the Celtic substratum and the loss of primary d: erosion is complete in the west, where the Celtic substratum was most vigorous; but the primary d becomes gradually more resistant to erosion as one progresses eastward, until, at the opposite extreme, we have the preservation of primary d as a trait representative of undiluted Old Navarro-Aragonese, an area of minimal Celtic penetration. One cannot help but wonder why Dutton generalizes vidieron with its primary d throughout Bsm, while he gives not a second thought to the doublets vido/vío.

³⁸ Malkiel cites one example in the preterite (DL, p. 155) in “Paradigmatic Resistance to Sound Charge . . . ,” p. 327: “Testimonias . . . qui esto vidieron et odieron,” from Cervera del Rio Alhama, dated 1212.

³⁹ Syncopated mintroso appears twice in the poem: 1949d, 1956b. The variants adarg[u]eros/adaragueros of 80c offer no easy solution, since they occur in this one passage alone and are not easily documented outside the Alex. Dutton emends Bsm 455b to read “por qui sue ge[nr]ación fue siempre fatilada,” and supports his decision by pointing out P-1235b “con cueyta del marido e de su generaçion,” in which the second hemistich scans as 8 syllables. An equally defensible choice for the critical version would be the Latin scansion generátio, a possibility Dutton himself recognizes, since Berceo so often falls back on the use of Latin stress where the rhythm requires it. (See B. Dutton, “Some Latinisms in the Spanish mester de clerecia,” Kentucky Romance Quarterly, 14, 45–59.) P-1235b can just as well be corrected with asyndeton or the omission of the second de. Bdu 84b “toda su generation por ellos fue perdida” can also be corrected with the Latinized reading generátio. Generación appears six times in the Alex with no literal sign of syncope: 333d, 404c, P-1235c, 1827d, 2025c, 2107b. 1827d is so corrupt it can hardly be relied upon as evidence in support of a hypothetical syncopated form.

⁴⁰ R. Menéndez Pidal, Manual de gramática histórica espahola (Madrid: Espasa-Calpe, 1958), pp. 73–74, 325–26.

⁴¹ J. Cornu, “Recherches sur la conjugaison espagnole au XIII^e et XIV^e siècle,” Miscellanea di Filologia e Linguistica in Memoria de Napoleone Caix e Ugo Angelo Canello (Florence, 1886), pp. 217–29.

⁴² Versification of the “Cuaderna Via” as Found in Berceo'¹ s “Vida de Santo Domingo de Silos” (New York: Columbia Univ. Press, 1905).

⁴³ I am indebted in a very special way to three scholars whose works have aided me constantly in the gathering of my Alexander materials: Lloyd A. Kasten of the Seminary of Spanish Medieval Studies in Madison, Wis., who has made available to me the revised version of the Tentative Dictionary of Medieval Spanish, now being prepared for publication by him and his colleagues; J. Homer Herriott, who some years ago prepared a word list for the MSS of the Alex; and the now deceased Bart E. Thomas, who in 1937 prepared a concordance of the works of Berceo as partial fulfillment for the doctorate at the Univ. of Wisconsin.