The reader will see an attempt to restore the Chanson de Roland, with musical notes, in Dr. Burney's Hist. ii. p. 276. See more concerning the Song of Roland, vol. iii. appendix, sect. ii. note M.
[S] [An eminent French writer, &c.] "M. l'Eveque de la Ravaliere, qui avoit fait beaucoup de recherches sur nos anciennes Chansons, pretend que c'est à la Normandie que nous devons nos premiers Chansonniers, non à la Provence, et qu'il y avoit parmi nous des Chansons en langue vulgaire avant celles des Provençaus, mais posterieurement au Regne de Philippe I. ou à l'an 1100."—v. Revolutions de la Langue Françoise, à la suite des Poesies du Roi de Navarre. "Ce seroit une antériorité de plus d'un demi siécle à l'époque des premiers Troubadours, que leur historien Jean de Nostredame fixe à l'an 1162, &c."—Pref. a l'Anthologie Franç. 8vo. 1765.
This subject hath been since taken up and prosecuted at length in the Prefaces, &c. to M. Le Grand's Fabliaux ou Contes du XII. & du XIII. Siécle, Paris, 1788, 5 tom. 12mo. who seems pretty clearly to have established the priority and superior excellence of the old rimeurs of the north of France, over the troubadours of Provence, &c.
[S2] [Their own native gleemen or minstrels must be allowed to exist.] Of this we have proof positive in the old metrical romance of Horn-Child, (vol. iii. appendix), which, although from the mention of Sarazens, &c. it must have been written at least after the first Crusade in 1096, yet from its Anglo-Saxon language or idiom, can scarce be dated later than within a century after the Conquest. This, as appears from its very exordium, was intended to be sung to a popular audience, whether it was composed by, or for, a gleeman, or minstrel. But it carries all the internal marks of being the production of such a composer. It appears of genuine English growth, for after a careful examination, I cannot discover any allusion to French or Norman customs, manners, composition or phraseology: no quotation "As the Romance sayth:" not a name or local reference which was likely to occur to a French rimeur. The proper names are all of northern extraction. Child Horn is the son of Allof (i.e. Olaf or Olave), king of Sudenne (I suppose Sweden), by his queen Godylde, or Godylt. Athulf and Fykenyld are the names of subjects. Eylmer or Aylmere is king of Westnesse (a part of Ireland), Rymenyld is his daughter; as Erminyld is of another king Thurstan; whose sons are Athyld and Beryld. Athelbrus is steward of K. Aylmer, &c. &c. All these savour only of a northern origin, and the whole piece is exactly such a performance as one would expect from a gleeman or minstrel of the north of England, who had derived his art and his ideas from his scaldic predecessors there. So that this probably is the original, from which was translated the old French fragment of Dan Horn, in the Harleyan MS. 527, mentioned by Tyrwhitt (Chaucer iv. 68), and by T. Warton (Hist. i. 38), whose extract from Horn-Child is extremely incorrect.
Compare the stile of Child-Horn with the Anglo-Saxon specimens in short verses and rhime, which are assigned to the century succeeding the Conquest, in Hickes's Thesaurus, tom. i. cap. 24, pp. 224 and 231.
[T] [The different production of the sedentary composer and the rambling minstrel.] Among the old metrical romances, a very few are addressed to readers, or mention reading: these appear to have been composed by writers at their desk, and exhibit marks of more elaborate structure and invention. Such is Eglamour of Artas (No. 20, vol. iii. appendix), of which I find in a MS. copy in the Cotton Library, A. 2, folio 3, the II. Fitte thus concludes:
"... thus ferr have I red."
Such is Ipomydon (No. 23, iii. appendix), of which one of the divisions (Sign. E. ii. b. in pr. copy) ends thus:
"Let hym go, God him spede
Tyll efte-soone we of him reed (i.e. read)"