Notation: Entered by Hand D in white mensural notation on four-line staves originally drawn for music in chant notation – staff system 3.
Disposition of parts: [Superius]-[Tenor]-[Bassus] below each other.
*Cambridge 1760 f. 2 »Dulcis amica dei« 3v Prioris PDF
*Copenhagen 1848 p. 430 »Dulcis amica dei« 3v PDF
*Laborde ff. 139v-140 »Dulcis amica dei« 3v PDF · Facsimile
London 31922 ff. 88v-89 »Dulcis amica dei« 3v
London 35087 ff. 61v-62 »Dulcis amica dei« 3v
Attaingnant 1540 no. 1a »Dulcis amica dei« 3v
Four-part versions (different Altus parts are added):
Cambrai 125-128 no. 181 »Dulcis amica dei« 4v (A “si placet”; notated a fourth lower)
*Paris 1597 ff. 4v-5 »Dulcis amica dei« 4v PDF · Facsimile
*Paris 2245 ff. 31v-32 [Without text] 4v (Altus incomplete) PDF · Facsimile
Sankt Gallen 462 p. 9 »Dulcis Maria dei« 4v Facsimile
Sankt Gallen 463 no. 140 »Dulcis amica dei« 4v (S and A only) Facsimile
Ulm 237 f. 41 »Dulcis amica dei« 4v
*Uppsala 76a ff. 55v-56 »Dulcis amica dei« 4v Prioris PDF
Antico 1521 no. 19 »Dulcis amica dei« 4v (A only)
Rhau 1538 no. 3 »Qui credit in Filium« 4v
Chicago Capirola no. 8 (lute)
Attaingnant 1529 no. 6 (lute)
Attaingnant 1531 no. 9 (keyboard)
Editions: Amiens 162 Edition no. 22 (PDF), Rokseth 1930 no. 4 (Laborde); Christoffersen 1994, Vol. I, p. 281 (Copenhagen 1848); Geering 1967 no. 1 (Sankt Gallen 462); Prioris 1982, Vol. III, p. 44 - and many other editions (see Christoffersen 1994, Vol. II, p. 160).
Text: Lauda to Virgin Maria, RH nos. 25737 and 36836.
Rosa vernans, Stella decora,
Tu memor esto mei
dum mortis venerit hora.
The short prayer is set in declamatory lines, articulated by fermatas. The successive entrance of the voices in its opening signals an imitative-polyphonic style, but the dominant feature soon becomes the parallel thirds between the upper voices. The three-part version is effectively entreating and quite elegant in its simplicity. In accordance with the popular lauda-style the upper voices are kept within a very narrow range (the tenor c’-f’) for the first three lines; only in the last line, “when the hour of death arrives”, the full ranges of an octave are explored, and the turn in sound in the flat direction underscores the plea.
The Amiens 162 version is in details and notation very close to the version, which was added to the Laborde chansonnier (ff. 139v-140) some time after c. 1480, or the textless three-part version originally copied into the last blank pages of the French chansonnier in Bibliothèque Nationale, Paris, MS f.fr. 2245, around 1500, or the French MS in Cambridge, Magdalene College, Pepys 1760, from the first decade of the 16th century, which attributes the song to the royal chapel master, Denis Prieur, known as Prioris. All three sources were connected with the French court (see the editions mentioned above, or the comments in Prioris 1982, Vol. III, pp. XI-XIII). In Paris 2245 the same copyist (probably) tried to modernize the setting by adding a contratenor. He composed it on the pages using some ideas that he probably had heard in four-part performances. When he reached bar 19, he had to change the bassus part too, made some errors, and soon abandoned the project.
The song was popular in the last decades of the 15th and the first half of the 16th centuries and enjoyed a wide circulation, often with a fourth part added to fill out the thin texture. However, the many different contratenors in the sources above mostly destroy the elegance of the three-part setting. The chansonnier from around 1500, Paris, Bibliothèque Nationale, MS f.fr. 1597, encumbers the song with a contratenor stretching over nearly two octaves and demonstrates how difficult it was to add to a setting perfect in itself (cf. the edition). A more throughout reworking of all the parts could be much more successful as shown by the publication by Petrucci of »Dulcis amica dei« as a four-part laude in Laude libro secondo, Venezia 1502 f. 20 (edited in Jeppesen 1935 no. 19).
The provincial collection, MS Copenhagen 1848, from Lyons c. 1524, has the song in a differing shape: The superius (prolonged with an extra brevis) starts with the bassus, which then follows the superius in parallel tenths, and the lines are separated by general semibrevis-rests, which cause the second line to be sung off-beat (see the edition). These differences emphasize the declamatory/improvisatory style of the song, and this version presents of all the versions the most natural declamation of the Latin poem. Even if this source is late, it might preserve a very early version of the song. However, another provincial source from the same area, MS Uppsala 76a from the first decade of the 16th century, contains the song in a four-part version, in which the usual start of the upper voice has been changed simply by scratching out the introductory rest and insertion of a brevis a’ as the third note. This is similar to what could have happened to Copenhagen 1848 version, so the standard version probably was the original. The rhythm has been revised in an idiosyncratic way in the Uppsala version. The three first lines all start with two brevis notes, which makes the text underlay difficult and clumsy!
Concerning the identity of Prioris, see the article by Theodor Dumitrescu, ‘Who Was “Prioris”? A Royal Composer Recovered’, Journal of the American Musicological Society 65 (2012), pp. 5-66 (comments on “Dulcis amica dei” pp. 31-36).
See also the comments and lists of related compositions in Christoffersen 1994, Vol. II, p. 160, and Fallows 1999, pp. 580-581 (also this song has been reworked as a basse dance).
PWCH December 2013