¹ Eighth-mode tracts are transmitted in four sources of the Old Roman manuscript tradition: Vatican City, Biblioteca Apostolica Vaticana, MSS Vat. lat. 5319, San Pietro f22 and San Pietro F 11 (De profundis only), and Cologny bei Genf, Martin Bodmer Collection, MS 74 (olim Phillips 16069). They are catalogued and indexed to their respective folio numbers in Cutter, P. F., Musical Sources of the Old-Roman Mass, Musicological Studies and Documents 36 (Neuhausen-Stuttgart, 1979).Google Scholar The folio numbers for the examples in the Bodmer–Phillips manuscript, not given by Cutter, are listed here: Qui regis israhel, fol. 8^V; Qui seminant, fol. 31^v; Desiderium, fol. 32; Beatus vir, fol. 32^v; De profundis, fol. 34; Commovisti, fol. 35^v; Iubilate deo, fol. 36^v; Laudate dominum, fol. 47^v; Ad te levavi, fol. 52; Qui confidunt, fol. 57^v; Sepe expugnaverunt, fol. 63^v. Space limitations prevent me from illustrating all of the analytic observations made in the article. Readers are referred to the complete edition of the Old Roman gradual by Stäblein, B., Die Gesänge des altrömischen Graduale Vat. lat. 5319, Monumenta Monodica Medii Aevi 2 (Kassel, 1970).Google Scholar The musical examples in the article are my own transcriptions from Vat. lat. 5319.

² Johner, D., Wort und Ton im Choral, 2nd edn (Leipzig, 1953).Google Scholar

³ Klauser, T., ‘Die liturgischen Austauschbeziehungen zwischen der römischen und der fränkisch-deutschen Kirche vom achten bis zum elften Jahrhundert’, Historisches Jahrbuch, 53 (1933), pp. 169–89;Google Scholar reprinted in Klauser, , Gesammelte Arbeiten, ed. Dassmann, E. (Münster, Westphalia, 1974), pp. 139–54.Google ScholarHucke, H., ‘Die Einfuhrung des gregoriani-schen Gesanges im Frankenreich’, Römische Quartalschrift fur christliche Altertumskunde und Kirchengeschichte, 49 (1954), pp. 172–85;Google Scholaridem, ‘GregorianischerGesanginaltroüischerund fränkischer Überlieferung’, Archiv für Musikwissenschaft, 12 (1955), pp. 74–87.Google ScholarConnolly, T.H., ‘Introits and Archetypes: Some Archaisms of the Old Roman Chant’, Journal of the American Musicological Society, 25 (1972), pp. 157–74.CrossRefGoogle Scholar

⁴ The Old Roman tradition lacks versions in the native style of the four tracts for the Easter vigil: Cantemus domino, Vinea facta est, Attends caelum and Sicut cervus. The versions of these tracts given in the Old Roman manuscripts are Frankish (i.e. standard Gregorian).

⁵ The rules of accentuation in standard Gregorian psalmody, as described in the Antiphonale monasticum (Tournai, 1934), p. 1225Google Scholar, treat stressed final monosyllables as unaccented:

Aliter diceres: monosyllabica vox in ultimo loco posita semper accentu privatur. Unde cum haec spondeum sequitur, ambo simul efficere putantur dactylum: v.g. Fáctus mm = Dóminus; cum autem dactylum sequitur, ambo simul efficere censentur di-spondeum: v.g. génuí te = Déus méus.

The Old Roman example is consistent rather with the practice of Dominican psalmody, as set forth in the Processionarium iuxta ritum S. Ordinis Praedicatorum (Rome, 1949)Google Scholar, which claims to transmit thirteenth-century usage. For a full discussion of this entire matter, see Chen, M. Y., ‘Toward a Grammar of Singing: Tune–Text Association in Gregorian Chant’, Music Perception, 1 (1983), pp. 84–122.CrossRefGoogle Scholar

⁶ Primary and secondary stress have been determined by the trisyllable rule: all polysyllables are proparoxytone if the penult is light or weak; otherwise they are paroxytone. In longer words, secondary stress is placed on all even-numbered syllables preceding the primary stress (Chen, ‘Toward a Grammar of Singing’, p. 89). Atonic polysyllables (prepositions, particles and true copulative conjunctions), exemplifying Chen's Rule 4 (ibid., p. 93), are declaimed as if they bore normal stress. The same rule specifies that monosyllables are declaimed as stressed or unstressed according to phrasal patterns. An example of the former is nos in line i; an example of the latter is te in line v.

⁷ Again, the Old Roman practice is consistent with thirteenth-century Dominican usage as transmitted in the modern liturgical books of the Dominican rite, rather than with the modern Gregorian practice described in the Antiphonale monasticum, p. 1225. See Chen, ‘Toward a Grammar of Singing’, p. 108.

⁸ The agreement of the three sources in the transmission of this unique melisma adds further weight to the speculation that the copyists had a written tradition to which they could refer in non-standard cases. For examples of phrase K₁, to which the passage in question is compared, see the Stiiblein edition, p. 236 on the words mittentes semina sua, p. 237 on dominorum suorum, p. 238 on domine sue and p. 240 on in hierusahm.

⁹ The tracts discussed in this essay are numbered i to xi; the complete list is given below at the beginning of part ii.

¹⁰ Johner, E.g., Wort und Ton im Choral, pp. 435–6.Google Scholar See also Ferretti, P., Estetica gregoriana (1934; repr. New York, 1977), p. 104.Google Scholar

¹¹ The examples may be found in the Stäblein edition, pp. 236–40. The passage in question, treated here as a constituent of a larger I phrase, is identical with the motif that elsewhere forms the core of the phrase labelled R. Although contextual differences may justify the inconsistency of labelling, a more thorough study of the phrase, and of the composition of I phrases generally, may yet yield a more elegant, less ad hoc solution to this analytical problem. However, since the passage occurs only five times in the whole corpus of eighth-mode tracts, the resolution of its status will have no great effect on the general outcome.

¹² The notion of contrastive variation presented here is related in important ways to the linguistic concept of the phoneme. For a useful general introduction to this subject, see Fromkin, V. and Rodman, R., An Introduction to Language, 2nd edn (New York, 1978), pp. 101–137.Google Scholar

More specific parallels to the phenomenon of contrast in plainchant may be found discussed in A. B. Lord's work on the oral epic, in which supposed departures from the expected prosody and vocabulary are shown to coincide with subtle disturbances in the poetic environment. Lord observed that singers, once they had solved a particular problem in verse-making, generally did not seek additional ways of expressing the same essential idea, so long as the poetic environment did not motivate them to do so. See The Singer of Tales (Cambridge, Mass., 1960), pp. 50–3.Google Scholar

¹³ Liber usualis (Tournai, 1953), p. 351.Google Scholar

¹⁴ The example may be found in the Stäblein edition, p. 236; see also p. 240 on the words non commovebitur. For further discussion of the R phrase, see n. 11.

¹⁵ Johner, , Wort und Ton im Choral, pp. 429–32.Google Scholar

¹⁶ Ibid., p. 213.

¹⁷ ‘Das Erhabene posaunt sich nicht aus’. Ibid., p. 430.