We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 12 April 2013
This analytical article aims to establish the melodic grammar of the in directum Milanese cantus. Deviations from the generic norms might be experienced in performance as emphasising the words with which they appear. Certain texts (particularly ‘domin-’) are often associated with melodic shapes found elsewhere in the genre on the same words, regardless of what one might expect to encounter formally. Chants that share similar liturgical functions, or which appear in close liturgical proximity, often share melodic material and formal strategies. Building on and responding to Bailey's The Ambrosian Cantus (1987), the present article seeks to appreciate the cantus melodies as they have been preserved, rather than using them as a lens through which to seek a putative (and entirely rationally structured) lost ancestor.
1 Chant analysis seeking to uncover generic norms was pioneered by scholars such as Frere, W.H., Antiphonale Sarisburiense (London, 1901–24)Google Scholar. For more recent examples see, inter alia, Edward Nowacki, ‘Studies on the Office Antiphons of the Old Roman Manuscripts’, Ph.D. diss., Brandeis University (1980); Hornby, Emma, Gregorian and Old Roman Eighth-Mode Tracts: A Case Study in the Transmission of Western Chant (Aldershot, 2002)Google Scholar; eadem, Medieval Liturgical Chant and Patristic Exegesis (Woodbridge, 2009); Maloy, Rebecca, Inside the Offertory: Aspects of Chronology and Transmission (New York and Oxford, 2010)CrossRefGoogle Scholar; Katherine Helsen, ‘The Great Responsories of the Divine Office: Aspects of Structure and Transmission’, Ph.D. diss., Universität Regensburg (2008). On this genre in particular, see Bailey, Terence, The Ambrosian Cantus, Musicological Studies 47 (Ottawa, 1987)Google Scholar. The present analysis is offered in a spirit of appreciation of Bailey's groundbreaking work and, in particular, of the pattern-matching methodology which, in its level of detail and its rigour, he pioneered for this repertoire.
2 For other Milanese chants labelled ‘cantus’, see Bailey, The Ambrosian Cantus, 16–22.
3 In my examples, melodic gestures associated with accented syllables are indicated by /. To a great extent, the manuscripts preserve an accent sensitivity characteristic of other formulaic chant genres. See, for example, Kainzbauer, Xaver, ‘Der Tractus Tetrardus – eine centologische Untersuchung’, Beiträge zur Gregorianik, 11 (Regensburg, 1991), 1–132Google Scholar; Hornby, Medieval Liturgical Chant, 38–40.
4 Bailey, The Ambrosian Cantus, 124.
5 Hornby, Gregorian and Old Roman Eighth-Mode Tracts, 12–13.
6 Antiphonaire Ambrosien de la British Library, Codex Additional 34209 (XIIe Siècle), Paléographie Musicale 5–6 (Solesmes, 1896 and 1900; reprinted 1996).
7 Bailey, The Ambrosian Cantus. D=Milan, Biblioteca Capitolare del Duomo, F.2.2 (twelfth century); Vh=Vimercate, S. Stefano, ssp B (thirteenth century); MUh=Muggiasca, S. Lorenzo (without number) (1387AD); Mh=a privately owned manuscript of the twelth century.
8 Bailey, The Ambrosian Cantus, 147–8.
9 For a detailed argument in support of the likelihood of progressive stereotyping in the oral transmission of Roman and Franco-Roman chant, see Pfisterer, Andreas, Cantilena romana: Untersuchungen zur Überlieferung des gregorianischen Chorals, Beiträge zur Geschichte der Kirchenmusik 11 (Paderborn, 2002)Google Scholar.
10 Benedictus es has the same melody on both Good Friday and Holy Saturday; on Holy Saturday there are fewer verses (see Table 18, below).
11 On the treatment of ‘anania azari-’, see Ex. 8 and discussion, below.
12 In Ad dominum verse 3, Mh and Vh have an extra Gaba at the end of ‘He-’; in these two manuscripts the melodic outline is thus even closer to that of the cognate phrases. The GFabaG figure in the melisma on ‘e-’ in Nisi verse 2 is mirrored exactly in Ad dominum verse 3 in Vh, where the other three manuscripts instead have GFGabaG. The cbca figure at the beginning of ‘bo-’ in Ecce verse 1 is mirrored exactly in Ad dominum verse 3 in L and Mh, but D and Vh instead have cbcba in Ad dominum verse 3. In Benedicam, ‘-cam’ has Gaba rather than aba in Mh, MUh and Vh, making the melodic shape identical to that of Nisi verse 2.
13 In Nisi v3, while ‘-ri-’ has cca in L, it has cba in Vh and Mh, and bcba in D; in the latter two shapes the profile is closer to the common shape marked ℑ in the example. Similarly, in Levavi verse 2, L has ca on ‘-li-’, but Mh, D and Vh all have cba. The shape bcba appears as bccba on ‘-tum’ in Ecce verse 2 in all the manuscripts except L. In the same phrase, the a found as the eighth note in the melisma on ‘-te’ is found only in L; the other manuscripts all have the same melodic shape as the other examples of phrase O2a and O2b.
14 The Roman and Franco-Roman second-mode tract phrase 1g has a parallel function as a second verse-opening phrase. See Hornby, Medieval Liturgical Chant, 49–50.
15 On the tendency for chants to take on characteristics of other chants sung in close proximity, see Nowacki, Edward, ‘Gregorian Office Antiphons and the Comparative Method’, Journal of Musicology, 4 (1985), 243–75CrossRefGoogle Scholar. See also Hornby, Emma and Maloy, Rebecca, Music and Meaning in Old Hispanic Lent: Psalmi, Threni and the Easter Vigil canticles (Woodbridge, 2013), chapter 3Google Scholar.
16 Bailey, The Ambrosian Cantus, 82.
17 Bailey writes that he cannot account for the existence of this phrase shape (The Ambrosian Cantus, 82). On the second-mode tract phrase 1c, see Hornby, Medieval Liturgical Chant, 61.
18 In Cantemus verse 2, ‘-cen-’ has Ga in MUh, D and Vh.
19 Although Super flumina has ababaG on ‘-nis’ in L, all the other three manuscripts in Bailey's edition have abaG, such as Benedictus es verse 12.
20 The G D G opening M2a in In convertendo and M2b in Supra verse 1 and Ad dominum verse 1; the Gabc(b)a opening M2a in Nisi verse 1 and M2b in Conserva me and Qui regis.
21 In L, Qui edebat is differentiated from the other phrases by having Gba rather than Gaba on ‘-de-’; the other three manuscripts have Gaba as in the other examples of phrase M2a.
22 Bailey, The Ambrosian Cantus, 94.
23 The E towards the end of the melisma is instead a D in Nisi verse 2 in Mh and Vh; a D in Ad dominum verse 2 in Mh, Vh and D; and a D in Levavi verse 2 in Mh, D and Vh. It is very possible that in fact there was a uniform melodic outline here with the low note notated casually – it is very common in early notations for an isolated low note to be treated without regard for accurate diastemy.
24 The E towards the end of the melisma is instead a D in all three sources apart from L in Ad dominum verse 3, Nisi verse 3 and Levavi verse 3. In Ecce verse 3, only manuscript D has D rather than E. As with similar occasions, the differences may be the result of notational casualness rather than differentiated melodic shapes. In Ecce verse 3, ‘-o-’ has bc in L but cc in Vh and D, and c in Muh.
25 Bailey, The Ambrosian Cantus, 98.
26 In D and Vh, the opening of Qui regis is not the FaG found in L. Rather, these manuscripts have FabaaG, like the cognate phrases in other chants.
27 The a–c–a near the end of Nisi verse 3 and Ad dominum verse 3 is filled in abcba in all manuscripts except L; abcba is more standardised. The melisma begins bb but b in Nisi verse 3 in Mh and D, in Levavi verse 2 in L and D, in Ecce v2 in MUh and D, and in Ecce verse 1 in D. The repeated c in the melisma in Levavi verse 3 is found only in L; the other manuscripts have a single c, as usual in phrase S3.
28 Bailey, The Ambrosian Cantus, 112.
29 The bracketed G in Domine exaudi is not present in L, but is present in all three other manuscripts.
30 Bailey, The Ambrosian Cantus, 62.
31 In Nisi verse 1, the E near the end is instead a D in Mh and Vh; in In convertendo this note is an E only in Vh, otherwise it is a D; in Sicut, L does not have this note, D has an E, and MUh and Vh have an E. As discussed above, such variants may be notational rather than performed. The ab at the end of ‘a-’ are present only in L and Vh.
32 The falling c–a on ‘u-’ in Ecce verse 1 is filled in as cba in the other three manuscripts; similarly cca on ‘mi-’ in Levavi verse 1 is instead cba in the other manuscripts.
33 The falling c–a on ‘ter-’ in Nisi verse 3, on ‘-cu-’ in Levavi verse 3 and on ‘cu-’ in Ecce verse 3 is filled in as cba in the other manscripts.
34 Bailey, The Ambrosian Cantus, 109–10, 116.
35 Nisi verse 1 and Ad dominum verse 1 (Ex. 3); Cantemus verses 5 and 6 and phrase O1b, ‘do-’ (compare Exx. 3 and 4); Levavi verse 3 and Supra dorsum verse 2 (Ex. 5); Ad dominum verse 2 (compare Exx. 5 and 6); Conserva and Laudate (Ex. 37, below); Cantemus verses 5 and 6, and Benedicam (Ex. 4). In Example 4, the Benedicam ending bG G is compatible with proparoxytonic cadences elsewhere (for example, in phrase O1a); the ba G ending of Cantemus verses 5 and 6 is not. This suggests that the textual parallel has been followed without Cantemus being fully integrated into the normal musical grammar of the genre.
36 On cognate texts sharing melodic identity, see, inter alia, Hornby, Gregorian and Old Roman Eighth-Mode Tracts, 146–8, and Hornby, Medieval Liturgical Chant, 73–6.
37 Bailey, The Ambrosian Cantus, 73.
38 See Hornby and Maloy, Music and Meaning, chapter 4.
39 Bailey, The Ambrosian Cantus, 103.
40 aca in the melisma is filled in abcba in Vh and D, and acba in Mh; Mh, D and Vh are closer to the S3 melisma than L in Nisi verse 1.
41 On Augustine's interpretation of this text, in which the ‘snare’ is the sweetness of earthly life, which God helps martyrs to escape, see Hornby, Medieval Liturgical Chant, 93.
42 Bailey, The Ambrosian Cantus, 92.
43 The bcba shape in the melisma (which does not directly parallel S3 or E2a) is not present in MUh.
44 As suggested by Bailey, The Ambrosian Cantus, 79.
45 Bailey, The Ambrosian Cantus, 125–6, hypothesises a connection with the practices of eleventh-century liturgical dramas. Much work has been done on the potential of early chant melodies to ‘read’ their texts in rhetorical ways since The Ambrosian Cantus was written in the 1980s. For a recent summary of the literature, see Hornby, Medieval Liturgical Chant, ch. 4.
46 On the potentially emphatic role of incomplete formulaic cadences in chant, see my ‘Preliminary thoughts on silence in western plainchant’, in Silence, Music, Silent Music, ed. Losseff, Nicky and Doctor, Jenny (London, 2007), 141–54Google Scholar.
47 Bailey, The Ambrosian Cantus, 92.
48 The GGFG within the melisma on ‘-dum’ is instead GGEG in Vh. The E in the same melisma is instead D in Vh and D.
49 Bailey, The Ambrosian Cantus, 114.
50 In Laudate verse 2, ‘-di-’ has bcba in manuscript D.
51 Bailey, The Ambrosian Cantus, 123.
52 See Hornby and Maloy, Music and Meaning, ch. 4.
53 In Conserva me, the second a on ‘-ser-’ is present only in L; the other three manuscripts have a phrase opening closer to Laudate. In Conserva me, ‘do-’ has abcc in all three manuscripts apart from L, which has abc.
54 ‘-cam’ is Gaba in Mh, MUh and Vh.
55 Noted by Bailey, The Ambrosian Cantus, 60.
56 Bailey sees this in chronological terms, viewing this as a late chant (The Ambrosian Cantus, 91).
57 Bailey, The Ambrosian Cantus, 83.
58 See Hornby and Maloy, Music and Meaning, ch. 5.
59 See Hornby and Maloy, Music and Meaning, ch. 5.
60 Bailey, The Ambrosian Cantus, 127, quoting Apel.
61 Bailey, The Ambrosian Cantus, 85.
62 Ibid., 127.
63 Ibid., 110.
64 Antiphonale Missarum; see Bailey, The Ambrosian Cantus, 83.
65 See Hornby and Maloy, Music and Meaning, ch. 5.
66 GF and EV show whether the verse is present in the Good Friday (GF) or Easter Vigil (EV) chant.
67 See Hornby, Medieval Liturgical Chant, 73–8; Hornby and Maloy, Music and Meaning, ch. 4.
68 See note 15, above.
69 Bailey, The Ambrosian Cantus, 52–4.
