No CrossRef data available.
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.
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 September 2008
It is well known that the two alleluias cited in the title are textings of the same tune – indeed they are only two of many textings of that tune. Along with the alleluia Video celos, these two were sung over a rather short space of time in the liturgical year: on the feasts of Christmas, St Stephen's and Epiphany. It happens that for another purpose, involving the transmission of chant in later manuscripts, I have collated the alleluia Dies sanctificatus and alleluia Vidimus stellam in over a hundred manuscripts, most of them late French sources. The choice of these manuscripts rather than others – French rather than German or English – was dictated in part by practical considerations, but also by the conviction that French sources are on the whole richer in variants and hence more revealing in the mechanics of their transmission. The approximately one hundred sources considered are a good majority of the surviving late sources, and the picture they give is, I believe, representative. In the course of this article, I will also have occasion to mention earlier manuscripts, both French and non-French. These are not the primary focus of my study, and therefore I have permitted myself many omissions in that respect. Important graduals, for example from St-Omer,2 Angers3 and Noyon,4 to name only three, have been untouched. My collation has shown a considerable number of variants, some of which are clearly regional in nature, while others are of a more ambiguous character. Having two textings of the same melody available supplies a useful control: readings that occur in both versions naturally have a strong claim to represent a true tradition, rather than a random alteration.
1 I should like to acknowledge here the very considerable assistance of Dr Thomas Davies Kozachek in the preparation and execution of this paper, including the preparation of the musical examples.
2 St-Omer, Bibliothèque municipale, 252.
3 Angers, Bibliothèque municipale, 91
4 London, British Library, Egerton 857.
5 Facsimile edition in Antiphonale tonale missarum, Xle siècle: Codex H.159 de l'Ecole de Médecine de Montpellier, Paléographie Musicale 8 (1901–05). The melody, with the Dies text, may be found beginning on p. 100 of the facsimile, and continuing on p. 103. The end of the verse is reduced to an unnotated cue, ‘quia hodie’, but since the music then repeats previously heard material through ‘lux magna’, it can be safely reconstructed through that point. The final phrase, ‘super terram’, is the published Solesmes version, which has near-unanimous support from the earliest manuscripts.The other textings of the melody are not present in the Montpellier manuscript.
6 I preserve the word ‘columns’ here, as in my ‘Evidence for the Traditional View of the History of Gregorian Chant’, Journal of the American Musicological Society, 40 (1987), 387Google Scholar, even though the columnar appearance found in the examples there is not present here. The term is preferable to ‘measure’ or ‘section’, and corresponds to the way in which the collation was done. In the full version of my own collation below the Dijon reading for each column stand the readings of the hundred or so manuscripts listed here, as well as a number of others.
7 This information has been taken over for the most part without change from Le graduel Romain: Edition critique par les moines de Solesmes. II: Les sources (Solesmes, 1957), and from Leroquais, Victor, Les sacramentaires et les missels manuscrits des bibliothèques publiques de France, 3 vols. (Paris, 1924)Google Scholar. Exceptions to this rule are noted as they occur.
8 E.g., Call, which is probably from Auberive, near Langres, but seems to owe little to Langres and more to its Cistercian origin.
9 On the reform of St-Corneille by monks of St-Denis see Robertson, Anne Walters, ‘The Transmission of Music and Liturgy from Saint-Denis to Saint-Corneille of Compiègne’, in Trasmissione e recezione delle forme di cultura musicale, Atti del XIV Congresso della Società Internazionale di Musicologia, Bologna, 27 Agosto-1 Settembre 1987, vol. III, Free Papers (Bologna, 1990), 505–14Google Scholar, and her ‘Benedicamus Domino: the Unwritten Tradition’, Journal of the American Musicological Society, 41 (1988), 17Google Scholar. Of the Compiègne manuscripts (sigla Cocl, Coc2, Coc3, Coc4), all but the first postdate the reform, which began in 1150. Cocl is generally dated to the middle of the twelfth century, but its content can be shown to be pre-reform in many ways, hence the appropriateness of the ‘before 1150’ dating.
10 This is not a wholly arbitrary assignment. While Les sources gives only ‘nord de la France’ with an inclination in the direction of Tournai (p. 28), its behaviour in what follows suggests rather a connection with the St-Denis/Compiègne group.
11 Evr2 was made in Paris for the use of Evreux and employs the Parisian melodic tradition throughout.
12 Like Evr2, Rog4 is a product of Paris, and uses the Parisian melodies, although it was made for the Norman town of Eu. See Hiley, David, ‘The Norman Chant Traditions – Normandy, Britain, Sicily’, Proceedings of the Royal Musical Association, 107 (1980–1981), 17CrossRefGoogle Scholar.
13 The Graduate triplex (Solesmes, 1979), pp. 49, 58Google Scholar, gives the readings of Laon 239 and St-Gall 359. I have consulted later St-Gall sources as well.
14 Antiphonale Missarum Sancti Gregorii, Xe Siècle: Codex 47 de la Bibliothèque de Chartres, Paléographie Musicale 11 (Solesmes, 1912), 107Google Scholar.
15 Le Codex 903 de la Bibliothèque Nationale de Paris, Graduel de Saint-Yrieix (Xle S.), Paléographie Musicale 13 (Solesmes, 1925–1930), 19.Google Scholar
16 The version of St-Yrieix (along with other Aquitanian manuscripts) has one small variant: in Dies, the pes on ‘ve-m-te’ (col. 42) is replaced by a single note.
17 See also the discussion of this cadence at Example 11 below.
18 See also the discussion at Example 10 below.
19 The neumes of ‘dies’ are tilted sharply upwards, so that the first punctum of the climacus is actually above the preceding virga, but since all of the textings have a compatible version it seems unlikely that a different meaning is intended.
20 On the problems posed by the trigon, see my ‘An Enigmatic Neume’, Themes and Variations: Writings on Music in Honor of Rulan Chao Pian, ed. Yung, Bell and Lam, Joseph S. C. (Cambridge, Mass., 1994), 8–30Google Scholar.
21 In fairness to the scribes it should be noted that errors were doubtless made during the various stages of my recording of the data. As far as I can determine, these are not excessively numerous, and they are spread fairly randomly among the sources. Such errors inflate the quantity of noise without materially affecting its nature.
22 See my ‘Parisian Sanctorals of the Late Middle Ages’, to be published in a Festschrift (1998) honoring Terence Bailey, edited by Bryan Gillingham and Paul Merkley. One may contrast this carelessness with the accuracy shown by the earlier sources. The pre-diastematic sources may conceivably conceal all sorts of pitch variants that are no longer recoverable, but there is no reason to suppose that they really do.
23 This poses a certain logical problem, to be sure. In the example just given, the improbable collocation of Aquitanian with a few northern sources is duplicated from Dies to Vidimus. Does this not show that the variant is not ‘random’ after all? In one sense, perhaps. This was notmerely a mistake, but rather an intended reading. Its origin, however, remains random (at least given our present state of knowledge). That it occurs in some but not all Aquitanian manuscripts, and in some but not all manuscripts from Chartres and from Langres, implies that it is somewhat below the level of a true local tradition, while still remaining above the level of error. The matter is not an easy one.
24 Circumstances that might have caused such partially local uses are easy to imagine: those that in fact did cause them are probably unknowable. In a roughly similar context, Max Haas puts it thus: ‘Die auf Handschriften gesrützte Beschäftigung mit liturgischer Einstimmigkeit des Mittelalters fuhrt mit der Zeit zu einer Reihe von Vermutungen, die sich durch Belege weder srützen noch widerlegen lassen. Dennoch drangen sie sich auf, da die Textwirklichkeit nicht ausreicht, sich die Lebensformen derer, die Texte oder Noten geschrieben haben, vorzustellen.’ (Mündliche Überlieferung und altrömischer Choral: Historische und analytische computergestutze Untersuchungen [Bern, 1997], 40.) In the case of the ‘sub-local’ readings, one might imagine a cantor or a book imported from another area, or a perhaps older singer who resisted modern ways of singing. These and other hypotheses are, as Haas reminds us, neither confirmable nor disprovable.
25 Not necessarily a written text, of course. My argument here is based on the relative unanimity of early sources (and hence, inferentially, of the early oral tradition) and the lack of unanimity later on. While the written evidence may point to the tenth or eleventh centuries as the transition between the two phases, any given variant may of course have entered part of the tradition at virtually any time and place. The phrase ‘established text’ is also problematic when applied to pre-literate musics. As Haas observes, a singer lacking any means of representing music visually may well have had notions of ‘same’ and ‘different’ that are not the same as ours (Mündliche Überlieferung, 12).
26 As observed by Stäblein, Bruno, ‘Graduale (Gesang)’, in Die Musik in Geschichte und Gegenwart, 5, col. 651.Google Scholar
27 It could be argued, at least provisionally, that chant is not ‘syntactic’ in the same way that Mozart is. Thus the very ending – in tonally syntactic music generally the most tightly controlled part of the composition – may well have been relatively free.
28 In this and subsequent examples, the top line (version a) is the reading of Dijon. The subsequent line or lines (versions b, c, etc.) show the variant(s) for the manuscripts listed.
29 For a full presentation of this structure as found in a number of verses, see Ferretti, Paolo, Esthétique grégorienne, trans. Agaësse, A. (Solesmes, 1938), pp. 187–90Google Scholar.
30 I assume here with everyone else that DE is the earlier formulation, although there is perhaps no absolute proof. The issue of E versus F in such contexts is carefully studied in Gajard, Joseph, ‘Les Récitations modales des 3e et 4e modes dans les manuscrits bénéventains et aquitains’, Etudes grégoriennes, 1 (1954)Google Scholar, passim. The monks of Solesmes dismiss this type of variant as irrelevant, as indeed it is to their purposes; see Le graduel romain, IV: le texte neumatique, 1, Le groupment des sources (Solesmes, 1960–1962), 22–3.Google Scholar
31 In fact Rog4 has the variant in Vidimus, but as we have seen, Rog4 is not from Rouen but rather from Paris for the use of Eu. The behaviour of Rog4 with respect to this variant is another small piece of evidence showing that it did not originate in Rouen.
32 This is not always the case, however. In column 5, Car6, Car8, Nim and Val (the eastern-most Aquitanian sources) join in a curious reading – GFED for the majority GEED – while a mixed group of nearly twenty manuscripts has GEFD. Here these four Aquitanian sources in effect replace an E with an F – the reverse of what they are supposed to do. It is not at all clear what has happened here, but the association of Car6, Car8, Nim and Val into a subgroup (‘eastern Aquitanian’) is plausible enough.
33 Naturally the pitches of Denl and Den4 are indeterminate, but as we have seen above, the neumes – clivis, climacus – support the reading GE FED ED in light of the fact that the later Sandionysian manuscripts have this version as well.
34 Both Cocl and Nof (which may not have been written at Compiègne itself) preserve pre-reform Compiègne usage, demonstrating that the connection of Otul is directly with Compiègne itself and not via the Sandionysian reform.
35 Denl has a pes preceding the strophae, possibly representing CD.
36 In fact, Rem7 shows a cadence DEED – the only manuscript to do so – and it retains that reading in all textings. Inaccurate writing is not a possible explanation here. I have grouped it with the other Reims sources because Rem7 is closer to them than to any other sources; but it would be unwise to make it a full and unquestioned member of the Reims group on this evidence. Still, given the reluctance of Reims to raise the E to F elsewhere, one might imagine that Rem7 preserves an older usage.
37 Of the others, only three – Rogl, Juml and Rog5 (also from Jumièges) show any kind of geographic convergence.
38 Note that Rog4 agrees here with the other sources from Rouen. Elsewhere it differs from these sufficiently to make a different (e.g., Parisian) origin wholly plausible.
39 Thus in column 40 Cha2 and Perl agree in an otherwise unknown reading. The third manuscript from Chartres, Prol, does not have this, but has a reading in column 39 that gives a closely similar effect to the passage as a whole.
40 Lor is definitely ascribed to St-Laurent-de-Longré, near Pouilly-sur-Loire in the diocese of Auxerre (Les sources, 97). With regard to Cyr ‘il s'agit peut-être de Saint-Laurent-de-Longré (?)’. (ibid., 104).
41 And it is contradicted by both the Paschal and post-Pentecostal alleluia series, which differ widely between the two manuscripts. The alleluias of Cyr are closest to those of Far; and those of Lor are not really close to those of any other source considered here, although there is some similarity to Auxerre, as well as to Sens and to Chartres.
42 Note that this implies multiple readings at least some of the time, while my experience is that more than three or so readings at any one point is somewhat unusual.
43 The problem of the purity of local traditions is a difficult one. Most scholars posit the existence of local traditions changing over time (e.g., Robertson, Anne Walters, The Service-Books of the Royal Abbey of Saint-Denis [Oxford, 1991], 102–5)Google Scholar. But a comprehensive examination of how local traditions actually work – particularly after the eleventh century – seems not to have been attempted as yet.
44 For a carefully reasoned and well-balanced discussion of different methodologies, see volumes 5 and 6 of Hesbert, René-Jean, Corpus antiphonalium officii, Rerum ecclesiasricarum documenta, Series major: Fontes 7–12 (Rome, 1963–1979), passimGoogle Scholar. Virtually each of Hesber's steps in constructing this work is preceded by an illuminating explanation.