¹ The St Cecilia gradual, now Cologny-Geneva, Bibliotheca Bodmeriana, Codex 74 (hereafter Bodmer 74), was written at and for the basilica of Santa Cecilia in Trastevere. Max, Lütolf, editor of the facsimile edition, Das Graduale von Santa Cecilia in Trastevere, Cod. Bodmer 74 (Cologny-Geneva, 1987)Google Scholar, aptly remarks, ‘nach der heutigen Quellenlage setzt mit dieser Handschrift die musikalische Überlieferung des sogannten “altrömisehen” Chorals ein’, but continues, ‘Von der Frage der altrömischen Tradition … abgesehen, stellt C 74, die als die älteste heute bekannte Musikhandschrift römischer Provenienz überhaupt gelten darf, auch für die Rezeption der sogenannten gregorianischen Überlieferung sowie für die Aufnahme einer beträchtlichen Zahl an Stücken aus dem Tropen und Sequenzenschatz des 9.-11. Jahrhunderts in Rom eine bemerkenswerte Quelle dar’ (the musical tradition of the so-called “old Roman” chant begins with this manuscript, according to [our] present-day sources… [Bodmer 74], which may be regarded as the oldest music manuscript of Roman provenance known today, presents a noteworthy source for the introduction to Rome of the so-called Gregorian tradition, as well as for the acceptance [there] of a considerable number of pieces from the repertory of tropes and sequences [dating] from the ninth to the eleventh centuries – quite apart from [its importance for] the question of the old Roman tradition: i, p. 14). It is to the phrase I have italicised that this paper is addressed.

² ‘Chant Notation in Eleventh-Century Roman Manuscripts’, Essays on Medieval Music in Honor of David G. Hughes, ed. Boone, G. M. (Cambridge, Mass., 1995), pp. 43–57Google Scholar. I read a shortened version of the present article entitled ‘Chant in Rome’ (which forms a sequel to the essay for David Hughes just cited) in September 1993 at a Colloque de Royaumont organised and chaired by Thomas Forrest Kelly.

³ Chant notation is ignored by most palaeographers, by almost all art historians and by some liturgical scholars. That is perhaps to be expected; it is more regrettable that some paleographers and art critics disregard or misrepresent the liturgical content and destination of the manuscripts whose script and illumination they discuss.

⁴ Martini, P. Supino, ‘Carolina romana e minuscola romanesca, Appunti per una storia della scrittura latina in Roma tra IX et XII secolo’, Studi Medievali, ser. 3, 15 (1974), pp. 769–93Google Scholar, and especially Roma e l'area grafica romanesca, secoli x-xii (Alessandria, 1987)Google Scholar – the indispensable guide to the existing codices and their scripts, with an extended bibliography and many plates. The user must be cautioned that the author's paleographical expertise is not matched by her liturgical knowledge. The manuscript BAV Archivio San Pietro B 79, for instance, is identified as a gradual instead of an Office antiphoner (p. 318 of the second citation above). Among the earlier and often contradictory studies of the scripts, Lindsay's, W. M. essay ‘The Farfa-Type’, Palaeographia Latina, 6 vols. (London, 1992–1999; rep. 1 vol., Hildesheim, 1974), part 3, pp. 49–51, was especially influential in stimulating further researchGoogle Scholar.

⁵ Berg, K., ‘Notes on the Dates of Some Early Giant Bibles’, Institutum Romanum Norvegiae. Acta ad Archeologiam et Artium Historiam Pertinentia, 2 (Rome, 1965), p. 169Google Scholar.

⁶ Garrison, E. B., Studies in the History of Mediaeval Italian Painting, 4 vols. (Florence, 1953–1962)Google Scholar. Also see revised extracts from Garrison's essays, published after his death in 1981 as Early Italian Painting: Selected Studies (London, 1984), vol. II, ManuscriptsGoogle Scholar. His successor, the Norwegian Knut Berg, while using Garrison's work and employing his terminology, corrected and amplified Garrison's conclusions. See Berg's, ‘Notes on the Dates of Some Early Giant Bibles’ and his Studies in Tuscan Twelfth-Century Illumination (Oslo, Bergen, Tromsø, 1968)Google Scholar.

⁷ For instance, ‘Saints Equizio, Onorato, and Libertino in Eleventh- and Twelfth-Century Italian Litanies as Clues to the Attribution of Manuscripts’, Early Italian Painting, II, pp. 367–85.

⁸ I am not convinced that Ottonian art is the chief or only source the style derives from. The style is found in Lombard art and artefacts and throughout the early Germanic word. (Cf. the astonishingly similar zoomorphic character of the decorated doorways of Norwegian stave churches, dating from the eleventh to the thirteenth centuries, which happen to be reproduced in plates 1–11 in Lean, M. B., ‘The Romanesque Dragon-Doorways of the Norwegian Stave Churches’, in Institutum Romanum Norvegiae. Acta…, 2 (Rome, 1965), pp. 177–94Google Scholar, adjacent to the article by Berg cited in note 5 above, or the insular decoration of the famous Lindisfarne Gospels – to name but two instances.) Zoomorphic interlace is ubiquitous from the seventh to the eleventh centuries in the decorative arts of western Europe.

⁹ Colour facsimiles of these initials can easily be examined in Lütolf, , Das Graduale von Santa Cecilia in Trastevere, I, plates la and lb, 2, 3 and 4, pp. 32–5Google Scholar, and in Stäblein, B. and Landwehr-Melnicki, M., Die Gesänge des altrömischen Graduale Vat. lat. 5319, Monumenta Monodica Medii Aevi 2 (Kassel, 1970), p. xiGoogle Scholar.

¹⁰ Closer study of the script, initials and notation of urban manuscripts from the eleventh century should help identify particular Roman scriptoria, about which very little is known. For instance, the similarity of the zoomorphic interlaced initials, even including the character of the red background dots, in Vat 5319 and in BAV, Vaticanus MS lat. 1274 (a lectionary for special feasts from the monastery of SS. Andrew and Gregory, Clivo Scauro) would suggest that the two books came from the same scriptorium: but from what scriptorium? Perhaps the groups of manuscripts known to have been written at Santa Cecilia in Trastevere would offer a starting point for such a study. (On the products of the Santa Cecilia scriptorium, see Berg, ‘Notes on the Dates of Some Early Giant Bibles’, pp. 167–73, where some of Garrison's conclusions are corrected.)

¹¹ Le Liber Pontificalis, 3 vols., ed. Duchesne, L., Bibliothèques des Ecoles Françaises d'Athènes et de Rome, 2nd ser., 3 (Paris, 1886; repr. Paris, 1955–57)Google Scholar. Most of Duchesne's later essays were published in the series Mélanges d'Archéologie et d'Histoire [de 1'] Ecole Française de Rome, of which he was for many years the editor. A selection was published posthumously in Scripta Minora. Études de topographie romaine et de géographie ecclésiastique, Collection de l'Ecole Française de Rome 13 (1973).

¹² Guy, Ferrari, Early Roman Monasteries. Notes for the History of Monasteries at Rome from the V through the X Century, Studi de Antichità Cristiana 23 (Vatican City, 1957)Google Scholar. Stäblein, B. summarised Ferrari's study in his introduction to Die Gesänge des altrömischen Graduale Vat. lat. 5319, pp. 45*–47*Google Scholar, but in certain respects tendentiously or inaccurately. At the time of their foundation in the sixth to the eighth centuries, the basilican monasteries were not responsible for ‘das tägliche Messopfer’: indeed, Mass may not yet have been offered daily in most basilicas, and in any case most monks and some abbots were not in orders; the monasteries were still chiefly lay institutions. See note 15 below.

¹³ Hamilton, B., ‘The Monastic Revival in Tenth-Century Rome’, Studia Monastica, 4 (1962), pp. 35–68Google Scholar. Also see Antonelli, G., ‘L'Opera di Odone di Cluny in Italia’, Benedictina, 4 (1950), pp. 19–40Google Scholar. The more recent study by Sansterre, J. M., Les moines grecs et orientaux à Rome aux époques byzantine et carolingienne (milieu du VI^e–fin du IX^e s.), 2 vols., Académie Royale de Belgique, Mémoires de la Classe des Lettres, 2nd ser., 66, fasc. 1 (Brussels, 1983; repr. 1993)Google Scholar, draws on sources not previously utilised (such as signatories to the acts of church councils) citing and comparing the views of the authors mentioned above and of many others. This illuminating study was not available to me when my paper was being prepared. Sansterre's conclusions do not however alter the brief summary presented here.

¹⁴ An exception: Leo IV (847–55) gave the monastery of St Lawrence outside the walls to Greek monks, with the obligation of singing the [Greek] office in the basilica. Leo IV was himself not Greek but Roman, trained in St Martin's monastery at the Vatican.

¹⁵ Hadrian I (c. 772–95), having restored the monasterium Honorii at the Lateran, commissioned the monks ‘to care for one choir…and those of St Pancratius [a second Lateran monastery] the other’. He ordered matins [= lauds?], prime, terce, sext, none and vespers to be sung. Presumably the office of vigils was already being sung. Nothing was said about Mass. However, when the monastery of St Stephen Major was added to the Vatican complex in the late seventh or early eighth century, it was arranged ‘quorum festa vigiliarum a monachis trium monasteriorum illic servientium cotidie per ordinem existentia atque nataliciorum missas [that is, ‘masses’ of saints' days – still infrequent in the eighth century – whatever missas may mean (not always the Eucharist)] in eodem loco celebrare [celebrari?]’. Yet Gregory III (c. 731-c. 741), according to the Liber Pontificalis i, p. 417 n. 14), ‘hoc constituit ut servatum [at St Peter's] secundum constitutum quod… a monachis vigiliae celebrentur et a presbiteris ebdomariis missarum solennia’. The most convincing though indirect evidence for participation of Roman monasteries in the singing of Mass (at least at the Vatican) comes from the Venerable Bede (c. 673–735), who describes how John the arch-chanter and abbot of St Martin's monastery at the Vatican was sent to England to instruct the Anglo-Saxon church in chant and liturgy. That John should have instructed the English only in the chants of the Office is improbable: his function St Peter's as arch-chanter would therefore seem to have included the singing of the chants of the Mass as well as of the Office.

¹⁶ But see the important study by Chavasse, A., La Sacramentaire gélasien (Vaticanus Reginensis 316): Sacramentaire presbytéral en usage dans les litres romains du vii^e siécle (Tournai, 1958)Google Scholar. Chavasse in the main presents a convincing case that a discernible Roman core within the old Gelasian sacramentary (that is, BAV, Reginensis MS latinus 316) was assembled precisely for use in the titular and suburban churches – even though this or that detail of his sometimes circular arguments may be questioned. We therefore possess most of the prayers and prefaces that non-papal celebrants used, at least in certain seventh-century titular churches of Rome; but we do not usually know exactly who these celebrants and their assisting ministers were, nor who sang the psalmodic chants of the Mass – if indeed they were sung at all in non-papal, non-stational Masses – nor (as already mentioned) what the function of the few basilicas having monasteries attached might have been at Masses there.

¹⁷ Ferrari, , Early Roman Monasteries, p. 4Google Scholar.

¹⁸ Gregory's biography of Benedict in the Dialogues hardly mentions the Rule. Gregory wished to portray Benedict as a holy man, an ascetic, a worker of miracles and the father of monastic life by reason of his holiness rather than as the wise father of monastic legislation.

¹⁹ A first step in this direction would appear to have been a canon of Eugene II (826) recommending that abbots should be in priest's orders so as to administer discipline and penance: ‘et ita observent ut statuta regularum [note plural] per omnia non inveniantur delinqui’ (quoted in Ferrari, , Early Roman Monasteries, p. 399)Google Scholar.

²⁰ Ibid., p. 4.

²¹ Galla Patricia, a friend of Gregory I, entered a convent near St Peter's and died there as a nun: hence the nickname for the refounded monastery. The present chapel of Santo Stefano degli Abissini marks the site.

²² As quoted in Supino Martini, ‘Carolina romana e minuscola romanesca’, p. 780 n. 23, from Schiaperelli, , ‘Le Carte antiche dell'Archivio Capitolare di S. Pietro in Vaticano’, Archivio della Società Romana di Storia Patria, 24 (1901), p. 446, no. viGoogle Scholar. Approximate translations of the notarial Latin of this and the immediately following excerpts – which in fact are partly Romance in their syntax – are appended here and in notes 23–6 below. ‘Stephen, the “religious”[?] archpriest of the venerable monastery of St Stephen, Christ's first martyr, which at St Peter's is called “the greater”, and John and Benedict, the second and third “religious” priests of the above-mentioned venerable monastery …’

²³ Ferrari, , Early Roman Monasteries, pp. 166–7Google Scholar: ‘Peter, the archpriest of the venerable monastery of SS. John and Paul and Benedict, the archpriest of the venerable monastery of St Stephen [the less] …’

²⁴ Ibid., pp. 237–8; also Supino Martini, ‘Carolina romana e minuscola romanesca’, p. 780 n. 23. ‘Crescentius … the “religious” archpriest of the venerable monastery of Christ's holy confessor [St] Martin, which is placed behind the apse of blessed Peter the Apostle ['s church], and Benedict who is called Galla and John and the other Benedict, the second and third “religious” priests, [?with the agreement of all of us congregations of priests?] …’

²⁵ Ferrari, , Early Roman Monasteries, p. 235Google Scholar. The passable Latin of the papal scrinium reads: ‘Leo IX grants a privilege to John, the archpriest of the venerable church of blessed Peter the Apostle, and the canons of the same church who have been ordained and are to be ordained [or ‘assigned to’] St Martin's monastery, in order that they may sing the divine offices [divini officia: ‘the services of God’] by day and by night for ever …’

²⁶ As quoted in Supino Martini, ‘Carolina romana e minuscola romanesca’, p. 780 n. 23, from Schiaparelli, ‘Le Carte,’ p. 468, no. xvi: ‘To John, the archpriest of the venerable church of the blessed Peter the Apostle and the canons of that church who have been ordained and are to be ordained [or ‘assigned to'] the greater monastery of St Stephen.

²⁷ See the obits for the founders of the convent and other tenth-century Romans in the mid-eleventh-century kalendar in the manuscript Rome, Biblioteca Vallicelliana F 85.

²⁸ Garrison, , Early Italian Painting, ii, p. 123Google Scholar: ‘It must… remain extremely difficult to distinguish manuscripts … made in different Roman scriptoria … from those … made in Farfa.’

²⁹ Supino Martini, ‘Carolina romana e minuscola romanesca’, pp. 785–7 and n. 37.

³⁰ Schuster, I., L'Imperiale Abbazia di Farfa (Rome, 1921; repr. Rome, 1987), p. 132Google Scholar. Also see McClendon, C. B., The Imperial Abbey of Farfa (New Haven, 1987)Google Scholar, for a succinct English summary of the history of Farfa.

³¹ See Schuster, L'Imperiale Abbazia di Farfa, for a detailed account. Legend aside, the community seems to have been founded between the years 680 and 700 by Thomas of Maurienne, a priest and pilgrim from the Savoy, who died in 720. Farfa was the first Italian abbey to receive the imperial privilege after Charlemagne's conquest of the Lombard Kingdom. Not surprisingly, the next three abbots were Franks, and the Benedictine Rule seems to have been introduced. The Emperor Louis II visited Farfa in 872. Saracens – the scourge of the Italian peninsula in the ninth century – laid siege to Farfa for years; Abbot Peter (890–c. 919) finally abandoned the abbey in 897, having split the community into three segments, one of which went to Rome, one to Rieti and one to the area of Fermo. The abbey buildings, seized by the Saracens as a convenient garrison, were accidentally destroyed by fire in 898. (See the quotations describing these events from Abbot Hugh's Destructio Monasterii Farfensis as given in Supino Martini, ‘Carolina romana e minuscola romanesca’, p. 787.) Internal decay set in: the divided and exiled community squabbled over lands and fiefs. Abbot Rimo (or Rimone, 920-c. 930) was poisoned while being bled at Rome (the monks of Farfa were renowned for their pharmaceutical and medical knowledge) and was buried in St Stephen's oratory, near the terme alessandrine – the area east of Piazza Navona mentioned above; see Supino Martini, ‘Carolina romana e minuscola romanesca’, p. 786 n. 37. Abbot Ratfredus (or Rotfredo) was poisoned in 936, and his abbacy was divided between his murderers, Campo and Hildebrand, who in turn fought each other. The Cluniac Dagibert (947–53), installed as abbot at Farfa only after a military onslaught by Alberic, was similarly disposed of, after prolonged but vain attempts at reform. Under the influence of the Ottonian emperors and some of the abbots they appointed, the community gradually bettered its way of life. Otto III, on his coronation trip to Italy, visited San Getulio, a Farfa dependency, in 996.

³² My dates, based on the notation and the liturgical organisation and contents of the manuscripts. Susan Boynton has kindly pointed out that these dates were independently confirmed some time ago by P. Dinter, who ‘argued that the manuscript must have been copied between 1030 and 1060 because the consecration of the new church (in 1060) is not recorded in it and because the feast of St Maurus was first celebrated after 1033 (introduction to Liber Tramitis Aevi Odilonis Abbatis, ed. Dinter, P., Corpus Consetudinum Monasticarum 10 (Siegburg, 1980), p. xl n. 93)Google Scholar’. I am grateful to Prof. Boynton for this reference.

³³ Although the San Saba lectionary has Roman minuscule script with typically Roman interlaced zoomorphic decorative initials and a Roman kalendar, its chant was neumed with Dijonesque central-French neumes. See the description of RoA 1383 in Boe, ‘Chant Notation in Eleventh-Century Roman Manuscripts’, pp. 49–56.

³⁴ Supino Martini has surprisingly little to say about Chigi C.VI. 177. In Roma e I'area grafica romanesca, pp. 254–5 and n. 139, she dates it ‘to the last decades of the eleventh century’ and suggests it was in Subiaco in the first decades of the twelfth century (pp. 37–8 nn.28, 30 and 32; see also p. 171, where she calls attention to marginal additions in a Beneventan hand, probably made in Subiaco). For the most part she treats this manuscript indirectly, while dealing with other manuscripts. Chigi C.VI. 177 is never fully discussed in her main text. Perhaps Supino Martini's reticence here was due to her erroneous description of the manuscript as ‘messale monastico’ in the earlier article, ‘Carolina romana e minuscola romanesca’, p.788 n. 40. In the later work, the manuscript is correctly referred to as a ‘breviario’. Garrison devoted a great deal of attention to this manuscript. In Garrison, , Studies in the History, i, p. 24Google Scholar and passim, he assigned it to Subiaco, a view he was later at pains to correct, as in Garrison, , Early Italian Painting, II, p. 84 n. 7, pp. 122, 124–9 and 136–9Google Scholar; ibid., III, p. 107 n. 1, p. 108 n. 1, and p. 280 n. 8. Also see his Selected Studies, II, pp. 69, 73 and 75, figures 2 and 3 (where a line of revised notation is cut off at the top of the figure). In addition, see his ‘Saints Equizio, Onorato, and Libertino’, and his studies on the local saints peculiar to Farfa devotion in Selected Studies, I, pp. 367–85. Garrison lists the kalendric Farfa symptoms found in Chigi C.VI. 177 (I copy the Italian forms of the saints' names that he uses): 9 June, Getulio and Companions; 8 July, Lorenzo, Abbot of Fara, and Susanna; 3 November, Valentino and Ilario; 10 November, Tommaso, Abbot of Farfa; and 23 December, Vittoria, Roman martyr. Garrison points out that the decorative initials have Roman features. He thought that the manuscript might have been written by Gregory of Catino; but this ascription is ruled out by chronology and paleography, as Supino Martini's extensive treatment of the script of Gregory of Catino makes clear. I do not see how adiastematic Dijonesque notation could have been written at Farfa or elsewhere in central Italy after 1060 at the latest. On the basis of the original notation, I therefore dated the earliest parts of the manuscript to the years 1030–50 (or possibly as late as 1060), before the Guidonian reform of notation had spread and while the influence of William of Dijon, who died in 1031, was still being felt at Farfa.

³⁵ Facsimile edition in Paléographie Musicale, vol. viii. See Corbin, S., Die Neumen, Paleographie der Musik 1/3, ed. Arlt, W. (Cologne, 1977), pp. 102–10Google Scholar, and especially Huglo, M., ‘Le Tonaire de Saint-Bénigne de Dijon’, Annales Musicologiques, 4 (1956), pp. 7–18Google Scholar. Also cf. Rankin's, S. ‘insular looped climacus’, mentioned in her ‘Neumatic Notations in Anglo-Saxon England’, Musicologie médiévale: Notations et séquences, Actes de la Table Ronde du C.N.R.S. à l'Institut de Recherche et d'Historie de Textes, 6–7 09 1982, ed. Huglo, M. (Paris, 1987), p. 132Google Scholar. Kenneth Levy reminds me that the ‘looped climacus’ appears in certain middle and northern Italian manuscripts, as Rome, Biblioteca Angelica MS 123, and Bodleian, Douce 222.

³⁶ According to Bulst's, N. chronology of William's life in Untersuchungen zu den Klosterreformen Wilhelms von Dijon (962–1031), Pariser Historische Studien 11 (Bonn, 1973), pp. 270–7Google Scholar, William returned to Italy (1) in 995 (pilgrimage to Rome, Benevento and Monte Sant'Angelo); (2) in 999–1001 (Rome, Farfa, Vercelli and Volpiano); (3) in 1003 (Fruttuaria, for the dedication and beginning of construction of t he new abbey); (4) in 1005 (for King Arduin's donation and grant of rights for Frutturia in William's presence); (5) in 1006–7 (consecration of the church of San Benigno at Fruttuaria); (6) in 1012 (Rome, where Benedict VIII granted papal protection to St Bénigne, Dijon); (7) in 1014–15 (Pavia and Rome); and (8) in 1026–8 (Ivrea, Rome and Susa). On William's foundation of the abbey of Frutturia, see ibid., pp. 115–45; on the visit to Farfa, see p. 189.

³⁷ See P. Thomas, ‘L'Oeuvre musicale de S. Odon’ (Dijon, 1950); Hourlier, H., ‘Remarques sur la notation clunisienne’, Revue Grégorienne, 30 (1951), pp. 231–409Google Scholar; and Picasso, G., ‘Usus e consutudines cluniacensi in Italia’, L'Italia nel quadro dell'espansione europea del monachesimo cluniacense, Atti del Convegno internazionale di Storia Medievale, Pescia 26–28 11 1981, in Italia Benedettina, 8 (Cesena, 1985), pp. 297–311Google Scholar. I thank Walter Simons for supplying the last citation.

³⁸ In the orally delivered version of my earlier study ‘Chant Notation in Eleventh-Century Roman Manuscripts’, I called attention to a combined evangeliary-sacramentary for Sundays and feasts, BAV, Archivio San Pietro F 12, which I thought might have been ‘from a scriptorium associated with the Vatican, probably the monastery of St. Martin’, following Jounel, P., Le culte des saints dans les basiliques du Latran et du Vatican au douzième siècle, Collection de l'Ecole Francaise de Rome 26 (Rome, 1977), pp. 34–5Google Scholar. This attribution can no longer stand: the manuscript may indeed have been copied from an exemplar based on Vatican usage; but (as Martini, Supino makes clear in Roma e l'area grafica romanesca, pp. 32–83 and n. 101Google Scholar) the invocation of certain saints never found in Roman books but found rather in books from Emilia and Tuscany and, more particularly, the name of the scribe – John Corbus [or Corvus], deacon and monk' (fol. 209^v) – and the evident monastic destination of the manuscript all render use at a Vatican ‘monastery’ in the mid eleventh century (my date) unlikely. (Supino Martini's date of ‘second half, eleventh century’ makes a Vatican destination even more unlikely. See above, on the evolution of the Vatican monasteries.) In the sacramentary, two tracts for the vigil of Pentecost, Adtende celū and Sicut ceruus, are cued with the opening neumes of the Gregorian melody. In the Easter vigil (fol. 128^r), Kyrie eleison is notated in full – an idiosyncratic version of segments of Kyrie XV, Dominator Deus, like the version in the Nonantola manuscript, Rome, Biblioteca Casanatensis MS 1741, fol. 12^r-v. The intonation for Gloria in excelsis that follows belongs to Gloria III (also unknown at Rome, like the Kyrie), whose earliest concordance north of the Alps appears to be in Paris, Bibliothèque Nationale, MS lat. 887, dated 1028 (?). In the section devoted to the gospels of SP F 12, Hely Hely of the Palm Sunday passion is notated in a style consistent with that of the sacramentary section of the manuscript and probably from the same hand. The notation is adiastematic, in the style that B. Stäblein called ‘central-French, Epoch I’. The three-note looped descending form is not used, but rather the standard climacus consisting of virga and two puncta. The manuscript, then, was written somewhere in central Italy but hardly Rome. How it entered the Archivio San Pietro is not known.

³⁹ See the edition of the life of William of Dijon (962–1031) by his nearly contemporary admirer and sometime companion, the garrulous chronicler and wandering monk Rodulfus Glaber, transcribed from an eleventh-century Fécamp manuscript, Paris, Bibliothèque Nationale MS lat. 5340: ‘Rodolfus Glabers Vita domini Willelmi abbatis’, ed. Bulst, N., Deutsches Archiv für Erforschung des Mittalalters, 30/2 (1974), pp. 450–87Google Scholar. See further, Glabri, RodulfiHistoriarum Libri Quinque, ed. and trans. France, J.Google Scholar, and Vita Domni Willelmi Abbatis, ed. Bulst, N., trans. France, J. and Reynolds, P. (Oxford, 1989)Google Scholar. See also il Glabro, Rodolfo, Cronache dell'Anno Milk (Storie), ed. Cavallo, G. and Orlandi, G. ([Milan], 1989), especially pp. 140–41 and 331 on FruttuariaGoogle Scholar.

⁴⁰ The Farfa origin of RoN Farf. 32 is established on fol. 1^v by the rubric ‘Incipit de uenerabili patre nostro Thoma. Le prima’. Thomas of Maurienne was the first Abbot of Farfa.

⁴¹ Garrison, Early Italian Painting, i, n. 4: ‘4th q. 11th c.?’, on the basis of the initials. Martini, Supino, in Roma e l'area grafica romanesca, pp. 253–4 and n. 134Google Scholar, says ‘la presenza di abbondanti legamenti caratterizza la sicura romanesca dell'Omiliario Farf. 32 della Bibl. Naz. di Roma…, manoscritto prodotto nell'abbazia farfense negli ultimi decenni del secolo XI, pregevole in ogni suo aspetto: grande formato, qualità della pergamena, decorazione, scrittura’.

⁴² I should like to call such notation as is found in the Lamentations of RoN Farf. 32 ‘protopeninsular’. Such notation is most often found in prefaces and other passages of anaphoral chant in peninsular missals from the late tenth century and throughout the eleventh. Conservative of their tradition, these books begin to employ Guidonian lines and clefs only in the mid twelfth century. From proto-peninsular notation, new liquescent forms were developed south of Rome, and their specific uses were greatly extended; the punctum distinction appeared and was regularised. From Rome onwards, liquescence on the contrary was limited to the special instances found in northern European notation, and the virga came to be used for all the rising notes of a series, instead of for the top note only as in the south. From Rome northwards, differentiation of punctum slant according to the direction of melody was never developed nor borrowed but seemingly was rejected – or at least was never consistently applied. Roman (or central-Italian) and Beneventan notations in their mature forms are distinct, though too often confused. That they both may derive from a common ‘proto-peninsular’ ancestor remains for now an attractive but unproven theory.

⁴³ Gribomont, J., ‘Les Prophètes de Belizo (San Pietro C 92)’, Miscellanea Codicologica F. Masai Dicata, 2 vols., ed. Cockshaw, P. et al. (Ghent, 1979), i, pp. 189–201Google Scholar.

⁴⁴ Enrico Marriott Bannister [Henry Marriott Bannister], Monumenti Vaticani de paleogrqfia musicale latina, Codices e Vaticanis Selecti Phototypice Expressi, 2 vols. (Leipzig, 1913)Google Scholar.

⁴⁵ SP C 105 – a pars hiemalis only – as catalogued in Salmon, P., Les Manuscrits liturgiques latins de la Bibliothèque Vaticane, Studi e Testi 251, 253, 260, 267 and 273 (Città del Vaticano, 1968–1972), iv, no. 13, bears the annotation ‘[fols.] 29–91Google Scholar, Versus sibyllae notés, leçons d'Isaie pour les vigiles de Noël; sermon pour Noël’. C. Stornajolo's manuscript ‘Inventarium’ of the Latin manuscripts in the Archivio San Pietro also gives notice of the neumed passage (p. 14). Also see Lambot, C., ‘La tradition manuscrite des sermons de saint Augustin pour la Noël et l'Epiphanie’ Revue Bénédictine, 77 (1967), p. 226CrossRefGoogle Scholar; Gamber, K., Codices liturgici latini antiquiores, 2 vols., Spicelegii Friburgensis Subsidia (Freiburg, Switzerland, 1968), ii, p. 559Google Scholar; Martini, Supino, ‘Carolina romana e minuscola romanesca’, pp. 791–2 n. 45, and Roma e l'area grqfica romanesca, pp. 59–62Google Scholar. Except in the first two references, no notice is taken of the neumed passage. An extensive article by LÖw, G., ‘II più antico sermonario di San Pietro in Vaticano’, Revista di Archeologia Cristiana, 19 (1942), pp. 143–83Google Scholar – inaccurate in other respects – dates the manuscript a century too early but nevertheless contains useful inventories of this and other Office homiliaries.

⁴⁶ For SP C 107, see Salmon, , Les manuscrits liturgiques latins, IV, no. 15, with the dates ‘s. xii, xiii’, and Stornajolo, , ‘Inventarium’, ii, p. 20Google Scholar, with the same dates. For SP C 106, see Salmon, IV, no. 14, with the date ‘s. xiv’, and Stornajolo, II, p. 20. (In SP C 106, the latest copy of the homiliary, the Song of the Sibyl has disappeared.)

⁴⁷ Etaix, R., ‘Le prologue du sermonnaire de'Alain de Farfa’, Scriptorium, 18 (1964), pp. 3–10CrossRefGoogle Scholar; the quotation is on p. 10. See the conclusions of Löw's article cited in note 45 above. Also see Hosp, E., ‘II Sermonariodi Alano di Farfa, Ephemerides Liturgicae, 50 (1936), pp. 373–81, and 51 (1937), pp. 210–41Google Scholar, and Ratti, A., ‘L'omeliario detto di Carlo Magno e l'omeliario di Alano di Farfa’, Reale Istituto Lombardo di Scienze e Lettere, Rendi Conti, ser. 2, 33 (1900), pp. 480–9Google Scholar.

⁴⁸ This Roman homiliary, widespread in Italy and Bavaria during the eighth and early ninth centuries, was replaced throughout the Carolingian empire by a very different one compiled by Paul the Deacon on Charlemagne's express order. The older homiliary (the original of Alanus' and Egino's editions) remained in use only at Rome. See Chavasse, A., ‘Le calendrier dominical romain au sixiéme siècle; l'épistolier et l'homélaire prégrégoriens’, Recherches de Science Réligieuse, 38 (1952), pp. 234–46, 41 (1953), pp. 96–122, and especially 41 (1953), pp. 111–122Google Scholar, ‘III. Un homéliaaire du sixième et du septième siècle’. Professor Joseph Dyer has kindly supplied additional references: Grégoire, R., Homéliaires liturgiques mediévaux: Analyse de manuscrits, Biblioteca degli ‘Studi Medievali’ 12 (Spoleto, 1980), pp. 223–44 (especially for SP C 105)Google Scholar; Chavasse, C., ‘Un homéliaire roman du VI^e et du VII^e siècle: Le sermonnaire des Saints-Philippe-et-Jacques et le sermonnaire de Saint-Pierre’, Ephemerides Liturgicae, 69 (1955), pp. 17–24Google Scholar; and Chavasse, A., ‘Le sermonnaire Vatican du VII^e siècle’, Sacris Erudiri, 23 (1978–1979), pp. 225–89CrossRefGoogle Scholar.

⁴⁹ My transcription, with manuscript spelling, punctuation, suspensions and abbreviations retained (except that the signs for pro, per, et, -que and -orum have been resolved). The text with modernised spelling and punctuation – and with the entire curse – is given in Löw, ‘II più antico sermonario di San Pietro’, p. 10, and twice by Supino Martini – first in ‘Carolina romana e minuscola romanesca’, p. 791 n. 45, where the opening is transcribed 'Omnibus sit notum tangentibus istum librum quod pro amore beati Petri apostoli fecit illum scribere Leo indignus et negotiens qui Schifo vocatur… and secondly in Roma e l'area grafica romanesca, p. 59 n. 47, with the incorrect beginning ‘Omnibus sit notum scribere Leo indignus et negotiens qui Schifo vocatur…’, where an entire line of copy has dropped out. The complete text is given in Gréoire, , Homéliaires liturgiques, p. 225Google Scholar. I translate: ‘[Be it known to all who handle] this book, which Leo, an unworthy merchant who is called SCHIFO, with devout intent for the love of blessed Peter the Apostle ordered to be written and completed, [which] he offered to Peter, the highest Prince of the Apostles and bearer of the keys of the heavenly kingdom, so that it might ever be kept in use irrejragabiliter [without hindrance?] for the readings by day and by night before his most sacred body. According to this regulation: [anathema and curses on him who steals it ] … I beseech you, O reader [of the lessons?], to pray for me a sinner – Leo the merchant, who is called Schifo – who had this book written for the redemption of my soul and offered it [when] written to the blessed apostle Peter with a confident will, in order that I may perceive him, the Prince of the Apostles, upon whom the head of the entire Church is founded, to be an advocate with the merciful judge, the Lord Jesus Christ, so that, having received pardon for my numberless sins by his [St Peter's] prayers, I may be worthy to be placed by God among his saints in eternal blessedness. Amen.’

⁵⁰ Negotiens: then current for the correct Latin negotians (modern Italian negoziante). The form with e is given in Nouum glossarium mediae Latinitatis, ed. Franz Blatt, iii-iv, cols. 1188–9, as occurring in Italian documents from around the year 1000: as for instance negociens in Cod. Lang., 999; negotiens in cartularies from Mica Aurea [Rome], 1003, and from Santa Maria in Via Lata [Rome], 1004 and 1032. Supino, Martini in Roma e l'area grafica romanesca, p. 60 n. 47Google Scholar, adds the important notice that a ‘Leo vir magnificus negotiator’ appears in two documents from Subiaco dated 20 February 965 (Reg. Subl. nos. 130, 131, pp. 180–2). Of this Leo she says ‘molto probabilmente da identificarsi con l'omonimo del 999 e con il donatore del'Omiliario’. See below.

⁵¹ Supino Martini, ‘Carolina romana e minuscula romanesca’, p. 792 n. 45, citing Schiaparelli, L., ‘Le carte antiche dell'Archivio Capitolare di S. Pietro’, Archivio della Società Romana di Storia Patria, 24 (1901), pp. 446–8Google Scholar; and Martini, Supino, Roma e l'area grafica romanesca, p. 60 n. 47Google Scholar: ‘l’arciprete del monastero vaticano di S. Stefano concede ad meliorandum una terra sementaricia a Leone detto Schifo, nobilis vir’. Stornajolo, ‘Inventarium’, p. 19, refers however to Archivio Fasc. 22 [2], capsula 61, and remarks, ‘Huius Leonis pseudo qui Schifo vocatur extat memoria in instrumento locationis territorii ei factae a presbyteris monasterii s. Stephani maioris an. 998.’ See also Grégoire, , Homéliaires liturgiques, p. 224 n. 3Google Scholar.

⁵² She objects to the possibility that SP C 105 might have been written at another Roman scriptorium, from an exemplar removed for the purpose from St Peter's, on the grounds that a book belonging to St Peter's would not have been lent to another scriptorium for copying. (The anathema in the dedication of SP C 105 tends to reinforce her opinion, if only weakly.) More convincingly, she emphasises the similarity of SP C 105 to Archivio San Pietro C 103, given to St Peter's by Iohannes Filius de Urso et Stephania: ‘Simili nella pergamena assai rozza, nelle dimensioni, nella composizione dei fascioli, nell'impostazione della pagina e nella semplicità delle iniziali calligrafice e decorate… i due manoscritti sono vergati ciascuno da una sola mano, purtroppo rimasta anonima… in minuscola romanesca’ (Rome e l'area grafica romanesca, pp. 60–1).

⁵³ See Bischoff, B., ‘Die lateinische Übersetzungen und Bearbeitungen aus den Oracula Sibyllina’ in his Mittelalterliche Studien: Ausgewählte Aufsätze zur Schriftkunde und Literaturgeschichte, 1 (Stuttgart, 1961), pp. 150–71Google Scholar, with an excellent bibliography in the notes; Iudicii signum and its Greek original enshrining the acrostic IHΣOYΣ XPEIΣTOΣ ΘEOY YIOΣ EΩTHP [ΣTAYPOΣ] are discussed pp. 154ff. Also see H. Leclercq, DACL, I/l, s.v. ‘Acrostiche’, cols. 356–72; the Greek text is given in col. 357. The earliest complete text is found in the address or manifesto purported to have been written or issued by the Emperor Constantine (λόγoζ őν έγραψε TΩ TΩN ATIΩN ΣYΛΛOΓΩ, ‘oratio quam inscripsit Ad sanctorum coetum’) and appended by the church historian Bishop Eusebios of Caesarea to his panegyric Life of Constantine some time between the Emperor's death in 337 and Eusebios' own death in 339 or 340. (See Vita Constantini, IV, cap. 32 (PG 20, cols. 1181–2), and Ad sanctorum coetum, cap. 18 (PG 20, cols. 1285–90), where the acrostic poem is attributed to the Erythraean Sibyl.) Probably Eusebios assembled the poem, at least in part, from already existing Sibylline verses to form the acrostic – the acrostic of the poem itself being drawn from the word IXΘYΣ, famous long before Eusebios: see H. Leclercq, DACL, VII/2, cols 1990–2086, s.v. IXΘYΣ, and especially cols. 2001–3 (quoting R. Mowat) and col. 2010, §10. I think Eusebios certainly composed the seven lines of the Greek text forming the acrostic ΣTAYPOΣ (‘cross’) – especially suitable for their pretended author, Constantine. Leclercq (s.v. ‘Oracle’, DACL, Xll/2, col. 2236) goes further: ‘Or maintenant, de qui étaient les vers mis sur le compte de la sibylle d'Erythrée? D'Eusèbe, évèque de Césarée. Du discours adressé al'assemblée des fidèles ils passèrent dans les collections des Oracles sibyllins [Bk. VIII, lines 217–50] et furent souvent reproduits, toujours inserés parmi les vers, que l'on attribuaità la prophetésse d'Erythrée.’ The vast collection of Sibylline utterances dating perhaps from the second century B.C. to at least the fourth century A.D. have been edited in Die Oracula Sibyllina, ed.Geffcken, J. (Berlin, 1902)CrossRefGoogle Scholar and in Sibyllinische Weissagungen, ed. Kurfess, A. (Munich, 1951)CrossRefGoogle Scholar. The Sibylline oracles were collected in the form given in these editions ‘hardly earlier than the beginning of the fifth century’ (Kurfess, p. 21) and show evidence of two previous redactions, one Jewish and another Christian. It is entirely probable that the Greek poem in its acrostic form was inserted into the collection from the pseudo-Constantinian manifesto published by Eusebios, as Leclercq asserts, although some at least of the verses must have been circulating earlier. The Latin writer Lactantius quotes four of the verses, separately and in Greek, in the seventh and last book of his Diuinarum institutionum, composed between 304 and 311 (PL 6, cols. 111–822). He also quotes many others that appear in bk VIII of the Oracula Sibyllina. (Lactantius, professor of rhetoric under Diocletian, having converted to Christianity, fled his job during the last persecution but was recalled from exile by Constantine to tutor his son. Lactantius died around 325.) The four verses in question, numbered according to the order of their appearance in the acrostic poem, are given below as they are found in bk VII of the Divine Institutes, followed by PL 6 column numbers and accompanied by the Latin translation that Augustine used in his City of God (see below). cited in Opera Quoduultdeo Carthaginiensis Episcopo tributa, ed. Braun, R., Corpus Christianorum, Series Latina 60 (Tournai, 1976)Google Scholar, In this edition the sermon appears on pp. 227–58, and the Song of the Sibyl on pp. 248–9. (I wish to thank Prof. J. Dyer for this important citation.) See also Grégoire, , Homéliaires liturgiques, p. 227Google Scholar, no. 11 (in SPC 105), and pp. 431–2, no. 10 (in the homiliary requested by Charlemagne for the night office from Paul the Deacon, monk of Monteccasino – see Charlemagne's letter, ibid., pp. 422–3).

⁵⁴ See Corbin, S., ‘Le Cantus sibyllae: Origine et premièrs textes’, Revue de Musicologie, n.s. 101–2 (1952), pp. 1–10Google Scholar – an inventory of manuscripts having notation for the Song of the Sibyl, with full reference to Anglès, H., La música a Catalunya fins al segle XIII (Barcelona, 1935), pp. 288–302Google Scholar and additional plates preceding, figures 35, 46, 48, 58 and 59. Anglès surveys Iberian, Aquitanian, northern French and a few Italian manuscripts of a similar nature and transcribes the first three verses of many of them in his seemingly impressive Taula 1. Unfortunately, his transcription of adiastematic and uncleffed notation in campo aperto is often questionable or even demonstrably flawed, usually in such a way as to suggest greater unanimity and less variety among readings than is likely to have been the case. Moreover, his transcription only three verses effectively hides important differences among the manuscripts in the way that text and music of the entire poem was presented – namely, whether the melody and the words ‘Iudicii signum … madescet’ of the first verse were repeated (a) as a respond, regularly or irregularly, after groups of internal verses as well as at the end, or (b) only at the end, like an antiphon to a psalm, or (c) not at all, being sung only once as the opening verse. Naturally enough, manuscripts in group a – to which most of Anglès's Catalonian versions belong – ignore the acrostic, hiding it from sight by not capitalising letters that begin whole verses, or by capitalising some such letters but not others, or by capitalising letters that begin half-verses as well. On the other hand, almost all manuscripts in groups b and c make the acrostic stand out by the use of large capitals aligned vertically, often in the left margin, to begin each verse, and sometimes by putting each hexameter in a single line to itself – which we might call the manuscripts of group d. It is immediately obvious from Fig. 6 that the revised version of SP C 105 belongs to group d and that it replaces the layout of either group b or of group c – or conceivably of group a, although no secure evidence remains of a refrain in the lower script legible on fol. 29^r, and fol. 29^v is too thoroughly erased to tell. (See the main text following and n. 55 below.) The Song of the Sibyl occupies an important, even crucial position in a recent article by Barrett, S., ‘Music and Writing: On the Compilation of Paris Bibliothèque Nationale lat. 1154’, Early Music History, 16 (1997), pp. 55–96CrossRefGoogle Scholar. The poem is quoted at the outset and reproduced complete from the manuscript, fols. 122^r-v and 123^r, in Fig. 1–3, which show the confusing double notation added to the manuscript. It is clear that this version originally belonged to our group a. Cues for the refrain ‘Iudicii signum’, original to the manuscript and predating the notation, were entered after the first eight half-verses (each half-verse being supplied with a capital letter in the margin) and thereafter following each set of four half-verses, so as to produce the melodic pattern R abab R ab R ab R …. ab R. But the cues for the refrain were inked out, although they are still mostly legible in Figs. 1 And they were never notated. The cancellation of the refrains in PaN 1154 allows the unfolding acrostic to be heard, marked as it is in alternate verses by two notes rising a fifth for the first syllables of lines 2 although visually the acrostic remains disguised by the fully capitalised half-verses. Certainly these alterations, as well as the additions of notation, must influence our view as to why the Song of the Sibyl was included in this nonliturgical collection and how the written and corrected text was used. The more thoroughgoing revisions to the Song of the Sibyl at the Roman night office may likewise stem from a growing sensitivity to the unique character of this text, evident elsewhere throughout the eleventh century. It would be gratifying to know the exact source of the Petrine melodic revision. (The citation of the article by S. Corbin on p. 60 of Barrett's study is incorrect: it should read as at the head of this note.)

⁵⁵ For example, the syllables ‘-cla futurus’, ‘-que fidelis’ and ‘in ipso’ are clearly legible on a second line. The erased layout in general thus resembled that of PaN 1154, but without capital letters to begin the parts of hexameters placed on the second line.

⁵⁶ Cf. the erasure of the Roman form of Gloria in excelsis ‘A’ in the Easter Vigil Mass in Vat 5319 and its replacement with Gloria Bosse 51; see Boe, J., ‘Gloria A and the Roman Easter Vigil Ordinary’, Musica Disciplina, 36 (1982), pp. 43–66Google Scholar. The legible original cadence in SP C 105 for the last three syllables of the first verse, ‘madescet’, CD D D (and similarly for final cadences of the second phrase member throughout, where they can be read), was replaced by E D D in the revised notation. Most of Anglès's earlier manuscripts have CD D D for these cadences.

⁵⁷ Discovered by Peter Jeffery, as reported at the annual meeting of the American Musicological Society in Chicago, November 1991.

⁵⁸ For a balanced summary of the period – Baronius' ‘saeculum obscurum’ – see Zimmermann, H., Des Papsttum im Mittelalter (Stuttgart, 1981), especially pp. 90–1Google Scholar. For more detailed studies, see Gregorovius, F., Geschichte der Stadt Rom im Mittelalter, ed. Kampf, W., i (Munich, 1978)Google Scholar; Bertolini, O., Roma di fronte ai Bizanzii e ai Longobardi (Bologna, 1941)Google Scholar; or Ullman, W., A Short History of the Papacy in the Middle Ages (London, 1971), pp. 111–13Google Scholar.