[2] Gevaert, François Auguste: La melopée antique dans le chant de l'église latine (Ghent 1895 Google Scholar; repr. Osnabrück 1967). The following recent studies may be mentioned. Hughes, David: ‘Evidence of the traditional view of the transmission of Gregorian chant’, Journal of the American Musicological Society 50 (1987), 372–404 Google Scholar, is a study of chant transmission based on the melodies of Gevaert's ‘theme 29’. Bailey, Terence and Merkeley, Paul: Antiphons of the Ambrosian Office II (Ottawa 1989)Google Scholar is a taxonomy of Ambrosian antiphons. The Old-Roman antiphon repertory is analyzed in Nowacki, Edward: Studies on the office antiphons of the Old Roman manuscripts (Ph.D. diss., Brandeis University 1980)Google Scholar; see also Nowacki, : ‘The Gregorian office antiphons and the comparative method’, Journal of Musicology 4 (1985–1986), 243–275 Google Scholar.

[3] Gevaert believed that the typical melodies came from Ancient Greek music. The short forms of the ‘themes’ were hypothetical Greek melodies, characteristic of modal octave species. This view is inconsistent with present knowledge of the music of Ancient Greece.

[4] It would be inappropriate to fix a year after which no tonaries are relevant, for theory and practice inevitably coincided only irregularly.

[5] A more refined and specific procedure for the layering of texts (not exclusively or necessarily in time) has been developed by Terence Bailey in his research on Ambrosian chant. Bailey has identified the biblical sources of all Ambrosian antiphons and he has divided these into four layers. In contrast to those in Gevaert's first layer, antiphons with psalmodic texts are divided into two groups: those whose text is taken from the psalm with which it is sung; and those whose text is taken from a psalm other than the psalm with which it is sung. For details of this inquiry, the relationship with melodic families and the implications for chronology, see Bailey, and Merkley, : The Antiphons of the Ambrosian Office (Ottawa: The Institute of Mediaeval Music), vol.1 (1989)Google Scholar, the liturgy and texts, vol.2 (1989), an edition of the antiphons with taxonomy of the melodies, vol.3 (forthcoming), commentary.

[6] See Guido of Arezzo: Micrologus, ed. van Waesberghe, Joseph Smits, Corpus Scriptorum de Musica 4 (1955), 144–5Google Scholar, translated, amongst others, by Apel, W.: Gregorian Chant (Bloomington 1958), 175 Google Scholar, and by Babb, W.: Hucbald, Guido and John on Music (New Haven 1978), 67 Google Scholar. I have written at length on this change and its implications in my book Italian Tonaries (Ottawa 1988)Google Scholar, of which it is the central topic. See also my articles ‘The transmission of tonaries in Italy’, Studies in Music from the University of Western Ontario 10 (1985), 51–74 Google Scholar; and ‘Tonaries, differentiae, termination formulas, and the reception of chant’, Festschrift Luther Dittmer (Ottawa, to appear in 1990)Google Scholar. Since completing my work on Italian tonaries I have studied on site almost every previously known tonary and several previously unknown ones. I have entered the lists of antiphons from large tonaries into a data base, as I did for my study of Italian tonaries, enabling me to identify and study conflicting assignations. I am presently engaged in interpreting the results and studying the relationship between assignments in tonaries and the modes and psalm tones in antiphoners.

[7] By differentia I mean a list of antiphons in a tonary, not the music sung to the doxology. This usage corresponds to that in the commentary of many tonaries and in the treatises associated with them. The judgement of the differentia by the initial note was first used in the tonary of Berno of Reichenau.

[8] Melodies of variable transposition or ambiguous range were the two chief remaining causes of difficulty.

[9] de Coussemaker, Edmond: Scriptorum der musica medii aevi nova series…, II (Paris 1867), 1–73 Google Scholar. New edition by Le Roux, Mary Prothase: The ‘De harmonica institutione’ and ‘Tonarius’ of Regino of Prüm (Ph.D. diss., Catholic University of America, 1965)Google Scholar.

[10] Gerbert, Martin: Scriptores ecclesiastici de musica sacra potissimum, I (St Blasien 1784), 230–231 Google Scholar; Le Roux, op.cit., 23. I translate as follows: “When frequently in the dioceses of your church, the choir singing the melody of their psalms resounded in confused voices because of dissonance of mode, and when I had seen your Reverence often disturbed by this, I seized the antiphoner; and, turning it diligently from beginning to end through the order, I distributed the antiphons which I found in it according, as I judge, to the proper modes… However, some add other divisions, which we judge to be superfluous. But lest we be reproved by superstitious musicians, we have taken care to make a note of these, below or above, in the margin, leaving it to the judgement of the skilful singer whether he wishes to regard them as necessary or superfluous.” On Regino see also Bower, Calvin: ‘Natural and artificial music: the origins of development of an aesthetic concept’, Musica Disciplina 25 (1971), 19–20, 24 Google Scholar; and Dyer, Joseph: ‘The singing of psalms in the early-medieval Office’, Speculum 44 (1989), 547–8Google Scholar.

[11] Huglo, Michel: Tonaires, Les: Inventaire, Analyse, Comparaison (Paris 1971), 182ffGoogle Scholar.; id.: ‘L'auteur du Dialogue sur la Musique attribué à Odon’, Revue de Musicologie 55 (1969), 119–171.

[12] Huglo, Les Tonaires, 71ff.; Bower, Calvin: ‘Natural and artificial music: the origins and development of an aesthetic concept’, Musica Disciplina 25 (1971), 17–33 Google Scholar.

[13] The date of the second section of Brussels 2750–65, the section containing the music and treatises, is unclear to me. The earliest estimate, one qualified by a question mark and ‘circa’, is 930, by Chartier, Yves: ‘Regino of Prüm’, The New Grove Dictionary. LeRoux, p. 14 Google Scholar, following the Brussels catalogue, suggests the second half of the 10th century, and Huglo is silent on the subject. I do not think that the script can be dated as precisely as this; it may be from the 10th or 11th centuries. The notation consists of plain Messine neumes. More precise dating will have to wait for new liturgical evidence.

[14] Cf. Merkley, , Italian Tonaries, 30 Google Scholar. One example is to be found in Rome, Biblioteca Apostolica Vaticana, Reg.lat.1616, in which the scribe apparently used the section of Guido's poem ‘Musicorum et cantorum’ beginning ‘Feci regulas pertas’ (‘I made the rules known’) as a prologue to the tonary.

[15] The Brussels tonary would then be similar to the group of Italian tonaries which assign the antiphon O beatum pontificem to the seventh differentia of the first mode. The scribes made this assignment in response to a preface to their tonaries common to all the manuscripts, and attributed in two of the sources to an ‘Abbot Odo’, which states that the seventh differentia is the correct one for this antiphon. In fact in these tonaries the seventh differentia is small, containing only from one to three antiphons, and the melodic incipit is duplicated exactly in the third differentia of the first mode, which is large; it seems likely, therefore that the scribes made up the seventh differentia just for this antiphon. See Merkley, , ‘Transmission’, 70–71, Italian Tonaries, 31 Google Scholar.

[16] Huglo, ‘L'Auteur’; Merkley, , Italian Tonaries, 32 Google Scholar.

[17] It should be pointed out that there is nothing in the text of the Epistola that indicates unequivocally that Regino compiled a tonary and not an antiphoner with modal indications. Although there are no extant antiphoners of this kind from the early 10th century, there are graduals with indications of mode.

[18] I have changed Gevaert's order, which was taken from the hypothetical Greek modes from which he believed the themes to have been derived. I have used letter-notation, because of its compactness, making it easy to align and compare melodies, and because it was the notation most used in early Italian tonaries. The convention of upper- and lower-case letters is the one followed in Micrologus. I have abandoned Gevaert's distinction between b-natural and b-flat, which derived from his own theoretical prejudices rather than specific sources.

[19] The theory of centonization, comparing the construction of chant with the mosaic structure of medieval poetry, was first proposed by Ferretti, Paolo in Estetica gregoriana (Rome 1934)Google Scholar.

[20] From Paulinus' biography of St Ambrose, written in 397, it is inferred that the practice of singing an antiphon before and after a psalm goes back to the 4th century.

[21] The manuscript is from St.-Maur-des-Fossés. Pieces added after the work of the main scribe (more recent offices) lack modal indications, which confirms that the modal indications are contemporary with the main text (Huglo, Tonaires, 113). In other words the duplication of material was planned from the start.

[22] This is Cologny near Geneva, Biblioteca Bodmeriana 77: see Merkley, , Italian Tonaries, 132 Google Scholar.

[23] It is difficult to establish the date of this source. The capitularies in the manuscript are for years in the early 9th century. The script has no pre-Gothic traits. The tonary is in a hand unique in the maniscript, and it is difficult to be precise in the dating of a small sample of Caroline script. It is safe to say that it was written in the 9th or 10th century.

[24] Huglo, Tonaires, 313.

[25] Gevaert, 331.