(Lat. tonarius, tonarium, tonale, toni, octo toni, intonarium, intonarius; Ger. Tonar; Fr. tonaire).
Liturgical book of the Western Christian Church in which the antiphons of the Office and the Mass and, by extension, the responsories and even other chants are classified according to the eight psalm tones of Gregorian chant. Tonaries are theoretically self-contained (e.g. the libellus of F-ME 351, ff.66v–75v), but were often copied in other liturgical books such as antiphoners, graduals, tropers and prosers, and in collections of musical treatises. Their terminology, of Byzantine origin, laid the foundations for the vocabulary of modal theory, in which the standard description of the church modes was developed in the 11th and 12th centuries through the division of the octave into a 4th and a 5th. In the absence of precise melodic notation, the tonary was indispensible to the memorization of the psalmodic endings for each of the eight tones. It was an important adjunct to the antiphoner during the latter's dissemination at the end of the 8th century, and its prolonged use in German-speaking areas until the end of the 13th century may be attributed to the continued need to memorize the chant.
1. Nomenclature and terminology.
2. Intonation formulae and model antiphons.
6. 11th- to 13th-century sources.
MICHEL HUGLO
The earliest tonaries, from the late 8th century, have no title but begin immediately with the title of the 1st tone (see below). Their contents came, however, particularly in Germany, to suggest a title. At first these varied: Toni (D-TRs 369, f.168v; Sl HB.XVII.17, f.227v etc.), Incipiunt toni (F-Pn lat.12584, f.216; Ps 1220, f.602v; I-Rvat Pal.lat.552, f.59 etc.), Incipiunt octo toni (the tonary of Regino of Prüm: CoussemakerS, ii, 3) or Incipiunt octo officiales toni (the tonary in the Winchester Troper: GB-Ccc 473, f.70v).
The term ‘tonarius’ appeared in the second half of the 10th century in the region of Lake Constance, in the abbey catalogue of Pfävers and in the Musica of Berno of Reichenau (GerbertS, ii, 63a). The term ‘tonale’ was used first in the Tonale sancti Bernardi of the Cistercians, and the terms ‘intonarium’ and ‘intonarius’ did not appear before the 13th century.
In the earliest tonaries, the first pair of tones have D as their final, and are both called protus, a term derived from the first Greek ordinal number. The pairs of tones with E, F and G as their finals are similarly termed deuterus, tritus and tetrardus respectively. Each pair contains one authentic (Lat. aut(h)enticus, aut(h)entus; see Huglo, 1974, pp.758–61) and one plagal (plagalis, plagis; from Gk. plagios) mode, which are distinguished by their higher or lower ambitus, respectively. The Greco-Latin terminology was current but unexplained in the first half of the 9th century; the first explanations appeared in the prologue of the Metz Tonary (see Lipphardt, 12) and in approximately 850 with Aurelian of Réôme (GerbertS, i, 40a).
Hucbald proposed a new system to replace this, according to which the tones were to be numbered from 1 to 8 (GerbertS, i, 119a): this terminology was adopted in all the Aquitanian tonaries, and, according to the anonymous early 11th-century Dialogus de musica from northern Italy (see Odo, §3), was by then an established custom (consuetudo; GerbertS, i, 259a). In the 9th-century Metz Tonary, no special terminology was adopted for the psalm tone endings (see §3 below), which were simply listed under the abbreviation Evovae (et in secula seculorum Amen). Later, however, a wide variety of terms was adopted as equivalents for ‘ending’: divisio, varietas, diffinitio, differentia, figura, modus, formula and so on (for further details, see Huglo, 1971, p.393).
Byzantine intonation formulae (enēchēmata), with nonsense ‘words’ set to them as identifications of the individual modes, are found in all tonaries until the mid-11th century, and in some as late as the 12th. These formulae end with long melismas on the ‘words’ noenoeane, for the authentic modes, and noeagis, for the plagal modes, and were introduced into Carolingian Francia. Medieval theorists were puzzled by them: Aurelian of Réôme questioned a Greek about them (GerbertS, i, 42), and the author of the Musica enchiriadis (GerbertS, i, 158b), Regino of Prüm (GerbertS, i, 247b) and Berno of Reichenau (GerbertS, ii, 77a) believed that they were meaningless syllables suited to vocalises, serving only to reveal the mode of a chant. In the 11th century, however, they were said to be derived from the Greek nous, ‘perception’.
Another method of modal identification lay in the use of model antiphons, introduced with the intonation formulae and ultimately displacing them. These antiphons, of unknown origin (they were not drawn from liturgical books), are based on New Testament texts as a literary elaboration of the numbers of the modes; this parallels their artistic elaboration in the decoration of Aquitanian tonaries or on the capitals that illustrate the tones in the abbey of Cluny and in Autun Cathedral: (1) Primum quaerite regnum Dei (Matthew vi.33)(2) Secundum autem simile est huic (Matthew xxii.39)(3) Tertia dies est quod haec facta sunt (Luke xxiv.21)(4) Quarta vigilia venit ad eos (Matthew xiv.25)(5) Quinque prudentes intraverunt ad nuptias (Matthew xxv.10)(6) Sexta hora sedit super puteum [common version] (John iv.6)Sexta hora ascendit in crucem [Aquitanian version] (Mark xv.25 or Matthew xxvii.45)(7) Septem sunt spiritus ante thronum Dei (Revelation iv.5)(8) Octo sunt beatitudines (Matthew v.3–11)Other texts are found in Berno of Reichenau (GerbertS, ii, 84ff), and at St Gallen and Augsburg (see Huglo, 1971, pp.234, 290).
These antiphons ended with the same melismas (neumae; see Neuma) as the intonation formulae; these served as touchstones by which the mode of a chant could be determined. In due course the neumae, otherwise known as caudas, jubili or stivae (PL, clxxxii, col.1130), disappeared from the tonary and came to be used in the liturgy at festivals, being limited to the Lauds and Vespers antiphons. Some 13th- and 14th-century motets were composed using the neuma as a tenor (see Huglo, 1971, p.338), and by the 16th century the neuma was played on the organ.
No examples of psalmody are given in the earliest tonaries, since the tradition was oral and was first committed to writing in the 10th-century Commemoratio brevis de tonis et psalmis modulandis (GerbertS, i, 213–29; ed. H. Schmid, Musica et Scolica enchiriadis, Munich, 1981, pp.157–78). Later tonaries usually include a psalm verse with notation, and sometimes also an example of the more ornate psalmody used for the Benedictus and Magnificat. These examples end with the first of the psalm tone endings (differentiae). Such endings were designed to smooth the transition from the psalm verse to the beginning of the repetition of the antiphon after the verse; accordingly, within each tone chants are classified by the endings best suited to them. This classification varies occasionally from one tonary to another, depending on the individual preferences of the compilers; the number of endings, too, is not the same for each tone and varies for any single tone from tonary to tonary. The order in which the endings are presented varies greatly. Early tonaries generally begin with the simplest and commonest endings, under which the largest number of antiphons is listed. In later tonaries the number of endings declined. Some later theorists established a rational order based on a musical principle: Berno of Reichenau classified the endings according to their proximity to the final of the mode (GerbertS, ii, 76a).
Among the psalm tones of the tonaries, one, the tonus peregrinus (‘wandering’ or ‘alien’ tone), is unlike any other: it has two recitation notes, whereas the regular eight tones have only one each. The theorists found difficulty in fitting it into the scheme of eight tones; it is often found after the 8th, but less because it has any connection with that tone than because it is foreign to Gregorian psalmody (see Huglo, 1971, pp.394–6, and J. Claire: ‘The tonus peregrinus: a Question Well Put’, Orbis musicae, vii, 1979–80, pp.3–14; see also Gallican chant, §10).
The tonary is an index of the psalm tone endings best suited to individual chants, and might thus be expected to be limited to antiphons of the Office and Mass. Nevertheless, responsory verses possess stereotyped formulae, varying according to their mode, which constitute a sort of very ornate psalmody; accordingly, the 9th-century tonary of Metz (see Lipphardt) and the tonary of Regino of Prüm (GerbertS, i, 231a) include some examples of responsories for each tone.
It is more difficult to understand the inclusion of graduals, alleluias and offertories of the Mass in the earliest of all tonaries, the late 8th-century tonary of St Riquier (ed. Huglo, 1971, pp.26–8). Not all the chants in the Gregorian repertory imply a dominant–final relationship (these elaborate non-antiphonal chants are used, in any event, only as illustrations of the division of the repertory into eight tones).
Individual chants are not assigned in every tonary to the same mode; in some cases the differences arise only because the chants may be assigned to an authentic mode in some tonaries and to a plagal mode in others, but in some cases there appears to have been some uncertainty over the choice of the modal pair (see Huglo, 1971, pp.402–9). But the schemes of classification in the earliest tonaries provide clear evidence of the conception of modality in the Carolingian era, if mistakes in classification (e.g. arising from errors in the transcription of the letters or numbers referring to the mode) are ignored. This Carolingian conception of modality was rather different from that which became universally established in the 13th century.
The earliest surviving tonary is that of St Riquier in the late 8th-century Psalter of Charlemagne (F-Pn lat.13159; ed. Huglo, 1952, pp.225–7; and 1971, pp.26–8); it contains only the first five tones and no Office chants. The tonary of Aurelian of Réôme dates from about 850 (GerbertS, i, 39–59; ed. Gushee). The tonary of Metz or Carolingian Tonary, which, according to Lipphardt, was compiled about 830, survives in four manuscripts: F-ME 351, copied in about 878 (ed. Lipphardt); D-W Helmst.1050 (catal.1152), of the 11th or 12th century, in which the chants are listed under each ending according to their position in the liturgical year; D-BAs lit.5 (Ed.V.9), from Reichenau, of the year 1001 (text collated by Lipphardt); I-Rc 54, from Nonantola, of the 9th century (analysed by Huglo, 1968, pp.22–8): this tonary, a copy of the Reichenau tonary, was probably brought to Italy by Jean Philagathos, private tutor to Otto III, who was named Abbot of Nonantola in 982 by the Empress Theophano.
The 9th-century tonary of the group of churches centred on St Denis and Corbie may be reconstructed from marginal letters opposite the introit and communion texts in three early graduals (see Huglo, 1971, pp.94–101): F-Pn lat.12050 (ed. R.-J. Hesbert, Antiphonale missarum sextuplex, Brussels, 1935/R, pp.CXXIII–CXXVI); the gradual from a private collection in Paris (pubd as PalMus, 1st ser., xvi, 1955); F-LA 118. The tonary of Regino of Prüm dates from about 900; Coussemaker's edition of it (CoussemakerS, ii, 3–73) is based on B-Br 2750–65. Several tonaries date from the 10th century: the tonary of the Alia musica (ed. Chailley); the tonary common to the various Enchiriadis manuscripts (see Huglo, 1971, pp.66ff); an anonymous tonary in a troper and proser of St Martial de Limoges (F-Pn lat.1240, dated 933–6) which contains the Byzantine intonation formulae but no model antiphons as did the later Aquitanian tonaries (see Huglo, 1971, pp.146–7); and the late 10th-century tonary of Odo of Arezzo (see also Odo, §2 (see also Huglo, 1971, pp.206–13).
From the 11th century the number of tonaries increased considerably and a division according to region occurred. Tonaries of an eastern group (from the German-speaking countries) represent a continuation of the Carolingian tradition – a complete tonary designed for practical use by singers, and covering the entire chant repertory; its use was greatest before the appearance of diastematic notation. Tonaries of a western group comprise mere abridgments, used for instruction, in which only a few examples from each chant category are cited to illustrate each tone.
Of the anonymous German tonaries, the most important is that in a Reichenau manuscript (D-BAs lit.5; see §5 above), in which the chants are classified alphabetically. This was probably the model for the tonary of Berno of Reichenau (d 1048) which survives in 17 manuscripts (list in Huglo, 1971, pp.266–7); Gerbert's severely mutilated edition of the Berno tonary quotes only the beginnings of the first six tones and omits the examples (see Huglo, 1971, pp.274–6; see Donato for edn after I-Rvat Pal.lat.1344). The tonary of Gondekar of Eichstätt (c1070) is partly derived from that of Berno; that of Frutolfus of Michelsberg (d 1103) is merely a slightly amended amplification of Berno's (ed. Vivell). Berno also influenced the Registrum tonorum of Udalscalcus of Maisach (d 1151), which survives in D-Mbs Clm.9921, W Gud.lat.4641 and in the fly-leaves of Melk.
Of the western tonaries the most homogeneous group is found in Aquitaine, and it may in turn be subdivided into a group from Toulouse and one from Limoges. The Toulouse tonaries survive in the following manuscripts: F-Pn lat.776, lat.1084, lat.1118, and GB-Lbl Harl.4951 (see illustration). Similar characteristics are present in a Narbonne tonary (F-Pn lat.780), a tonary in a troper from St Martial (Pn lat.1240), and some Paris fragments (Pn lat.7185). The Limoges tonaries are found from the 10th century in the following manuscripts: F-Pn lat.909, lat.1084, lat.1118, lat.1121, and US-BEm 88, from St Allyre, ?Clermont-Ferrand (RISM, B/III/4, 1994, pp.141–2). From these two sub-groups are derived the tonaries of F-Pn lat.7211 and n.a.lat.443, and of I-Nn VIII.d.14. Spanish tonaries were also linked with the Aquitanian tradition, although they are all short and there are complex problems in linking them with the earliest sources; they survive in E-Bac Ripoll 74 (10th-century) and GB-Lbl Add.30850 (from Silos).
Abridged tonaries, usually anonymous, appeared in the langue d'oïl regions north of the Loire from the 10th century, and increasingly from the 11th century (F-Pn lat.4995, f.38v; T 96; I-Rvat Reg.lat.1638, f.126, pp.315ff). The collected works of Odorannus of Sens (d 1046) include a tonary of this kind (I-Rvat Reg.lat.577; see Bautier and Huglo). The tonary of Guillaume de Dijon (990–1031), the instigator of a monastic reform, is probably the finest of the French tonaries; it survives in F-MOf H.159 (PalMus, 1st ser., viii, 1901–5/R) and is supplemented – for the Office chants – by a 13th-century copy at the beginning of an antiphoner from Fécamp (F-R 245 [A 190]).
In England short tonaries were sometimes transcribed in liturgical books (see §iii below). The most important English tonary, however, was the Sarum tonary (Tonale secundum usum Sarum et universalis ecclesiae, ed. Frere), compiled in the early 13th century like the Sarum gradual and antiphoner. This tonary is detailed, and preceded by remarks on theory; it corresponds in every respect with the Sarum antiphoner (AS, 1901–25/R), and contributed to the tradition of sound theoretical and practical training of cathedral choristers in England. Complete or in part it survives in numerous 14th- and 15th-century manuscripts (Longworth House, Longworth, Berkshire, the Denchworth Breviary; GB-En Adv.18.2.13a; Mayer Museum, Liverpool, 12016, originally from Cardington, Bedfordshire; Lbl Arundel 130; Occ 44; SB 175; I-PAc 98).
Intermediate groups of tonaries were borrowed at times from the eastern and western groups described above. These borrowings are found most notably in tonaries associated with the Liège school, and also in Swiss tonaries. The rise of the Liège school is reflected in anonymous treatises of the second half of the 9th century, such as the De musica sive modulatione cantus and, more particularly, the Quaestiones de musica (ed. Steglich), both of which contain musical formulae borrowed from tonaries. This Liège school had contacts with both Metz and Bavaria. Tonaries associated with it include that ascribed (apparently without foundation) to Sigebert de Gembloux (1030–1112; B-Br 10078–95), and that in the De musica of Johannes Afflighemensis, formerly regarded as of the school of Normandy (see Huglo, 1971, pp.299–301).
The earliest examples of Gregorian chant in Switzerland seem to belong to the western group described above, but the tonary of the 10th-century Hartker Antiphonary (CH-SGs 390–91: PalMus, 2nd ser., i, 1900/R; Monumenta palaeographica gregoriana, iv/1–2, Münsterschwarzach, 1988) contained intonation letters, a novel feature that later influenced tonaries in the eastern group. The six Latin vowels (i.e. including both i and y) and the two Greek long vowels are used to indicate the psalm tone appropriate to an antiphon, and the vowel for the tone is followed by one of the consonants b, c, d, g, h, k, p and q (rather than an ordinal numeral) to indicate the psalm tone ending to be used. This system of pairs of letters became traditional not only in Swiss tonaries, but also in antiphoners and graduals, and it came to be used in other regions influenced by Switzerland, especially Swabia and Bavaria, probably by way of Weingarten and Ottobeuren (an abbey with contacts with the Swiss abbey of Einsiedeln). Intonation letters ultimately spread as far as Kremsmünster in Upper Austria and to Monza (I-MZ C.12/75).
The short English tonaries transcribed in liturgical books include the Octo officiales toni in the Winchester Troper (GB-Ccc 473, f.70v). At the beginning of each tone there are verses, written by Berno of Reichenau for his tonary (GerbertS, ii, 84).
Since early tonaries are heterogeneous it is unnecessary to distinguish between those of secular and those of monastic origin, as must be done with antiphoners (indeed, many of those cited in earlier sections are monastic in origin). From the 12th century, however, the increasing systematization of the liturgy of the religious orders is reflected in the development of distinctive tonaries (see Huglo, 1971, chap.11). The Cistercian tonary, or Tonale sancti Bernardi, was a product of the liturgical reforms of Bernard of Clairvaux (1090–1153) and the consequent unification of chant in the Order. In it the number and complexity of the psalm tone endings were reduced. It represents a systematic application to the tonary of the Regule by Guido of Eu (CoussemakerS, ii, 150ff; ed. C. Maître, La réforme cistercienne du plain-chant: étude d'un traité théorique, Brecht, 1995, pp.108–233; and survives complete (GerbertS, ii, 265–77; PL, clxxxii, 1153ff), in abridged form and also in a form comprising only the notated sections and omitting the theoretical exposition (as in antiphoners such as GB-Lbl Eg.2977).
The Dominican tonary, included in the graduals, was based on that of the Cistercians but was not as systematic as the latter; it was sometimes preceded by a short prologue, Omnis cantus ecclesiasticus (see Huglo, 1967, p.124). Dominican antiphoners were preceded by an extract from the Tractatus de musica of Hieronymus of Moravia rather than a tonary. A similar prologue of the Omnis cantus ecclesiasticus occurs in several noted Franciscan books, but the Franciscan tonary is difficult to reconstruct because the archetypal Franciscan liturgical books do not survive. Some Augustinian manuscripts (not of the Premonstratensians) contain tonaries, but there is no evidence that any unified tonary was compiled for the Order. Only two Carthusian tonaries survive (Parkminster, Charterhouse A.33, and F-G 124), and they are rather dissimilar, possibly representing Carthusian practice before and after the reform of chant in that Order (see Becker).
Tonaries continued to be copied until the 16th century, but apparently became manuals for teaching the theory of the eight tones rather than practical manuals for oral instruction in the chant. In this field, however, they are of purely documentary interest since late tonaries are only abridgments, and they provide much less information about the development of the concept of modality than can be obtained from treatises. The main interest of the tonary lies in the earliest examples – for the light cast on the conception of modality held by the earliest orderers of the Gregorian repertory in the Carolingian era. Late tonaries are full of mnemonic devices and verses summarizing the rules concerning the intonations or psalmody for each mode; these appeared mainly from the 12th century, particularly in Germany. There are some tonaries in verse, for example that of Hugo Spechtshart (1332). Tonaries still occur in early printed books, such as that of Dietrich Tzwyvel (Tonarius qui vulgo Primum querite dicitur, Cologne, 1505 or 1515). See also Psalm, §II.
P. Jaffé, ed.: ‘Des Abtes Udalskalk … Registrum tonorum’, Archiv für die Geschichte des Bisthums Augsburg, ii/1 (1858), 68–78
W.H. Frere, ed.: The Use of Sarum, ii (Cambridge, 1901/R), 237ff [edn of the Sarum tonary]
F.X. Mathias: Die Tonarien (diss, U. of Leipzig, 1903)
R. Steglich, ed.: Die Quaestiones in musica: ein Choraltraktat des zentralen Mittelalters und ihr mutmasslicher Verfasser, Rudolf von St Trond (1070–1138) (Leipzig, 1911/R)
C. Vivell, ed.: Frutolfi Breviarium de musica et tonarius (Vienna, 1919)
P. Wagner: ‘Ein kurzer Tonar veröffentlicht und erklärt’, Gregorius-Blatt, liii (1929), 97–114
E. Omlin: Die Sankt-Gallischen Tonarbuchstaben: ein Beitrag zur Entwicklungsgeschichte der Offiziumsantiphonen in Bezug auf ihre Tonarten und Psalmenkadenzen (Engelberg, 1934)
J. Smits van Waesberghe, ed.: Johannis Affligemensis: De musica cum tonario, CSM, i (1950)
M. Huglo: ‘Un tonaire du Graduel de la fin du VIIIe siècle’, Revue grégorienne, xxxi (1952), 176–86, 224–33
K.-W. Gümpel, ed.: Hugo Spechtshart von Reutlingen: Flores musicae (Wiesbaden, 1958)
M. Cocheril: ‘Le “Tonale sancti Bernardi” et la définition du ton’, Commentarii cistercienses, xiii (1962), 35–66
J. Chailley, ed.: Alia musica: traité de musique du IXe siècle (Paris, 1965)
W. Lipphardt, ed.: Der karolingische Tonar von Metz (Münster, 1965)
C.T. Russell: The Southern French Tonary in the Tenth and Eleventh Centuries (diss., Princeton U., 1966)
M. Huglo: ‘Règlement du XIIIe siècle pour la transcription des livres notés’, Festschrift Bruno Stäblein, ed. M. Ruhnke (Kassel, 1967), 121–33
M. Huglo: ‘Un troisième témoin du “tonaire carolingien”’, AcM, xl (1968), 22–8
M. Huglo: Les tonaires: inventaire, analyse, comparaison (Paris, 1971)
R.-H. Bautier and M.Huglo, eds.: Odorannus de Sens: opera omnia (Paris, 1972)
M. Huglo: ‘L’introduction en Occident des formules byzantines d'intonation’, Studies in Eastern Chant, iii, ed. M. Velimirović (London, 1973), 81–90
M. Huglo: ‘Comparaison de la terminologie modale en Orient et en Occident’, IMSCR XI [Copenhagen 1972], ed. H. Glahn, S. Sørensen and P. Ryom (Copenhagen, 1974), 758–61
H.J. Becker: Das Tonale Guigos I: ein Beitrag zur Geschichte des liturgischen Gesanges und der Ars musica im Mittelalter (Munich, 1975)
L. Gushee, ed.: Aureliani Reomensis Musica disciplina, CSM, xiii (1975)
G. Donato: Gli elementi costitutivi dei tonari (Messina, 1978)
P. Merkley: Italian Tonaries (Ottawa, 1988)
M. Huglo: ‘Die Tonare’, Grundlagen und Ansätze der mittelalterlichen Musiktheorie von der Spätantike bis zur ottonishen Zeit, Geschichte der Musiktheorie, iv: Das Mittelalter (Darmstadt, 1999), 76–102