> (I) Specified chromatic inflections in vocal sources
(II) The theorists' statements
(III) Instrumental tablatures
(IV) Doubling the subtonic
(V) General conclusion
(I) Specified chromatic inflections in vocal sources
This section is devoted to providing an overall framework for our study: when and where did explicit leading notes appear in Medieval vocal repertories? For what regions and periods is it possible to document the existence of implied leading notes? And when did explicit accidentals finally take over from them?
Of course, all these aspects will not yield any definitive answer to our general question. They will nevertheless allow us to set a general context, which is a prerequisite for our inquiry.
In spite of the fact that hexachordal signs (b molle or b quadratum) do appear quite often in medieval polyphony, apparently they were not used to specify leading notes earlier than the end of the thirteenth century:
ex. Adam de la Halle (1245/50-?1285/8), Bonne amourette
[See WILKINS, N. (ed.), The Lyric Works of Adam de la Hale (Chansons - Jeux-Partis - Rondeaux - Motets), American Institute of Musicology, 1967, p. 57, Nr. 14, bars 3-4. See also REANEY, Gilbert (ed.), Manuscripts of Polyphonic Music, 11th—early 14th century (=RISM B IV1), p. 400, who gives a transcription of the incipit in ancient notation (so to speak a "quasi-facsimile"), whereby one can indeed see both accidentals, creating the double leading note cadence.]
This example of Adam de la Hale is probably one of the very first instances of a double-leading-note cadence. Rarely to be found in the late-thirteenth century repertory, the formula was to grow in importance significantly during the fourteenth century, in France as well as in Italy. By that time, however, many a chromatic inflection already seems to be left unwritten, as shown, for example, by a comparison between the various manuscript sources of the celebrated Notre-Dame mass of Guillaume de Machaut. The half a dozen manuscripts through which the work has become known to us all disagree as to the number and placing of explicit leading notes.
[See VAN, G. de (ed.), Guglielmi De Mascaudio opera I: La messe de Nostre Dame, Rome, American Institute of Musicology, 1949, critical notes p. ii ff. Chromatic accidentals appear most sparingly in a French copy dating back to the 14th century (Paris, Bibl. nat. franç. 9221), whereas a late manuscript (Paris, Bibl. nat. franç. 1585, 15th century) is remarkably complete in this respect.]
Moreover, even if one takes into account every chromatic inflection in every one of the six manuscripts, a few will still remain missing, as shown in the following passage:
ex.: G. de Machaut, Notre-Dame mass, Kyrie II
[See ibid., p. 5/65.]
The same reasoning also holds for pieces by Italian composers of the same period. In the following example by Jacopo da Bologna, the middle voice (c natural) most probably lacks a sign of chromatic inflection, for the harmonic augmented fifth between the upper voices (c natural-g sharp) seems inconceivable in the context of the musical language of the time:
ex.: Jacopo da Bologna, Lux purpurata -- Diligite iusticiam
[See PIROTTA, N. (ed.), The Music of Fourteenth-Century Italy, 1967- (CMM 8), vol. 4, p. 41/23 (manuscript fragments of Padua, Biblioteca Universitaria, N° 684 et 1475.]
On the other hand, fourteenth-century England represents a special case, owing to the coexistence of two rather distinct styles: the former one is typically English, essentially homophonic, with few dissonances, and practically devoid of any explicit "attraction accidentals" ("attraction accidentals" should be understood as chromatic inflections whose purpose is to connect more closely two melodically related pitches -- typically an upper or lower leading note); as for the second style, it is very close to that of the continental composers of the same period, and displays all the complexities of French Ars Subtilior, including rather frequent signs of chromatic inflection.
To sum up, the fourteenth century displays an obvious development in the number of explicit leading notes. But it is no less obvious that there are already other, implied ones; yet to specify their exact number would be an extremely difficult task.
In the early-fifteenth century sources, English (Old Hall manuscript) as well as continental (Trent Codices...), one still can find explicit attraction accidentals; and it is also possible to show (applying the criteria that we have used above for fourteenth-century sources) that other ones are necessarily implied. Thus there is no break between the fourteenth and fifteenth centuries in this respect.
On the other hand, as the fifteenth century proceeds, the number of specified accidentals tends to decrease steadily (as Willi Apel and Margaret Bent have pointed out). It is difficult to say whether this trend corresponds to a mere simplification of the notation, or to some more fundamental change in the importance ascribed to attraction by the composers of the time (as Willi Apel suggests).
[See APEL, Willi, "Musica ficta", in Harvard Dictionary of Music, Cambridge, Mass., Harvard University Press, 1966, p. 466. BENT, M., "Musica Ficta", §1(i-iv), in The New Grove Dictionary of Music & Musicians (1980), t. 12, p. 802-806a.]
The Trent Codices
The case of the Trent Codices is especially interesting. These constitute the most important collection of pieces by fifteenth-century French and Italian composers -- to which have been added numerous English works. The seven manuscripts of this set are in fact composite, being formed by instalments which were originally independently circulated and copied, before they were bound into volumes -- and in a rather arbitrary manner at that: hence the many difficulties in trying to date them.
If we examine these Trent Codices from the point of view of attraction accidentals, we may observe the utmost rarity of the latter in the most recent instalments (after about 1450): we will not find more than one or two of them in each piece, often even none. But it would be premature to conclude that their disappearance is only a question of chronology. Indeed, among these manuscripts, one should distinguish two quite different types of sources: the former, large-sized ones, were intended to be used as performance material, and bear many signs of correction. The other, small-sized ones, are less accurate; they were devoted mainly to the conservation of the repertory, and served as models for further copying. The latter include the most recent codices (Trent, Nr. 88 to 91), i.e. precisely those which are most sparing in attraction accidentals. Hence the disappearance of these accidentals seems not to be properly speaking a matter of chronology, but one of notation technique.
To sum up, there seems to be little hope of unambiguously defining the moment at which attraction inflections disappear from vocal sources; the situation is more complex, and varies from one manuscript to the other, even if the overall tendency is clearly towards a decrease in the number of explicit accidentals.
Late-fifteenth century/sixteenth century
In the works of such late-fifteenth century Franco-Flemish composers as Regis, Barbireau or Ockeghem, one still can find a few explicit attraction accidentals, but such cases have become very rare:
ex.: Jean Ockeghem, Missa "Mi-mi", Kyrie
[See PLAMENAC, Dragan (ed.), Johannes Ockeghem, Collected Works, New York, 2 vol., 1959-66; vol. 2, p. 1/7-8. Chigi manuscript C VIII.234 (end of 15th century), and two other manuscripts of the Sistina. These three sources all have the same accidentals for this passage (see critical notes, ibid., p. xvi).]
By the way, it is interesting to note that this work could date back to the beginning of Ockeghem's career, as it was probably copied in a (now lost) manuscript of Bruges in 1475/76.
[See STRHOM, Reinhard, Music in Late Medieval Bruges, Oxford, Clarendon Press, 1985, p. 30]
On the other hand, attraction accidentals are virtually nonexistent in Ockeghem's mature period, as well as in the works of all great Franco-Flemish composers who succeeded him, up to the mid-sixteenth century. Exceptions to this rule are extremely rare; I found only one case for Josquin, in an Italian manuscript -- and none for all other contemporary Flemish composers. The following passage, a well-known piece by Nicolas Gombert, is no exception indeed, since the accidentals (f sharp and e flat, in the edition of Corpus Mensurabilis Musicae) do not appear at all in the facsimile of the original edition (Susato, Antwerp, 1545):
ex.: Nicolas Gombert, Le chant des oiseaux
[See SCHMIDT-GÖRG, J. (ed.), N. Gombert opera omnia, 1951-75 (=CMM 6), vol. 11, p. 2/42. Compare: SUSATO, Tielmam: Le dixiesme livre [des chansons] contenant la Bataille à quatre de Clément Janequin , Bruxelles, édition culture et civilisation, 1970 (Corpus of Early Music 11), fol. 4v-6.]
The reappearance of sharps during the sixteenth century
Italy was the place where explicit raising accidentals first reappeared. The most advanced author in this respect seems to have been Elzear Genet (alias Carpentras), a very neglected figure nowadays, but one of the greatest celebrities of his time. A master of the papal chapel from 1514 to 1521, he retired in Avignon (he had been born in southern France), where he supervised the publication of his works, which appeared in four volumes (1532-35). In these volumes, one can find a large number of explicit accidentals, intended as Picardy thirds as well as leading notes:
ex.: Elzéar Genet (Carpentras), Magnificat
[See SEAY, A. (ed.), Elziarii Geneti (Carpentras) opera omnia, American Institute of Musicology, 1972-3 (=Corpus Mensurabilis Musicae 58), vol. 4, p. 104/144-155.]
Other Italy-based composers were to follow Carpentras in this respect, albeit several years later. This is the case for Verdelot, Arcadelt or Costanzo Festa. At mid-century, with the advent of the chromatic madrigal, some composers tend to write down all necessary accidentals. But other authors (some of them very famous) still resist the trend until the end of the century (Willaert, Lassus, Palestrina).
Twenty or thirty years went by before the Franco-Flemish composers in turn adopted the Italian manner. In the North, the evolution starts as late as 1560, with authors like Vaet, Regnart, Claude le Jeune, Paschal de l'Estocart. From 1580 onwards, raising accidentals can be considered as entirely explicit in that repertoire as well.
During the whole period, England clearly represents a special case. The most important polyphonic manuscripts of the early sixteenth century (Eton Choirbook, Lambeth Choirbook, Caius Choirbook) still seem to be unaware of explicit attraction accidentals. Later on, the English manuscript tradition becomes very confused, in part because of the religious conflicts, which resulted in the loss of much mid-century material. The remaining sources (Forest-Heyther partbooks, Henrican partbooks, British Library MS Roy. Add. 74-77, Gyffard partbooks) underwent numerous emendations by later hands, including many additional accidentals. The dating of such alterations turns out to be a very difficult task, often even an impossible one. Important printed sources do not appear earlier than 1575 (Tallis and Byrd, Cantiones quae ab argumento sacrae vocantur, London), at which date the situation begins to improve again. In this publication of 1575, attraction accidentals are to be found rather frequently, sometimes in combination with simultaneous false relations (imperfect octaves) which soon became a distinctive feature of English music.
The period between 1530 and 1575 therefore remains a very obscure one as concerns the use of attraction accidentals in English music.
Continued: (II) The theorists' statements >>>
© Vincent Arlettaz
- last update: 27 December 2002 -
Vincent Arlettaz's homepage