Note: The following text differs from the paper publication in some small details of layout, especially the numbering and placing of the examples, in order to facilitate on-screen reading.

Since the music examples are inserted as in-line images at their appropriate locations in the text, live links from text references are provided only where these occur at a greater than usual distance from the relevant examples.

Footnotes open in the window below this main text; bibliographical references also open in this same window, from which you should use your browser's Back button to return to the footnote text. Footnote reference numbers and other links remain red after they have been visited, and the footnote references also work in reverse (i.e. from footnotes back to main text reference), to facilitate navigation between the two windows. The Appendix opens into the present (i.e. top right) window. For ease of reference, paragraphs are numbered in [blue].

I am extremely grateful to Margaret Bent, Bonnie Blackburn and David Fallows for their suggestions and comments on this material, either during the conference or at draft stage, though this does not imply their agreement with all of the content. I am also grateful to the editors of the volume for permission to reproduce the article here.

[1] The purpose of this article is not so much to set out a concrete thesis, but rather to open up discussion in an area of notational history and editorial policy that has become curiously sidelined by convention. At issue here are not simply points of notational nit-picking, but also genuinely musical implications for how we think about the music on both local and structural levels, how we wish it to be performed, and how we might even begin to relate these questions to the evolution of stylistic expectation during the course of the fifteenth and early sixteenth centuries.

[2] Although the greater part of this discussion takes place more 'beyond' or around Okeghem's Ma bouche rit than actually on it, the original springboard was a tiny variant reading occurring at one point in this song, as transmitted in the Mellon Chansonnier (Example 1).1

[3] Lest my interpretation of the status of this variant be misunderstood, it should be made clear at the outset that it is almost certainly not an accurate representation of Okeghem's hypothetical original text, at least if its uniqueness among some seventeen manuscript and printed sources is in any way statistically significant. But it is nevertheless a rhythmically interesting reading, since it seems strongly to imply, for this source anyway – and bearing in mind the longer notational tradition of this kind of composite figure – that not only should the last three coloured minims (marked 'x' above) be construed unequivocally in sesquialtera, but by implication the preceding group probably should as well. In turn, this has repercussions for surroundinggroups in this piece, and the ways in which they may be intentionally differentiated from othergroups. This might then be seen as having direct corollaries in, at least, the notation of other pieces in this particular source, and the subject immediately opens into a wider debate on how one is to construe this particularcoloration figure more generally during the fifteenth century, and into thorny but fundamental questions of source-based versus work-based editorial decisions.

[4] Received wisdom on this figure, certainly in the English-speaking world, is heavily determined by Apel's account in his Notation of Polyphonic Music, especially the later editions.2 As a term, indeed, 'minor coloration' or minor color is, as far as I can see, a fabrication by Apel himself: previous writers on notation (e.g. Wolf, Riemann, Bellerman, Ernst Praetorius)3 have a number of examples which employ the figure, but the limited discussion tends to be focused on sixteenth-century theoretical writings, which emphasize its equivalence with the dotted figuration, though there is certainly acknowledgement that it grew out of earlier fifteenth-century sesquialtera practice. But the awkward question of how and when the changeover came about from true sesquialtera to dotted equivalence is distinctly glossed over – unsurprisingly, since it is clearly a subject not amenable to easy prescription. Over the last fifty years, however, Apel's tag, minor color, has seemed very convenient, since all editors and scholars of fifteenth-century music know that there does seem to be an identifiable figure there, waiting to be given a name. The trouble is that, in separating the figure out from other coloration usages the assumption was planted in students' and scholars' minds that it was a special case which needed a special (if historically fictitious) name. The more one looks at the evidence, though, the more it seems clear that for a large part of the fifteenth century there was nothing special about it at all, and its effective suppression in many editions today should rather be read as a symptom of the seductions of normalization for editors confronted with apparently random or inconsistent patterns of colour and dots in their sources, and their (understandable) desire to focus on a reductive Urtext policy rather than address the particular historical or musical nature of the variants. (Since this article is taking Okeghem as a certain point of departure, it is worth remembering, as an example of this assumption of normalization, that the recent edition of the Okeghem songs in the AMS Collected Works4 is but one instance where so-called minor coloration is always presented as dotted, without distinction as to source pattern, possible chronological considerations, or apparently any other criteria.)

[5] This said, it should be emphasized that after the turn of the sixteenth century the theoretical evidence, from both Italy and northern Europe, has moved fairly unequivocally towards the interpretation of individual colouredandgroups in minor prolation as generally requiring assimilation into a dotted rhythm. The Appendix below gives some sample texts taken from a number of the major names from this period: Pietro Aaron (1516–after 1545), Ornithoparchus (1517), Sebald Heyden (1540), Glareanus (1547), Zarlino (1558), even stretching to Thomas Morley at the end of the century (1597). A particularly consistent stance is taken by Aaron across three treatises: the passage from Book I, Chapter 36 of the Toscanello, along with the previous Chapter 35, prompted further discussion and some disagreement between Aaron and Spataro, as documented in their correspondence;5 Chapter 36 also raises the question of the related figure, here in major prolation, but which also presents a problem in minor prolation, to be discussed later below in relation to Ma bouche rit and an extract from Tinctoris's Proportionale. All of these sample sixteenth-century texts basically state that normal practice in minor prolation is to perform the colouredand by extensionfigure as equivalent to dotted; Glareanus and Zarlino go some way towards historicizing the issue, by acknowledging an earlier triplet version, but without providing much detail. Sebald Heyden (and some others not cited here) explicitly make a distinction between the single figure and a longer, coloured sesquialtera string, the latter usually calling for a real, sustained tripletization. One or two other writers from the early sixteenth century use the term 'imperfection by a quarter' (imperfectio (quo)ad quartam partem) to theorize the effect of the figure (e.g. Sebastian Felsztyn in his Opusculum musicae mensuralis of 1519),6 emphasizing the relationship to the normal imperfection by a third resulting from coloration of perfect note-values.

[6] I would draw especial attention to the relative lateness of these theoretical sources, when viewed from the stance of, say, the study of mid-fifteenth-century song. I have not yet found any treatise from before 1500 which unequivocally pre-echoes the early sixteenth-century statements; the nearest is perhaps an ambiguous sentence in Book 3 of Adam of Fulda's Musica (1490), which could perhaps be interpreted as indicating an ambivalence in the reading of colour in imperfect mensurations: 'Although in these places coloration also removes a third part in terms of beat, it nevertheless does not completely accomplish this in terms of rhythm.'7

[7] The most important and influential treatise from just before the turn of the century was, of course, Gafori's Practica musicae (1496),8 whose extensive sections on mensuration, coloration and proportion leaned heavily, though not uncritically, on the work of Tinctoris twenty or more years earlier. In Gafori's work, there is no explicit discussion of what we call minor coloration in imperfect prolation as a recognized unit with its own problems of interpretation, separable from other coloration issues. In common with Tinctoris9 and a number of other fifteenth-century writers he draws attention to the general problem of recognizing, when one sees a string of, say, five, six or seven black minims, which are to be read as minims under true coloration and which are semiminims (or, in Tinctoris's case, minims under duple proportion, since his linguistically pedantic side could not accept a value less than the minima.) But even here, neither Gafori nor Tinctoris takes the opportunity to discuss explicitly the particular figure of black, let alone the simplefigure by itself, though there is some more tangential treatment of the former in Tinctoris, which I shall come to shortly. It seems, simply, not to be regarded as a problem, not a special case worth separating from normal sesquialtera usage in minor prolation.

[8] There are, nevertheless, a number of musical examples in the Gafori Practica which employ the blackfigure en passant, aside from any theoretical elaboration of its meaning. In one or two cases it is not possible to determine whether the writer intends the figure to be read as triplets or dotted; but in a few others, the weight of probability is certainly in favour of triplets. In Book 4, Chapter 1, for instance (Example 2), the Tenor coloration figure seems unequivocally to wish to be read homorhythmically with the voidunder explicit 3/2 proportion in the Cantus.

[9] Gafori has just written a normal dottedin the Tenor, and if he had wanted to make a rhythmic distinction between Tenor and Cantus at the 3/2 passage he could/would certainly have written the dotted figure again. In Miller's 1968 translation of the treatise, however, the coloration is transcribed rather perversely as dotted, though Young's translation, published the following year, does indeed tripletize the coloration.10

[10] Example 3, taken from Book 4, Chapter 5, shows clearly a longer sesquialtera chain in the Cantus of the type which even the sixteenth-century theorists accept as triplets or 'hemiolia' proportion; but, more importantly for the earlier fifteenth-century context being examined here, the example shows that from a musical aesthetic or stylistic point of view, Gafori has no problem with simultaneous triplets and dots even as late as 1496.

[11] Example 4 is another instance, rather like the Tenor of Example 2, where a distinction between coloured and dotted figures is clearly made in the notation, which can be heard as having a real musical effect, even within this rather abstract context following passages of sesquiquinta (6/5) and subsesquiquinta (5/6) proportions, and especially with the more languorous coloration triplet syncopated over the 'barline'.

[12] Tinctoris, writing in the 1470s, uses the figure in minor prolation more sparingly even than Gafori in those musical examples which were obviously composed himself: indeed there seems to be only one unequivocal instance, in an example devoted to 14:5 proportion in the Proportionale,11 and here it is not by itself determinable which sense Tinctoris has in mind. But, as with Gafori, there is no indication in the theoretical discussion that this single unit is to be treated any differently from other coloration usages or longer sesquialtera chains; and in two instances, involving criticism of another composer's practice (Busnoys and Pasquin) there is a strong implication that he regards the triplet interpretation, that is true sesquialtera, as the correct one, not only for theunit, but for the first two elements of thefigure, and another related figure, the coloured(i.e. with the semibreve divided into its component halves). In Book 3, Chapter 4 of the Proportionale, Tinctoris takes Busnoys to task for what he regards as notational malpractice, apparently in a passage from his motet Animadvertere -- long thought to be from a lost work, but recently suggested as part of the anonymous Gaude celestis domina surviving in Vatican Library, Cappella Sistina MS 15.12

Ab hiis uero tribus pariter articulis Busnoys unicus dissidet, qui suas emyolias etiam per impletionem notarum designatas suppositione istius cyphre 3 iterum et iterum signat, ut patet in isto moteto suo Animadvertere:

The only composer, indeed, who disagrees equally with each of these three points [i.e. how, when and where proportions should be indicated] is Busnoys, who, having already indicated his passages of emyolia by filling in the note-heads, then proceeds to add the figure 3 repeatedly under them, as can be seen in this motet of his, Animadvertere:

In quo superfluus quia pro signo sufficit notarum impletio, diminutus quia licet signo cyphrali indigeret unica cyphra non satisfaceret, et inordinatus quia quod preponendum est supponit cunctis esse perhibetur.13

Here it is clear to all that this is superfluous, since as a means of indication the filling-in of the notes is sufficient; but also deficient, since, even granted the requirement for a numerical signature, one figure alone would not be enough; and misplaced, because what should be put in front he places underneath.

[13] The thrust of Tinctoris's triple criticism here concerns the use of the repeated 3s in addition to the coloration, since in his view the 3s are deficient as a proportional sign (which should more correctly be 3/2) and misplaced (should be in front of the relevant notes, not underneath them). But for our purposes it is the first of Tinctoris's points which is the most relevant, namely that the use of figures is in any case redundant, since the rhythm is already adequately expressed by the coloration alone. He is quite clear that this is indeed a true sesquialtera ('suas emyolias') – indeed, Busnoys may well have used the figures knowingly as a strictly inessential confirmation – and there is no possible doubt about the triplet interpretation of both the coloration units of the Tenor secundus. Gafori later refers back to Tinctoris's discussion in his Practica, is a little more indulgent of the composer's sins, and uses an identical rhythm, with a single initial 3 in front of the black notes, as illustration;14 but interestingly this example is taken from the Sanctus of Busnoys's L'Homme armé mass instead, so perhaps Tinctoris's misleading or confused Animadvertere reference had perplexed Gafori as well as recent scholars.

[14] A second instance of indirect discussion by Tinctoris of these kinds of coloration group occurs in the chapter of the Proportionale devoted to the superparticular genus (Book 1, Chapter 6). Tinctoris has just given advice on how one should discriminate between minims blackened under sesquialtera and those under dupla (i.e. semiminims), and then – importantly for our subject – how black minims placed in association with black notes of higher value completely adopt the coloration nature of those higher-value notes.15 He then gives an example from a mass by Pasquin – clearly a composer not well known to Tinctoris, for whom he has little regard either musically or notationally:

Nescio tamen quis Pasquin in plerisque passibus sue misse autenti prothi irregularis distonite omni arte ac melodia expertis quoad primum ab omnibus dissentit; nec mirum, nam et sibi ipsi in 'Cum sancto spiritu', quod valde ridiculum est, contrariatur, quom in exordio nobiscum, in fine autem contra nos taliter operatus sit:16

Someone called Pasquin, however, in a considerable number of passages of his mass in the irregular authentic first mode – a work distoned17 and lacking all skill and melody – disagrees with everyone in the first of these respects. Nor is this surprising, since he even contradicts himself in the 'Cum sancto spiritu', which is quite absurd: for at the beginning his working is in line with the present writer, but at the end is in disagreement, thus:

[15] The example given is divided into two halves, and Tinctoris states that part of the absurdity of Pasquin's notation lies in its self-contradiction: 'for at the beginning his working is in line with the present writer, but at the end is in disagreement'. The opening shown is precisely that colouredwhich formed the first unit of the Busnoys example: Tinctoris is in no doubt as to its true sesquialtera nature. In the second half of the example, the solecism committed by Pasquin, in Tinctoris's eyes, is that the composer has assimilated the colouredfigure with its dotted equivalent, in the imitative exchanges between Discantus and Tenor. For Tinctoris, the blackpair should by rights be read in true sesquialtera (this first minim 'adopting the nature' of the semibreve coloration), and the final two black minims are coloured under dupla, in other words are semiminims. Tinctoris has nothing, notationally speaking, against this kind of coloration string per se, only against its mistakenly conceived rhythmical equivalence to the dotted unit.

[16] Of course, the very existence of the Pasquin example does demonstrate that such a notion of equivalence was in the air in the early 1470s; but it is significant, I think, that Tinctoris is obliged to dig out such an obscure work (even in his own eyes) to make his point, whereas innumerable examples of the blackfigure, not necessarily in such clear juxtaposition to the dotted equivalent, could be found in many works at his fingertips. The implication is that, in the early 1470s anyway, Tinctoris is assuming and recommending that such coloration is precisely what it says it is. There is no special case to plead.

[17] So far in this article the argument has been working backwards, from a chronological point of view. But the citation of thefigure in Busnoys and Pasquin, closely related both notationally and historically to thefigure, gives a neat opportunity to return to the earlier history of these kinds of local rhythmical decoration in the first half of the fifteenth century.

[18] At risk of slightly over-simplifying the historical process, one can detect two principal strands of notational practice which fed into the evolution of the various coloration figures under discussion. There is an Italianate strand, reaching back to the advanced stages of trecento divisio notation and its later assimilation of French practice in the early fifteenth century. Here the insertion of brief sequialtera or triplet passages becomes a very common, almost clichéd, decorative gesture: Example 5, taken from O spirto gentil, tu m'ay percosso by Prepositus Brixiensis, after Oxford, Bodleian Library, MS Canon. misc. 213, has been chosen to show its use in a piece that employs flagged semiminims in white notation, emphasising again the differentiation of sesquialtera minims and true semiminims.

[19] This kind of sesquialtera decoration is also very common in early fifteenth-century French music, as represented again, for instance, in Canon. misc. 213. The other strand is an English one, visible through Dunstaple, Old Hall, and the mid-century carol manuscripts, involving not only the 'classic'pairing but the particular cadential gesture still sometimes known as the 'English figure', though certainly traceable in other corners of the repertory. This involves, usually in tempus perfectum, the colouredunit (already seen above in the later examples), followed either by a further trochaic pair of, or a straight binary pair of minims: an instance of the latter is given as Example 6(a), from Dunstaple's Puisque m'amour, after EscA.

(a) first main cadence (bars 5–7)

(b) final cadence (bars 31–3)

[20] In no theoretical source of this period, from, say, the late fourteenth-century Tractatus figurarum or the Berkeley treatise, through to Tinctoris's contemporaries such as Giorgio Anselmi, Guilielmus monachus, Ugolino of Orvieto, or Ramos de Pareia,18 can I find any hint that coloration passages of this sort are anything other than true sesquialtera. The(or) figure is particularly interesting from a theoretical point of view, since its correct interpretation depends on the coloured breve being considered perfect, subsequently imperfected by the two minims, and hence strictly to be construed in modern terms as (a) in the following, rather than the more common transcription as (b):

[21] This caused problems for some theorists in France and Italy, since they maintained that what had already been imperfected by coloration (i.e. the breve) could not be further imperfected. But in English usage, as studied in Old Hall, for example, by Margaret Bent,19 it is clear that in these circumstances the coloured breve here, as in some other coloration contexts, could indeed retain its perfect quality, within the time-span already reduced by a third; and it is this usage which transferred, for example, to the French chanson repertory, despite the theoretical discomfort. Example 7 shows an instance from Binchois's Dueil angoisseus, again after EscA (Burgundy–Flanders, late 1430s): here the use of the figure 3 makes it doubly sure that the singer knows that it is a true sesquialtera that is needed, even though Tinctoris, as we have seen in the Busnoys example, later showed his disapproval of this practice, in his eyes at once redundant and deficient, though to many at the time simply confirmatory.

[22] Examples 8 and 9 show some further examples of groups of two and threefigures in Binchois songs, which must certainly be tripletized: hence it is positively perverse not to tripletize the single groups as well, especially since the scribe is perfectly happy to distinguish the dotted version of the figure (Example 8a) when he wishes.20 (It may be noted at this point that Wolfgang Rehm's edition of Binchois is one of the only standard editions of fifteenth-century music that does systematically transcribeorcoloration groups as triplets.)21

(a) bars 21–5

(b) bars 32–6

(a) bars 19–22

(b) bars 28–33

[23] From a real musical point of view, the more one looks at, especially, the song repertory of the first three-quarters of the fifteenth century, the more clearly it emerges that composers, or scribes, or others responsible for the notational detail of manuscript compilation, often made a specific point of playing off the more languorous nature of these coloration figures against the tighter dotted equivalent. This is already implicit, in fact, in the use of the 'English figure' in Dunstaple: Margaret Bent has noted in her book on the composer that Dunstaple never uses the figure at a final cadence;22 and his song Puisque m'amour is a good example of how the 'looser' – or perhaps in some sense 'weaker' – sesquialtera figure articulates the first main cadence (Example 6a), whilst the final cadence is tightened up with the explicitly dotted version (Example 6b).

[24] Once one begins looking for more subtle instances of this kind of musical nuancing, they are not hard to find. The final cadence of Dufay's Helas mon dueil, as transmitted in its only source, Porto 714 (Oporto, Biblioteca Pública Municipal, MS 714), is a case in point; and the citation of this source is additionally interesting in such a context, since the manuscript seems to have been compiled under the English supervision of Robertus de Anglia in Ferrara in the 1460s.23 Example 10 gives the final cadential melisma, which starts at the bottom of the contour with a dotted figure exchanged between Tenor and Superius; then in the middle of the contour the rhythm is stretched in both voices into the more languorous sesquialtera group; then it returns via the intermediary of a dotted sesquialtera group to the tighter, final dotted cadence.

[25] Example 11 presents two extracts from (probably) Binchois's rondeau Comme femme desconfortee, as notated in the Mellon Chansonnier (New Haven, Yale University, Beinecke Library for Rare Books and Manuscripts, MS 91).24 Towards the end of Example 11a is a very striking homorhythmic moment at the words 'Desire la mort': the scribe of Mellon is in no doubt that the particular sound of the drooping coloration triplets in all voices is what he wants at this place; and by extrapolation there seems no reason to doubt that the distinctions made in Mellon elsewhere in the piece between colour and dots mean exactly what they say. (See, for instance, the preceding phrases in the superius at Example 11a). If this is so, and bearing in mind previous comments relating to the Gafori and Binchois examples, we should be particularly alert to the fact that the Mellon scribe has no problem with simultaneous colour and dots in the phrase leading up to the main medial cadence (Example 11b); it should be noted also, incidentally, that this provides yet another instance where the coloration in the superius and contratenor 'tightens up' to dots at the cadence itself.

(a)

(b)

[26] It is hardly a coincidence that I have all but juxtaposed in this discussion Tinctoris's views on these coloration groupings and a belief that the texts of Mellon seem to wish to be interpreted by and large with true sesquialtera coloration. The relationship between Tinctoris and Mellon has been pondered by a number of scholars for some years now;25 and whilst it is inappropriate to digress too far in this direction, it is worth putting on record that the possibility that Tinctoris was in fact the scribe of the manuscript can by no means be ruled out. One intriguing part of the jigsaw, which deserves more detailed exposition elsewhere, is that the decorated initial letters of the last four songs in the collection, followed by the final composer ascription, can be read as 'manu Jo. tinctoris':

Ma dame de nom
Alas, alas, alas is my chief song
Nos amys, vo vous abusés
Virgo Dei throno digna

Jo. tinctoris

[27] I believe that, depending on one's exact terms of reference for calculation, the chances of this being coincidental are at least 5000:1. Although I do not wish to become distracted by this potentially fascinating digression here, it is worth flagging up the fact that if Tinctoris were indeed the scribe, and not just compilator in a more general sense, it raises very intriguing questions about the nature of textuality in this period, the extent to which composers' musical texts could be considered by their peers to be authoritative, and the extent to which a theorist of Tinctoris's stature regarded it as proper, or on the contrary ultra vires, to re-edit these texts according to his own beliefs.26

[28] Examples 12 and 13 show two further instances, chosen more or less at random, from other songs by Okeghem. Example 12 is a phrase taken from the Tenor of Les desleaux ont la saison, as transmitted in Dijon, Bibliothèque Municipale, MS 517 some time in the 1470s.

[29] Read in the context of the foregoing, this seems to be another clear example of a fine nuancing of rhythm proceeding from two languorous triplet groups to a tightening-up into dots. The only other source for this song, the Laborde Chansonnier (Washington, Library of Congress, MS M2.1 L25 Case), in the third layer dating from the 1480s, gives dots throughout: far from indicating, however, that the coloration is simply an orthographic variant for the same rhythm (as many editors would maintain), this shows rather how the process of normalization was already setting in by this time, and that from a musical point of view, those finer rhythmical nuances from a slightly earlier period were no longer being cultivated.27 (David Fallows, in a recent provisional attempt to order Okeghem's songs chronologically, places both Les desleaux and Ma bouche rit around 1460.)28 The Dijon reading of Les desleaux also provides a good example of how the triplet figure is quite happy in a syncopated position (i.e. in transcription, starting on an off-beat quaver rather than a main crotchet beat). The more one looks, the more instances of this one finds in the mid-century French song repertory, and it is again a feature which goes back at least to Dunstaple, who uses it numerous times also in his mass music. It may, indeed, be one of several factors which led to the erosion of a true triplet interpretation, since it is even easier for this syncopated version to slip into an orthochronic, dotted pattern than the main-beat version.

[30] Example 13 shows the opening of Okeghem's D'ung autre amer from the Nivelle de la Chaussée chansonnier (Paris, Bibliothèque Nationale de France, Département de musique, MS Rés. Vmc. 57: Loire Valley, 1460s). This is an instance of the descending opening figure which became something of a cliché; even though the figure is later commonly notated as dotted, the Nivelle coloration is another straightforward example of the() figure, and there is absolutely no reason to doubt the intention of this source to have the opening sung as a true sesquialtera, tightened up to dots in the succeeding phrase.

[31] Finally, returning to the Mellon text of Ma bouche rit with which this article began, the logical extension of these arguments is that thefigure later in the song (Example 14b) should be read after the manner of Tinctoris's Pasquin critique, as a single, true sesquialtera group followed by two semiminims (or minims blackened under duple proportion), carefully distinguished from the orthochronic, dotted version occurring both earlier in the first half of the song (Example 14a), and immediately after this coloration group in Example 14b, leading to the repeat of the second section of the piece.

(a) bar 18

(b) bars 34–6

[32] Once one tunes in to the musical possibilities evoked by these kinds of differentiation, there are a number of songs in Mellon and elsewhere in which larger-scale structural concerns are played out. The detailed teasing-out of the implications of these for our reading of this repertory must await a future study. As one instance, however, the very opening song of the collection, Busnoys's Bel accueil, presents a rhythmical contour across the whole piece which is similar to that seen already within the final melisma of Dufay's Helas mon dueil. The opening, rising imitative passages consistently use the dotted figure in all voices; then, after the medial cadence the music breaks into triplet, 'minor' coloration figures, exchanged, again, among all voices (including, incidentally, a rhythm identical to the full version of the 'English figure'); and finally, towards the end, the music returns to the tighter dotted formulation for the final cadence. This is a straightforward and easily comprehensible tripartite contour evoked by the scribe, editor, or possibly the composer. (Mellon is, after all, a particularly strong source for Busnoys songs.)29 But in other pieces, rhythmical and relational patterns are evoked which do not always fit so comfortably with our own post-Enlightenment notions of rational structure.

[33] In the end, I hope that it is clear what I am suggesting and what I am trying to be careful not to suggest. I am certainly suggesting that much received wisdom in this area of notation has been fuzzy, and often too concerned with normalization based on either anachronistic criteria or a reluctance to acknowledge some of the subtleties of source patterning. As I outlined at the beginning, there is no doubt that a process of assimilation between 'minor coloration' and dots did take place over a period of time: the reasons for this are doubtless rather complex, and may stem from factors as straightforward as speed increase; or the possible normalizing effects of performance involving more than one to a part in repertories other than song, leading to practices which may have subsequently leaked back across generic boundaries; or less tangible movements in musical aesthetics, away from the more finely gradated rhythmic nuances of the early to mid-century – perhaps increasingly heard as over-fussy – towards a simpler, more orthochronic approach.

[34] This is not to suggest in any simplistic way that that the minor coloration versions always represent earliest and best texts: the source patterns, when they are more completely untangled, will doubtless display a much more multi-dimensional quality. A specific case of this occurs in Binchois's Mon cuer chante, where interesting coloration readings (both on main beats and in syncopation) occur in Pavia, Biblioteca Universitaria, MS Aldini 362 (probably dating from the 1460s in Piedmont or Savoy) in passages where Munich 902 – a source much closer, as we have seen, to the composer in the 1440s – presents simple minim movement or dotted figures.30 It is probably true, indeed, that in many cases, at the level of actual fifteenth-century performance practice in this song repertory especially, both coloration and dotted figures are only two essentialisations of a rather wider repertory of inégalité practice: David Fallows suggested as much twenty years ago in his dissertation on Robert Morton,31 though this intriguing avenue has yet to be pursued further. The musical realities of the fifteenth century in this respect, and the reasons for their evolution, are doubtless just as difficult to quantify and assess as, say, the assimilation of dots and triplets in the piano music of Schubert and Schumann. No doubt scribes and composers in the fifteenth century did make mistakes in their notation and understanding; no doubt they sometimes committed errors of musical judgement even within the terms of their own aesthetic criteria. But this is insufficient reason to sweep under the carpet all of the shades of musical meaning offered in the sources, and insufficient reason to deny the music some of its most characteristic subtleties simply on grounds of blunt consistency.