MCE 3: Loyset Compère, Ave virgo gloriosa (Galeazescha)

Edited by Daniele V. Filippi

The ascription to ‘Loyset’ is reported both in the index of I-Mfd 1, where it refers to the whole cycle (in its three-motet form), and over the first motet (I-Mfd 1, f. 171v). The attribution to Compère was extended to the whole eight-motet version in I-Mfd 3 based on the stylistic consistency of the cycle (see JEPPESEN 1931, FINSCHER 1954, and FINSCHER 1964).

I-Mfd 1, ff. 143v-149r (3 motets)

I-Mfd 3, ff. 125v-135r (8 motets)

Motets nos. 1, 2 and 4 are attested both in I-Mfd 1 (Scribe A) and in I-Mfd 3 (Scribe I), where they are part of an eight-motet cycle labelled ‘Galeazescha’ and provided with loco rubrics. The version of motet 4 (Ave sponsa verbi summi) in I-Mfd 1 is shorter. Besides this, the versions of the three concordant motets diverge as to many superficial traits (distribution of music on the openings; voice designations; text underlay; use of minor color; use of ligatures; insertion of accidentals; ways to indicate the correct mensural interpretation of values in ternary sections) and several textual variants of minor importance (esp. regarding rhythmic details and short ornamental notes); in I-Mfd 1 only, there remain at least three uncorrected errors (with as many visible corrections by a different, unidentified hand). Therefore, even though many of the differences may depend upon scribal habits or choices, there seems to be no particular reason to think that the version I-Mfd 3 derives from the one in I-Mfd 1, and even the direct descent from a common ancestor seems unlikely. The present edition adopts I-Mfd 3 as the main source for all the motets; except for an uncorrected error (added in a second moment), the source is substantially correct and reliable.

The cycle Ave virgo gloriosa, consisting, in the complete form attested in the third of the Gaffurius Codices (Librone 3), of eight motets for five voices, belongs to the ‘core’ group of the so-called motetti missales.[1]

Whereas the long version in Librone 3 has no attribution, the shorter, three-motet version in the earlier Librone 1 bears the attribution to ‘Loyset’ [Compère] both in the index and at f. 143v, over the first motet.[2] That the attribution to Compère had to be extended to the whole eight-motet version was suggested by Knud Jeppesen,[3] and later demonstrated by Ludwig Finscher on the basis of the stylistic consistency of the cycle;[4] all subsequent scholars have accepted the extended attribution.

The title/rubric ‘Galeazescha’, which, as we will see, opens up a thought-provoking connection with the Sforza court, appears only in Librone 3, at f. 125v, over the first motet. ‘Missa’ is probably implied, as happens for regular polyphonic masses in Librone 3 (see f. 99v: O venus bant) as well as in Librone 2 (e.g. at f. 56v: Io ne tengo quanto te); in fact, in the autograph index of Librone 1, Franchinus Gaffurius used the term ‘missa’ for two of the motetti missales cycles listed there.[5] It is worth noting, however, that the Galeazescha does not appear in Gaffurius’s comprehensive index of the masses contained in Librone 3.[6]

The short version of the cycle is included among the ‘motetti missales consequentes’ in the index of Librone 1, but the component motets lack the loco rubrics; in Librone 3, instead, the loco rubrics appear as follows:

1. Ave virgo gloriosa caeli iubar loco Introitus
2. Ave salus infirmorum loco Gloria
3. Ave decus virginale loco Credo
4. Ave sponsa verbi summi loco Offertorii
5. O Maria in supremo sita poli loco Sanctus
6. Adoramus te Christe (2.p. Virgo mitis) ad Elevationem
7. Salve mater salvatoris loco Agnus
8. Virginis Mariae laudes loco Deo gratias

At the end of the Bassus part in no. 5 (Librone 3, f. 131r) there is a ‘Verte folium’ (‘Turn the page’) rubric pointing to no. 6; moreover, no. 5 ends with breves (with fermatas), whereas all the other motets in the cycle end with longae; Nolan Gasser, therefore, considered the two items as one motet in three partes.[7] More important than classifying nos. 5 and 6 as one or two motets, however, is to acknowledge that they form a Sanctus–Elevation complex (with no. 5 loco Sanctus, the first part of no. 6 ad Elevationem and its second part, though not labelled as such, post Elevationem) accompanying the most sacred moments of the liturgy with a precious and virtually seamless layer of polyphony.[8]

The version of motet 4 in Librone 1 is shorter than the one in Librone 3: the section ‘Gaude, virgo […] Pietate solita’ is missing, and the motet ends at m. 39. Besides this, the versions of the three concordant motets diverge as to many superficial traits (distribution of music on the openings; voice designations; text underlay; use of minor color; use of ligatures; insertion of accidentals; ways to indicate the correct mensural interpretation of values in ternary sections) and several textual variants of minor importance (especially regarding rhythmic details and short ornamental notes); in Librone 1 only, there remain at least three uncorrected errors (with as many visible corrections by a different, unidentified hand: see the Critical apparatus). Therefore, even though many of the differences may depend upon scribal habits or choices, there seems to be no particular reason to see the version in Librone 3 as deriving from the one in Librone 1, and even the direct descent from a common source seems unlikely. The present edition adopts Librone 3 as the main source for all the motets; except for an uncorrected error (a superfluous rest added later in no. 3: see the Critical apparatus), the source is substantially reliable.     

The presence of the short version in the earlier manuscript, Librone 1, raises some difficult questions. As said, Gaffurius included it in the column of the index dedicated to the motetti missales, a strong sign of his persuasion that it belonged to this ‘genre’, in spite both of its shortness, compared to the other five cycles listed there, and of the absence of the loco rubrics.[9] Was the short version designed as such by Compère himself, possibly as a ‘kernel’ that was later developed into a longer cycle? The composite nature of the text (on which more below) makes it difficult to speculate about the ‘completeness’ of the short version.[10] A tiny musical detail, however, counters the idea that Compère was directly responsible for it: at the end of Ave sponsa verbi summi in Librone 1, the Tenor secundus re-enters after the rests only to sing the final longa at the unison with the Tenor primus,[11] instead of singing the entire last phrase together with his ‘twin’, as always happens at the conclusion of the previous two motets of the short version, as well as in all the other motets of the longer version in Librone 3 (as can easily be seen in the modern score; in the long version the two tenors keep alternating until they sing at the unison the final phrase, ‘Pietate solita’, at mm. 56–58). Rather than a composer’s choice, it looks like a manipulation by someone not entirely familiar with the distinguishing compositional feature of the cycle (the ‘split tenors’, on which see below), who carelessly truncated the longer version.

Perhaps the scribe who copied the short version in Librone 1 (Scribe A)[12] simply did not have a complete and reliable source of the cycle available. Gaffurius – who, as the inclusion of some attributions in the index of Librone 1 seems to prove,[13] was better informed than his scribe – might have known that the cycle was originally longer and that it belonged with good reason to what he called ‘motetti missales consequentes’. Therefore, he listed it as such in the index, and at the earliest convenience, years later, obtained a copy of the full version and had it included in Librone 3.[14] In the light, however, of what recent research on motet cycles has revealed about the flexibility of such works and their dynamic interaction with liturgy, it cannot be excluded that the short version was purposely designed or trimmed to only partially cover the mass ritual.[15]

* * *

As said, the title/rubric Galeazescha constitutes the most explicit association between the cycle and the Sforza court, namely that of Galeazzo Maria Sforza, who was duke of Milan from 1466 until he was murdered on 26 December 1476, and whom Compère served as a singer from 1474. The partial homonymy between Galeazzo and his son Gian Galeazzo (who succeeded his father as a child, first under the regency of Duchess Bona, and then under the progressive usurpation of his uncle, Ludovico il Moro), has given rise to alternative proposals,[16] but the reference to Galeazzo seems unquestionable, both in view of his own agency in matters of music and considering Compère’s chronology. We are, in any case, faced with two main sets of questions, one regarding the genesis of the cycle, the other its inclusion in the Libroni – that is, in sources belonging to Milan’s Cathedral, rather than to the ducal chapel in which Compère was a singer, and dating from decades after Galeazzo died and Compère left Milan.

The most straightforward hypothesis regarding the genesis of the cycle is that Compère composed it during his tenure at the ducal chapel while Galeazzo was alive, that is, between July 1474 and December 1476. Alternatively, he might have written it immediately after Galeazzo’s murder, but before leaving Milan, that is, between late December 1476 and February 1477 (or shortly thereafter).[17] The second option, according to which the Galeazescha was conceived right from the origin as a posthumous tribute to the duke and to accompany rituals for the repose of his soul, was advanced by David Crawford.[18] On the one hand, Crawford’s reading of the texts and overinterpretation of devotional commonplaces seems ‘fanciful’, as Gasser put it.[19] On the other, if the idea of a posthumous homage and musical intercession fits all too well in the picture of Galeazzo’s troubled succession, there is very little in the Marian texts compilation to suggest either political allusions or suffrage and mourning. There was surely no dearth of liturgical and devotional texts explicitly ‘pro defunctis’ with which to put together a cycle of commemorative Staatsmotetten.[20]

Paul Merkley surmised that the term Galeazescha ‘would mean “Galeazzo-like”, and this would suggest a later date of composition or compilation […] making it a product of Ludovico Sforza’s rule’.[21] As I have already argued elsewhere,[22] in order to expose the linguistic weakness of this argument it suffices to cite a document of 1472, according to which the biggest bombard of the Sforza artillery was named ‘Galiazescha victurioxa’ (‘the victorious Galeazzesca’).[23] As a matter of fact, the name itself cannot tell us anything precise about either the chronology or the function of the cycle – not to mention that we do not know whether the title/rubric was assigned to the cycle from its origin in the 1470s, or if it was a retrospective label attached to it at a later time.

Putting off, for a moment, the question whether and how the compilation of the text and the compositional design of the cycle reflect a specific commission by Galeazzo, or more loosely his taste, let us consider the problem of the cycle’s inclusion in the Duomo musical manuscripts. In general terms, the acquisition of Compère’s work is part of a larger transfer of sacred compositions from the environment of the ducal chapel to the Duomo. Various factors may have entered into play, including Gaffurius’s agency, but the close relationship between the Sforzas and the cathedral, and the frequent visits by the dukes and members of the household for dynastically relevant rituals and anniversaries, or on important feasts of the liturgical year, provide the most logical explanation.[24] But what was the point of having the cycle in Librone 1, finished in June 1490[25] – almost fourteen years after Galeazzo’s death and during Ludovico il Moro’s progressive ascent to power – let alone in Librone 3, surely compiled after 1500, and possibly as late as 1507, that is, in the first period of French occupation after the fall of the Sforzas?

Before trying to answer, we should at least spell out some assumptions that have perhaps been too easily taken for granted in previous scholarship. The first assumption is that inclusion in the Gaffurius Codices implied destination for performance: this is not necessarily self-evident, but recent research on the Libroni does encourage us to maintain the notion.[26] The second assumption is that the cycle’s special association with Duke Galeazzo was (and remained) recognizable, at least for a more or less selected circle of cognoscenti: I should once more remind the reader that the title/rubric in Librone 3 is the only hint of this association, which would be utterly impossible to guess based on Librone 1 alone, or on the cycle text. But after all, even allowing for the protean nature of motet cycles and their wider circulation, it is hard to contradict the idea that the motetti missales were pretty much Sforza-related:[27] can it be mere chance that the only motetti missales cycle in Librone 1 surely composed after Galeazzo’s death and by a non-member of the Sforza chapel, Gaffurius’s Salve mater salvatoris, is preceded in the manuscript by (and likely connected with) a motet praising Ludovico il Moro (Salve decus genitoris)?[28] After accepting the second assumption too, a final, perhaps smaller, ‘leap of faith’ is required: namely, to admit that a Marian cycle without any more specific ‘pro defunctis’ or dynastic connotation could be repurposed in order to commemorate Duke Galeazzo and pray for his soul.

Now, having gone through the necessary assumptions, let us try to answer our original question about the possible meaning of the Galeazescha in the different political scenarios of the 1490s and 1500s in Milan. In the late 1480s and the first half of the 1490s, Gian Galeazzo, son of Galeazzo, was the legitimate duke: the desire to commemorate his deceased father, at the same time honouring the (almost) namesake heir, is easily understandable. As the art historian Federico Maria Giani demonstrated, the Veneranda Fabbrica (the vestry board of the cathedral) and the duke partnered in cultivating the memory of Galeazzo at the Duomo during the early 1490s. On 8 January 1492, for instance, the deputies of the Fabbrica decided to bring to completion the altar of St Joseph, ‘started already since a long time for the devotion of the good and indelible memory of the most illustrious and excellent Lord Galeazzo Maria Sforza Visconti, father of the most illustrious and excellent Lord Gian Galeazzo’.[29] According to Giani, the officials of the Fabbrica used works in memory of Galeazzo (including the altar and its stained-glass window) as ‘a currency of exchange’ in their dealings with the éminence grise Ludovico.[30] Even after the death of Gian Galeazzo (21 October 1494) and Ludovico’s investiture as duke (26 May 1495), however, the ostentation of dynastic continuity was a means of legitimation for the former de facto usurper now become duke. Work on the altar of St Joseph, for instance, continued over the 1490s.[31] After Ludovico’s fall in 1499–1500, the memory of Galeazzo paradoxically retained the same meaning in terms of legitimation, but this time in a pro-French and anti-Ludovico perspective: this might help explain the inclusion of the cycle within the ‘Francophile’ repertory of Librone 3.[32]

An additional circumstance needs to be mentioned: Compère was part of the King of France Charles VIII’s retinue during his Italian expedition of 1494–95. The composer might easily have met with Ludovico and his wife Beatrice in Asti, where they went to pay homage to the king in September 1494, or in Pavia, when Charles VIII visited the moribund Gian Galeazzo on 14 October. Ferrante d’Este, who also participated in the French expedition, famously wrote to his father Ercole, duke of Ferrara, about a meeting with Compère in Casale Monferrato, reporting that the composer ‘was extremely sorry not to be able to furnish Your Lordship with any good compositions because the only works he has with him are old ones’.[33] As Joshua Rifkin further suggested, ‘the appearance of five motets and a Magnificat setting by Compère, all copied about 1495–7, in the Vatican choirbook I-Rvat C.S. 15 may represent a souvenir’ of the composer’s probable stay in Rome in January 1495, during the French occupation of the city.[34] Was the full version of the Galeazescha among the ‘old works’ in Compère’s trunk? Was it perhaps gifted to Gian Galeazzo or Ludovico? Might it have subsequently reached Gaffurius via the Sforza environment, to be included, years later, in the next big manuscript he prepared for the Duomo, under drastically changed political circumstances?

* * *

A new set of questions awaits us regarding what makes this specific cycle ‘Galeazzesque’. We may wonder whether its association with Galeazzo depends upon external circumstances, such as a special occasion for which it was composed and/or performed, or if it might be inherent in the design of the cycle, even though there are no explicit signs comparable to those in Gaffurius’s Salve decus genitoris for Ludovico il Moro or in other dedicatory works of the time. At the present state of research, the first hypothesis is at a dead end: in the absence of even indirect documentary evidence, any attempt to match the generically Marian text of the cycle to a precise occasion seems fruitless. In order to explore the second hypothesis, we should identify the distinguishing features of the cycle, the most striking of which is undoubtedly the ‘split tenor’.[35] Compère makes the two tenors alternate antiphonally without any overlap, except for the following passages, in which they sing together at the unison, as if to underline structurally and rhetorically relevant moments:

  • the final phrase of each motet;
  • the conclusion of the middle section in the loco Credo motet, no. 3 (mm. 65–75), underscored by fermatas at mm. 69–71;
  • the emphatic repetitions of the ascending formula d-f-g-a in the loco Sanctus motet, no. 5, always with at least one fermata: sung on ‘o Maria’ at the opening and at mm. 48–50, on different words at mm. 59–61 and at the conclusion;
  • the incipit and conclusion of the Elevation section in no. 6 (mm. 1–8 and 29).

Clearly, such a ‘strangely “unreal” five-part notation’, as Finscher put it,[36] conspicuously deviates from the ordinary writing in contemporary five-voice polyphony. The only real parallel is to be found, significantly, in two motets from another motetti missales cycle in Librone 1 attributable, with some caution, to the same Compère (Ave domine Iesu Christe),[37] whereas other five-voice works with alternating tenors are only superficially similar and originate from different compositional models.[38] Compère’s strategy is in several ways enigmatic, but a crucial aspect, not adequately elucidated by previous scholars, illuminates, if not its raison d’être, at least its working: namely, the inherent, indeed genetic association between this peculiar rendering of the tenors and their derivation from cantus prius facti.[39] Only some of the pre-existing materials have been identified so far (see below), but even a superficial study of the tenor parts reveals that they are constantly tuneful, rhythmically straightforward, neither treated as real cantus firmi nor heavily ornamented or paraphrased, and organized in short and often symmetrical-periodical phrases (see the synopsis of the tenor parts in the Appendix). The identified borrowings indicate that Compère deliberately maintained the melodic content and geometry of the pre-existing tunes; since most of the cycle texts derive from sequences and comparable metrical songs/poems, it is inevitable to surmise that even the corresponding unidentified melodies, with their remarkable phraseological structures, originate from the same repertory (or possibly from the bordering realms of cantio and lauda). Furthermore, the rhythmic and metric shape of the tenor phrases, often relying on simple patterns, might reproduce what was already traditional for the corresponding songs, whether in monodic rhythmicized performances or in simple polyphonic renderings.[40] The split tenors, then, are not merely the result of a special performance practice applied ex post facto to an ordinary four-voice motet, but rather the special projection-in-performance of a more far-reaching compositional project. It is by no means a normal tenor part that Compère decided to split in two, but rather one that sews together portions of carefully selected pre-existing melodies, retaining their tuneful character. From this exceptional tenor part springs the overall compositional concept of the cycle: the variable geometry of the tenor phrases, with patterns of repetition, variation, and periodicity, induces the antiphonal rendition, and the symmetries in the polyphonic complex partially reflect, partially contradict those of the tenor itself. As a result, the Galeazescha is an impressive tour de force in which the varietas of the polyphonic treatment and the adoption of unifying strategies counterbalance the quodlibetical tendency to sectionalization triggered by the tenor.

* * *

Similarly to those of Hodie nobis and Ave domine Iesu Christe, the text of the Galeazescha is an assemblage from various sources. Unlike Hodie nobis, clearly focused on a single feast (Christmas), the present cycle, though permeated by Marian devotion, does not have a clear emphasis on any specific liturgical occasion.[41] The sources are mostly sequences and rhymed prayers, all in verse (though with some interpolations that momentarily blur the metrical structure), and praising or invoking the Blessed Virgin. The only exception to both rules, that is, the only passage in prose and not addressed to Mary, is the prima pars of motet 6, ‘Adoramus te Christe et benedicimus tibi, quia per sanctam crucem tuam redemisti mundum’: needless to say, the reason is its purpose of accompanying the Elevation of the Host.[42] Whereas some source poems are used only once, others are quoted in two, three, or even four motets:

  • Ave virgo gloriosa in nos. 1, 5 (with interpolations), and 7;
  • Ave virgo virginum in nos. 2, 4, 6, and 7;
  • Ave caelorum regina in nos. 2 and 8;
  • Salvatoris mater pia in nos. 1 (with interpolations), 2, and 4;
  • Salve mater salvatoris in nos. 5 and 7;
  • Verbum bonum et suave in nos. 2, 4, and 5.

This finding alone gives an idea of the complex weaving of verses from different origins that characterizes the compilation. The texts have various metrical forms: the prevailing ones are verses of three paroxytone lines of eight syllables with a concluding proparoxytone line of seven syllables (3×8p+7pp, following Norberg’s classification);[43] similar verses of three lines (2×8p+7pp); and verses of three proparoxytone lines of seven syllables (3×7pp). There seems to be no inherent rationale in their alternation with other metrical forms.

As to the contents, like many other Marian texts the Galeazescha can be seen as a sort of troped macro-litany, with plentiful epithets, frequently introduced by such salutations as O, Ave, Salve, Gaude. The very beginning of the cycle is emblematic: the Mother of God is saluted as virgo gloriosa, caeli iubar, mundi rosa, caelibatus lilium, gemma pretiosa, virginale gaudium, florens hortus, fons puritatis. Further appellations range from the obvious mater, domina, and regina, to flos, lucerna, medicina, porta, refugium, salus, sponsa, stella, and to the biblical mater veri Salomonis and vellus Gedeonis. The emphatic repetitions of the invocation ‘O Maria’ constitute a unifying feature (see motets 4, 5, 7, and, with special rhetorical weight, the conclusion of motet 8). It is possible to identify recurring motives, all drawing from well-established Marian topoi (e.g. Mary’s intercessory power, her role as New Eve, the paradoxical combination of virginity and maternity, or the praise that the faithful owe to her), but without any clear progression or logic, except that of varietas and accumulation. On the other hand, a certain regularity in size (considering the quantity of text and the number of verses or blocks per motet) is a sign of the purposeful character of the assemblage. 

The cantus prius facti identified with certainty are in motets 1 (from the sequence Veni virgo virginum, sung to the melody of Veni sancte spiritus), 2 (from the sequence Ave virgo virginum, again sung to the melody of Veni sancte spiritus), 6 (from the melody of the sequence Mane prima sabbati),[44] 7 (from the sequence Salve mater salvatoris and again from Ave virgo virginum), and 8 (from the sequence Virginis Mariae laudes, sung to the melody of Victime paschali laudes).[45]

Furthermore, Patrick Macey proposed that the tenor parts at mm. 17–32 (‘Gaude virgo salutata…’) of motet 4 are based on the tune of the hymn Ave maris stella:[46] if that is the case, we should postulate a substantial paraphrase and an alteration of the tune’s geometry that makes it aurally unrecognizable. I rather think that Compère quoted a different sequence melody, possibly sharing the modal ‘mould’ with the hymn, or loosely deriving from it. Macey also suggested, more plausibly, that measures 25–28 in motet 1 (‘Dans fluenta gratiae’, immediately preceding the block sung on Veni virgo virginum / Veni sancte spiritus) are based as well on a phrase from Veni sancte spiritus.[47] Two other suggestions, in my view, are to be rejected. Dino Faggion, followed by Gasser, proposed that the tenors at the very incipit of the cycle are based on the opening motive of the Kyrie from the Gregorian Missa IX ‘cum jubilo’.[48] The notes do partially coincide, but it is a very common modal turn of phrase: and why should Compère heavily restructure a Kyrie motive in order to build a tuneful AAB melody and match it with the verse of a sequence? It seems safer to imagine that in this case too he found the melody ready-made in the sequence repertory. Finally, Finscher suggested that measures 53–58 in motet 4 (‘Et regnare fac renatos…’) are based on verse 3 of the sequence Stabat iuxta Christi crucem:[49] the melodic relationship seems vague enough, and I cannot find evidence that the sequence from which the text passage is taken, Salvatoris mater pia, was ever sung to this tune. Admittedly, however, Compère did not mechanically associate one text to one melody: for instance, he set verses from Ave virgo virginum to at least two different tunes (without mixing them in the same motet, though: the tune of Veni sancte spiritus is used in nos. 2 and 7, that of Mane prima sabbati in no. 6), and conversely employed the same Veni sancte spiritus tune for verses from different sequences (Veni virgo virginum in motet 1 and Ave virgo virginum in motet 2). The inherent interchangeability of sequence melodies undoubtedly poses further complications for the understanding of Compère’s behaviour and the identification of other cantus prius facti.

Even though only a minority of the tenor passages have been traced back to pre-existing materials, Finscher remarked that the Galeazescha ‘seems to be built upon pre-existent melodies from beginning to end’ and he tried to reconstruct some of the possible cantus prius facti.[50] Macey, in turn, noticed the ‘decidedly tuneful, foursquare and popular cast’ of several passages and suggested that at least four of them may derive from contrafacted ‘secular tunes’: motet 2, mm. 50–58 (‘Ave veri Salomonis…’); motet 3, mm. 76–95 (‘O domina dominarum…’); motet 6, mm. 31–65 (‘Virgo mitis, virgo pia…’); and motet 7, mm. 28–48 (‘Tu veniae vena…’).[51] If the four examples highlighted by Macey speak for themselves, in other cases it is the geometry of the melodic phrases that hints at pre-existence, rather than at Compère’s invention. See for instance mm. 5–33 in motet 5 (‘In supremo sita poli…’): the text, consisting of two quatrains, derives from the sequence Salve mater salvatoris; the melody in the split tenors does not resemble the standard one for the sequence,[52] but, significantly, in both verses it follows the same AABC pattern as the standard melody.

Generally speaking, four are the defining ingredients in the Galeazescha compositional project: the pre-existing tenor, its polyphonic projection, the mensural patterns, and the presence of recurring motives and blocks. As said, Compère stitched together a series of pre-existing, modally uniform melodies (in protus): their at times antiphonal or periodic phrase structures, common among sequences, possibly prompted the unusual splitting among two alternating voices, in the way just described.[53] The Cantus often echoes, amplifies or paraphrases the Tenor segments (see e.g. motet 8): in this and other ways, the pre-existing melodic substance permeates and organizes the polyphonic complex, engendering symmetries and repetitions, whereas in other moments it remains submerged and its symmetry is modified or contradicted (see e.g. motet 5, mm. 51–61 and 62–72, in which the repetition of the Tenor is partially ‘hidden’ by the change of metre, obtained through coloration, and by the different polyphonic construction). As it has often been observed, recurring mensuration patterns contribute to unify the cycle:[54]

motet signs
1 q 3    
2 u 3    
3 u2 3 u2 3
4 q 3 q 3
5 u 3 u color
6 p color p2 3
7 u 3 u 3
8 q 3 u 3

Notably, all the motets, irrespective of their initial mensuration, shift at a certain point to a sesquialtera proportion (except in nos. 1 and 2, the alternation is repeated twice).

As to the recurring motives and blocks, the most remarkable case regards the formula pointed out above discussing the passages sung in unison by the Tenors in motet 5. In fact, the ascending formula d-f-g-a and its descending contrapposto g-f-e-d (both contained, by the way, in the opening phrase of the Tenor in motet 1) occur in the following passages:

  • motet 4 (desc. form): mm. 38–39 (‘o Maria’);
  • motet 5 (asc. form): incipit in breves, mm. 1–4 (‘o Maria’); mm. 24–25 (variant) (‘o Maria’); mm. 48–50 (‘o Maria); mm. 59–61 (‘pietatis’); mm. 70–72 (‘sanctitatis’);
  • motet 7 (desc. form): explicit, mm. 61–62 (‘o Maria’);
  • motet 8 (desc. form): mm. 57–58 (variant) (‘o Maria’); mm. 68–69 (‘o Maria’); explicit, mm. 71–73 (‘o Maria’).

Furthermore, the descending form closes the Tenors in all the motets (except no. 5, closed by the ascending form, corresponding to the seamless passage from loco Sanctus to ad Elevationem discussed above).

* * *

Let us return, in conclusion, to the question about what makes this cycle ‘Galeazzesque’. Besides recognizing the quite extraordinary features of Compère’s compositional project, it is ultimately difficult to judge whether they have anything to do with Galeazzo’s own musical or spiritual predilections. In his article of 1996,[55] Macey discussed a passage from the Historia patria (publ. 1503) by the Milanese courtier and historian Bernardino Corio. According to Corio, a few days before his death Galeazzo ordered his singers to sing at daily Mass a ‘versiculo tolto ne l’officio dedicato a li defuncti [recte from the little office of the Blessed Virgin]: “Maria mater gratiae, mater misericordiae etc”’. Macey noticed that the corresponding stanza from the hymn Memento salutis auctor, or at least ‘Galeazzo’s verset’, recurs in the texts of several motetti missales cycles, including the Galeazescha (here, however, in a modified or ‘troped’ form in motets 7 and 8). Macey connected the emphasis on this verset with Galeazzo’s supposed special devotion to Our Lady of Graces.[56] Successively, however, Gasser somewhat weakened the case by demonstrating that this devotion was in fact widespread in Lombardy and hardly a prerogative of Galeazzo.[57] Corio’s colourful anecdote retains its suggestive power, but on a more factual level it is, for the time being, impossible to establish whether, or in which proportion, the cycle’s texts and cantus prius facti are of local origin or of French import. The motetti missales were, as said, certainly Sforza-related and rooted in aspects of Milanese spirituality and liturgy,[58] some of their texts undoubtedly circulated in Milan,[59] but the lack of contemporary Milanese sources with notated sequences, inter alia, complicates the identification and tracking of the Galeazescha raw materials.[60] Pending new discoveries, then, we cannot help but conclude that the exact reason for the cycle’s title/rubric, and the Galeazzesqueness of the Galeazescha, remain a tantalizing enigma.

* * *

As stated above, I have adopted Librone 3 as the main source for the present edition. In addition to the MCE Editorial methods, I have applied the following rules: if an accidental is missing in Librone 3 but present in Librone 1, I have included it in the score; similarly, if a portion of text is missing in Librone 3 but present in Librone 1, I have included it in the score in roman type; in both cases I have added an annotation in the Critical apparatus.

As declared in the MCE Editorial methods, the text underlay adopted in the score ‘should be regarded more as a proposal than as a prescriptive solution’. The treatment of text in Librone 3 is rather discontinuous: even in motets like no. 3, for instance, that have an unusually accurate underlay, there are inconsistencies and blanks, notably in the Altus (see the Critical apparatus). In motet 4, the melismatic Altus has virtually no text underlaid in the first half, and the Bassus seems, from a certain point, strangely out of sync with the other voices. In motet 5 Altus and Bassus have just the incipit and the last part of the text. In such cases, I have proposed text underlay without accounting for the details in the Critical apparatus (added text, however, is marked in italics in the score).

Milan, November 2020


A synopsis of the tenor parts in the cycle Ave virgo gloriosa (Galeazescha)

[if you do not see the pdf in the frame below, you can download it here]

[1] On the nature, origin, and cultural meaning of the motetti missales, and of motet cycles broadly considered, I refer the reader to the MCE General Introduction and to the literature quoted there.

[2] For the status quaestionis about the dating of the Libroni, see my ‘The Making and the Dating of the Gaffurius Codices: Archival Evidence and Research Perspectives’, in Reopening Gaffurius’s Libroni, ed. Agnese Pavanello (Lucca: Libreria Musicale Italiana, forthcoming).

[3] Knud Jeppesen, ‘Die 3 Gafurius-Kodizes der Fabbrica del Duomo, Milano’, Acta Musicologica, 3, no. 1 (1931), 14–28, at 16, n. 2.

[4] Ludwig Finscher, ‘Die Messen und Motetten Loyset Compères’ (Ph.D. diss., Georg-August Universität, Göttingen, 1954), 259–60; Ludwig Finscher, Loyset Compère (c.1450–1518): Life and Works, Musicological Studies and Documents, 12 ([Rome]: American Institute of Musicology, 1964), 92.

[5] ‘Salve mater salvatoris gaffori | cum tota missa’; ‘Ave domine Iesu Christe | cum reliquis totius misse’. See the digitization of the index in Gaffurius Codices Online (henceforth: GCO), Schola Cantorum Basiliensis, accessed 6 November 2020, and

[6] Masses stricto sensu, then. See GCO,

[7] Nolan Ira Gasser, ‘The Marian Motet Cycles of the Gaffurius Codices: A Musical and Liturgico-Devotional Study’ (Ph.D. diss., Stanford University, 2001), 310.

[8] See Agnese Pavanello, ‘The Elevation as Liturgical Climax in Gesture and Sound: Milanese Elevation Motets in Context’, Journal of the Alamire Foundation, 9, no. 1 (2017), 33–59; Motet Cycles between Devotion and Liturgy, ed. Daniele V. Filippi and Agnese Pavanello, Scripta, 7 (Basel: Schwabe, 2019), Introduction, 6.

[9] Rubrics which, in any case, appear in the other two cycles by Compère (Hodie nobis) or attributed to him (Ave domine Iesu Christe), but not in those by Gaspar van Weerbeke (Ave mundi domina and Quam pulchra es) or by Gaffurius himself (Salve mater salvatoris). See Daniele V. Filippi, ‘Breve guida ai motetti missales (e dintorni)’, in Codici per cantare: I Libroni del Duomo nella Milano sforzesca, ed. Daniele V. Filippi and Agnese Pavanello, Studi e saggi, 27 (Lucca: Libreria Musicale Italiana, 2019), 139–69, at 144 and 148–51.

[10] Gasser, ‘The Marian Motet Cycles’, 310, n. 158 observed that ‘The three motets copied into MilD 1 […] display considerable unity in terms of textual content, and thus form a successful cycle in their own right’ and that ‘the three motets are well unified in terms of text construction: all three are formed using a similarly diverse compilation pattern – in contrast to motet 3, which relies mainly on one source; moreover, all three motets contain a verse from the sequence, Salvatoris mater pia’. I do not see, however, text-based arguments solid enough to postulate the pre-existence of the short version.

[11] See GCO,

[12] According to the classification adopted in GCO-Inventory and in Martina Pantarotto, ‘“Scripsi et notavi”: Scribes, Notators, and Calligraphers in the Workshop of the Gaffurius Codices’, in Reopening Gaffurius’s Libroni, ed. Pavanello.

[13] See my introduction to MCE 2.

[14] Cristina Cassia, ‘Gaffurius at the Mirror: The Internal Concordances of the Libroni’, in Reopening Gaffurius’s Libroni, ed. Pavanello, suggests an alternative scenario. The short version in Librone 1 might simply be the result of an error in the copying process: the scribe inadvertently skipped motet 3, and after realizing his mistake decided to cut short motet 4 and move on to a different work. Some inconsistencies in the copy of motet 4, including the anomalous behaviour of the tenors noted above, might be explained in this sense. It strikes me as unlikely, however, that neither Scribe A nor, a fortiori, Gaffurius bothered to settle the situation. Scribe A, for instance, fixed another skip at f. 9v by sewing a slice of paper at the bottom of the page (see GCO, ‘Restoration of Librone 1 (2019)’, As to Gaffurius, he did not hesitate to add new works to the Libroni wherever empty folios were available, even if consecutive sections were to be scattered in the manuscript (see Pantarotto, ‘“Scripsi et notavi”’).

[15] See Gasser, ‘The Marian Motet Cycles’, 304: ‘Compère wrote all seven motets as a single cycle, likely during his tenure at Milan (1474–77) […] motets 1, 2, and 4 were later editorially singled out to form a separate motetti missales cycle – demonstrating the later flexible performance practice of the genre – before the entire cycle was reassembled for inclusion in MilD 3.’

[16] Gaetano Cesari proposed to associate the cycle with Gian Galeazzo: see Cesari, ‘Musica e musicisti alla corte sforzesca’, first published in Rivista musicale italiana, 29 (1922), 1–53, then included in the fourth volume of Francesco Malaguzzi Valeri, La corte di Lodovico il Moro, titled Le arti industriali – La letteratura – La musica (Milan: Hoepli, 1923) and published as an offprint from the same (ibid.). Already Finscher, Loyset Compère, 92, n. 15 refuted the hypothesis, in consideration of Compère’s departure from Milan shortly after Galeazzo’s murder, in February 1477. Other scholars suggested that the title deliberately referred both to Galeazzo and Gian Galeazzo: see Thomas L. Noblitt, ‘The “Motetti Missales” of the Late Fifteenth Century’ (Ph.D. diss., University of Texas, 1963), 36–37, and David E. Crawford, ‘Review of: Thomas Lee Noblitt - The Motetti Missales of the Late 15th Century’, Current Musicology, 10 (1970), 102–8, at 105–6. See also Agnese Pavanello, ‘The Non-Milanese Repertory of the Libroni: A Potential Guide for Tracking Musical Exchanges’, in Reopening Gaffurius’s Libroni, ed. Pavanello.

[17] See Lora L. Matthews and Paul A. Merkley, ‘Josquin Desprez and His Milanese Patrons’, Journal of Musicology, 12, no. 4 (1994), 434–63, at 449–52.

[18] Crawford, ‘Review of: Thomas Lee Noblitt - The Motetti Missales of the Late 15th Century’, 102–8.

[19] Gasser, ‘The Marian Motet Cycles’, 314.

[20] A remarkable illumination with Duchess Bona dressed as a widow and praying to the crucified Christ is included in one of her books of hours: see Edith W. Kirsch, ‘An Unpublished Book of Hours of Bona of Savoy, Duchess of Milan’, in Opere e giorni: studi su mille anni di arte europea dedicati a Max Seidel, ed. Giorgio Bonsanti and Klaus Bergdolt (Venice: Marsilio, 2001), 395–402, fig. 7.

[21] Paul A. Merkley, ‘Ludovico Sforza as an “Emerging Prince”: Networks of Musical Patronage in Milan’, in Music and Patronage, ed. Paul A. Merkley (Farnham: Ashgate, 2012), 255–70, at 267.

[22] Filippi, ‘Breve guida ai motetti missales’, 164, n. 125.

[23] Luca Beltrami, La Galeazesca vittoriosa: Documenti inediti sul ‘530’ delle artiglierie sforzesche, ed. Andrea Beltrami (Milan: Tipografia U. Allegretti, 1916), esp. 77.

[24] See Filippi, ‘Breve guida ai motetti missales’, 165–66; Daniele V. Filippi, ‘Where Devotion and Liturgy Meet: Re-Assessing the Milanese Roots of the “Motetti Missales”’, in Motet Cycles between Devotion and Liturgy, ed. Filippi and Pavanello, 53–91, at 67–73; Daniele V. Filippi, ‘Operation Libroni: Franchinus Gaffurius and the Construction of a Repertory for Milan’s Duomo’, in Sounding the Past: Music as History and Memory, ed. Karl Kügle (Turnhout: Brepols, 2020), 101–14; and Pavanello, ‘The Non-Milanese Repertory of the Libroni’.

[25] A few compositions were probably added later, but surely not those copied by Scribe A: see Pantarotto, ‘“Scripsi et notavi”’ and Daniele V. Filippi, ‘Gaffurius’s Paratexts: Notes on the Indexes of Libroni 1–3’, in Reopening Gaffurius’s Libroni, ed. Pavanello.

[26] See Codici per cantare, ed. Filippi and Pavanello; Filippi, ‘Operation Libroni’; and Reopening Gaffurius’s Libroni, ed. Pavanello.

[27] The relevance of this idea for wider historiographic controversies, notably those concerning Josquin, is spelled out in Joshua Rifkin, ‘Milan, Motet Cycles, Josquin: Further Thoughts on a Familiar Topic’, in Motet Cycles between Devotion and Liturgy, ed. Filippi and Pavanello, 221–336.

[28] See Cristina Cassia’s introduction to MCE 4.

[29] Archivio della Veneranda Fabbrica del Duomo di Milano, Ordinazioni capitolari, 4, f. 56v: ‘perfici debere altare sancti Joseph in prefata maiori ecclesia, iam longo tempore inchoatum ob devotionem bone et nunquam delende memorie illustrissimi et excellentissimi domini Galeaz Marie Sfortie Vicecomittis prelibati illustrissimi et excellentissimi domini Johannis Galeaz patris’. On the feast day of St Joseph in 1466 Galeazzo Sforza had made his official entry into Milan as successor to his father Francesco.

[30] See Federico Maria Giani, ‘Ricerche per l’altare di San Giuseppe nel Duomo di Milano’, Concorso. Arti e lettere, 7 (2015), 5–65, at 15–16 and the documents published in appendix.

[31] See again Giani, ‘Ricerche per l’altare di San Giuseppe’.

[32] See Pavanello, ‘The Non-Milanese Repertory of the Libroni’.

[33] Letter of 7 October 1494, edited and translated in Lewis Lockwood, ‘Music at Ferrara in the Period of Ercole I d’Este’, Studi Musicali, 1 (1972), 101–31, at 115 and 129–30.

[34] Joshua Rifkin, Jeffrey Dean, David Fallows, and Barton Hudson, ‘Compère, Loyset’, Grove Music Online, 2001, accessed 30 July 2020, The section on the ‘Life’ was originally written by Rifkin for the 1980 edition of the New Grove and then revised by Dean and Fallows.

[35] I borrow this expediently concise definition from Fabrice Fitch, ‘Loyset Compère and the Motetti Missales Cycle “Ave Domine Jesu Christe”’, Journal of the Alamire Foundation, 10, no. 2 (2018), 293–304.

[36] Finscher, Loyset Compère, 103.

[37] See MCE 2.

[38] See Finscher, Loyset Compère, 103, n. 36 and Fitch, ‘Loyset Compère and [...] “Ave Domine Jesu Christe”’, 300.

[39] Of course scholars of the Galeazescha noticed both aspects separately, but without drawing any meaningful link between them. Finscher, Loyset Compère, for instance, discusses cantus prius facti at p. 102, observing that the cycle seems entirely based on pre-existing materials, then moves on to illustrate the unifying devices used by Compère (including tonality and mensural patterns), and then touches upon the ‘amazing technique’ of the split tenors at p. 103. See also Fitch, ‘Loyset Compère and [...] “Ave Domine Jesu Christe”’, 300–1, who suggests a parallel between the split tenors in Galeazescha and Ave domine Iesu Christe and some contemporary pieces based on litany tones, but to draw a conclusion opposite to mine: namely, that the Galeazescha ‘originated as a four-voice piece’ and that the split tenors depended upon a ‘local tradition of performance practice’ rather than resulting from Compère’s ‘compositional decision’.

[40] See Marco Gozzi, ‘Sequence Texts in Transmission’, in Motet Cycles between Devotion and Liturgy, ed. Filippi and Pavanello, 157–87, and the literature quoted there.

[41] I will limit myself here to an overall view of the cycle texts. For the individual motet texts, see the commentaries by Eva Ferro in the present edition and, for a precise account of the text sources, the corresponding records in the Motet Cycles Database, Schola Cantorum Basiliensis, accessed 23 July 2020, The most substantial of the few still unidentified passages is ‘Tu veniae vena […] regno praesentemur’ in motet 7.

[42] The same text, from a well-known and multipurpose antiphon for the Holy Cross, is sung at the Elevation in the motetti missales cycles Ave domine Iesu Christe (see MCE 2.6) and Natus sapientia (see MCE 8.5). Furthermore, it is included in another Elevation motet from the Gaffurius Codices: the anonymous Bone Iesu dulcis Christe in Librone [4], ff. 76v–77r (see Motet Cycles Database, and GCO,

[43] Dag Ludvig Norberg, Manuel pratique de latin médiéval (Paris: A. et J. Picard, 1980).

[44] I cannot establish at the moment whether the sequence from which the corresponding text derives, Ave virgo virginum, was ever sung to this tune.

[45] See the melodies and annotations in the field ‘Cantus prius facti’ on each motet page in this edition, drawing (with some qualifications) from Patrick P. Macey, ‘Galeazzo Maria Sforza and Musical Patronage in Milan: Compère, Weerbeke and Josquin’, Early Music History, 15 (1996), 147–212, at 167–73 and 206–10, and from Gasser, ‘The Marian Motet Cycles’, 316–20. For a synoptic view, see the chart in the Appendix.

[46] Macey, ‘Galeazzo Maria Sforza and Musical Patronage in Milan’, 171–73, 208.

[47] Ibid., 206. I was not able to ascertain whether the sequence Ave virgo gloriosa, from which the textual passage derives, was ever sung to the Veni sancte spiritus tune, as happened with many other metrically corresponding sequences.

[48] Loyset Compère, Messe, Magnificat e Motetti, ed. Dino Faggion, Archivium Musices Metropolitanum Mediolanense, 13 (Milan: Veneranda Fabbrica del Duomo, 1968), p. vii; Gasser, ‘The Marian Motet Cycles’, 319.

[49] Finscher, ‘Die Messen und Motetten Loyset Compères’, 262 and 265.

[50] Finscher, Loyset Compère, 102. He did that, however, by stripping them of rhythmic values, which seems at least partially misleading in view of the wide (but still scarcely known) tradition of rhythmicized performances of sequences and neighbouring genres.

[51] Macey, ‘Galeazzo Maria Sforza and Musical Patronage in Milan’, 206–10. As Bonnie Blackburn kindly pointed out to me, the second passage has a resemblance to L’homme armé.

[52] As given, for instance, in Carl-Allan Moberg, Über die schwedischen Sequenzen: eine musikgeschichtliche Studie, Veröffentlichungen der Gregorianischen Akademie zu Freiburg/Schweiz (Uppsala: Almqvist & Wiksell, 1927), no. 10, or in Andrea Kovács, Monuments of Medieval Liturgical Poetry in Hungary: Sequences, 2 vols., Musica Sacra Hungarica, 1 (Budapest: Argumentum, 2017), I: 178–80.

[53] It is even possible that the alternating tenors reflect existing performance practices for sequences and related genres, but at the present state of research this remains merely a suggestion.

[54] See Finscher, Loyset Compère, 103 (but with several imprecisions in the table of mensuration signs), Gasser, ‘The Marian Motet Cycles’, 316, and Clare Bokulich, ‘Meter and the “Motetti Missales”’, in Motet Cycles between Devotion and Liturgy, ed. Filippi and Pavanello, 397–427, at 403 (taking into account only the shorter three-motet version of Librone 1).

[55] Macey, ‘Galeazzo Maria Sforza and Musical Patronage in Milan’, 159–66 and 175–80.

[56] Ibid., 155–59.

[57] Gasser, ‘The Marian Motet Cycles’, 315; Nolan Gasser, ‘“Beata et Venerabilis Virgo”: Music and Devotion in Renaissance Milan’, in Crossing Boundaries: Issues of Cultural and Individual Identities in the Middle Ages and the Renaissance, ed. Sally McKee (Turnhout: Brepols, 1999), 203–38, at 215–25.

[58] Filippi, ‘Where Devotion and Liturgy Meet’.

[59] Ibid., 80–89.

[60] On the other hand, no hint has emerged so far in favour of a northern origin for the Galeazescha comparable to the identification of the rare sequence In honore matris dei, apparently exclusive to Cambrai, for the cycle Ave domine Iesu Christe: see my introduction to MCE 2. For other examples of ‘northern ingredients’ in the motetti missales, see Felix Diergarten, ‘“Aut propter devotionem, aut propter sonorositatem”: Compositional Design of Late Fifteenth-Century Elevation Motets in Perspective’, Journal of the Alamire Foundation, 9, no. 1 (2017), 61–86, at 79–80, and Agnese Pavanello, ‘Praying to Mary: Another Look at Gaspar van Weerbeke’s Marian Motetti Missales’, in Motet Cycles between Devotion and Liturgy, ed. Filippi and Pavanello, 339–80.

Expand all
Go item
MCE 03-01 Ave virgo gloriosa caeli iubar
text/xml thumbnail
MCE 03-02 Ave salus infirmorum
text/xml thumbnail
MCE 03-03 Ave decus virginale
text/xml thumbnail
MCE 03-04 Ave sponsa verbi summi
text/xml thumbnail
MCE 03-05 O Maria in supremo sita poli
text/xml thumbnail
MCE 03-06 Adoramus te Christe : Virgo mitis
text/xml thumbnail
MCE 03-07 Salve mater salvatoris
text/xml thumbnail
MCE 03-08 Virginis Mariae laudes
text/xml thumbnail