The anonymous cycle Gaude flore virginali is one of the two motet cycles with loco rubrics related to the Mass liturgy entered in the manuscript Mus. Ms. 3154, known as the ‘Leopold Codex’, preserved in the Bayerische Staatsbibliothek in Munich.^[1] The work, transmitted only in this source, was copied in the first and older section of the manuscript (gatherings i-xviii) by the main scribe, including the subsequent motet cycle Natus sapientia.^[2] This section of the manuscript was originally copied and foliated separately, as was the second section, made from many originally independent fascicles (gatherings ixi-xlviii) bound together around 1511.^[3] The continuous foliation and numbering of pieces were added in the nineteenth century.^[4] Compiled over a period of time covering more than four decades (from around 1466 to around 1511) probably in Innsbruck by scribes associated with the church of St. Jacob and with the imperial court chapel,^[5] the manuscript collects a large repertory of sacred music (mainly Mass movements, Magnificats, hymns, sequences, and motets) and a few secular pieces, mostly handed down anonymously.^[6] A large number of scribes (forty-three scribes altogether) was involved in the copying especially in the second section of the manuscript. In the first part the scribe responsible of most of the copying work, whom Thomas Noblitt classified as Scribe A in his study and edition of this source, was joined by seven other copyists, whose interventions were limited to some additions to the main corpus of works.^[7]

By studying and identifying the paper watermarks in the manuscript, Noblitt proposed specific dates for the gatherings and hence established a relative chronology within the different parts of the manuscript. According to his research, Gaude flore virginali, like the following Natus sapientia, was copied on paper produced in 1476.^[8] One can thus assume that the copying took place around the same time.^[9] Scribe A organized his work relatively systematically by genre.^[10] It is not surprising, therefore, that he transcribed the two motet cycles with loco rubrics next to each other, beginning on the last verso of gathering iv and continuing on gathering v.

Gaude flore virginali and Natus sapientia are the only known fifteenth-century motet cycles explicitly, related to the Mass liturgy through paratexts, preserved in a source outside of Milan. The presence of loco rubrics as well as the inclusion of a section for the Elevation in consonant chordal polyphony make the function of the music as polyphony for the Mass manifest and directly connect the motets to the repertory of motetti missales in the Milanese Libroni. The shared features have led scholars to frame the Munich cycles historically in relation to the motetti missales and to assume for both a Milanese provenance or, alternatively, a direct relationship with the Milanese environment (at least in terms of knowledge of the local musical practice).^[11] Briefly discussed in the dissertation of Gerhard Croll on the motets of Gaspar van Weerbeke,^[12] the Munich cycles were first analysed in detail by Thomas Noblitt in his dissertation of 1963 focusing on the motetti missales repertory.^[13] Noblitt pointed out the connection of these cycles to Milan and proposed for both an attribution to Johannes Martini on account of stylistic and especially biographical details. Since Martini worked in Milan in 1474 for the Sforza chapel together with Gaspar van Weerbeke and Loyset Compère and he was supposed to have had close contacts with Innsbruck and with Paul Hofhaimer in particular, Noblitt considered him the most probable candidate for composer of the Munich cycles.^[14] Later, however, in his edition of Munich 3154, Noblitt withdrew his conjectural ascription.^[15] No further attributions have been proposed.^[16]

Gaude flore virginali consists of six motets, which partially bear loco rubrics similar to those attested in Loyset Compère’s cycles in the Milanese Libroni.^[17]

Although the first three motets were not supplied with rubrics, they were likely meant to be sung during earlier parts of the mass, possibly corresponding to the ‘loco Introitus’, ‘loco Gloria’ and ‘loco Patrem’ of the Milanese motetti missales.^[18] The longer motet no. 5 ‘loco Sanctus’ includes a section ‘ad elevationen’ as well as a section ‘post elevationem’, although neither was provided with a specific heading. In fact motet no. 5 incorporates two of the seven stanzas of the poem on which the motet cycle is based as well as an interpolated prayer for the moment of the Elevation, characterized by an emphatic chordal style (beginning with ‘Domine Iesu propitius’).^[19] The section serving as music ‘post elevationem’ contains the text of the stanza ‘Gaude humilis beata’. The longer motet thus corresponds to the Elevation-complex comprised by two motets (nos 5 and 6) in the cycles by Gaspar van Weerbeke and Loyset Compère included in this edition (MCE).^[20] Although structured as six instead of eight motets,^[21] the cycle Gaude flore with its internal subdivisions essentially matches the layout of the Milanese examples.^[22] This induced Noblitt to present the cycle in his edition as the Milanese cycles are presented in the Libroni, thus arbitrarily splitting the fifth motet into two and adding the rubric ‘Ad elevationem’ for the part covering the sections ‘Domine Iesu propitius’ and the following ‘Gaude humilis beata’.^[23] Noblitt’s presentation, however, obscures an essential aspect of the implicit performance practice, documented by this cycle and emerging from the analysis of the entire core corpus of the motetti missales. Namely, the consecration ritual was entirely encompassed by music, and the motets were meant to be sung continuously over the words of consecration, the Elevation, and the following prayers – regardless of their notation as visually separate or connected units or sections.^[24]

Unlike the motet cycles by Weerbeke and Compère in the Milanese Libroni, Gaude flore virginali was composed using a single poem in seven stanzas, to which two brief textual supplements in the form of invocations were added (at the Elevation, as mentioned, and at the end).^[25] This choice makes it clear that the multipartite work was conceived from the beginning as an unitary composition and that the motets were meant to be performed in the notated order. This is not necessarily true for the motet cycles that set various, distinct texts.^[26] In this regard Gaude flore virginali is close to Natus sapientia, which is also based on a single rhythmic poem: there too the stanzas, related to the horologia passionis, are not interchangeable with each other and imply a consecutive performance.^[27] Both Munich cycles therefore set in polyphony a single poetic text centred on a specific theme, suggesting that the idea of the motet cycle could have originated from that of the multi-stanza poem.^[28]

Ascribed to Thomas Becket, archbishop of Canterbury, Gaude flore virginali is structured as a sequence, in which the spiritual joys of the Virgin Mary are praised.^[29] According to rubrics recorded in various sources, its writing was inspired by the Virgin Mary herself: she appeared to Becket in order to give an account of her heavenly joys, in turn requiring adequate rejoicing by the faithful.^[30] The prayer became popular because of its ascription to the saint and the deep significance attributed to the contemplation of Mary’s joys: it was thought that, thanks to this devotion, the faithful would enjoy the presence and assistance of the Virgin Mary at the moment of death. Prayers praising Mary’s joys were also provided with indulgences.^[31] It is thus not surprising that the poem enjoyed a wide circulation as a sung prayer and that it was set to polyphony during the fifteenth century.^[32] Singing it at Mass, as in the case of the Munich motet cycle, added a more personal dimension to the priest’s celebration of the Eucharist, since, as underlined above, it was believed to have beneficial effects for the faithful’s own life and afterlife. The choice of Gaude flore virginali for the composition of this motet cycle, therefore, was probably not motivated by an association to a specific liturgical feast, such as the Assumption, as Thomas Noblitt proposed.^[33] Rather, the cycle provided music for a Marian prayer of particular significance that was appropriate to be recited or sung on many different occasions, especially within Marian services. The textual choice of the Munich setting discloses a direct connection with devotional practices, as attested in contemporary books of hours, and, more specifically, with the reciting or singing of ‘free’ prayers during mass.^[34]

According to research carried out by Daniele Filippi in Milan, the poem Gaude flore virginali is attested in a Milanese manuscript dated 1476, but in a very different version than in the Munich cycle.^[35] However, the poem seems to have enjoyed a wider circulation mainly in Northern Europe, particularly in England and in the Franco-Flemish area.^[36] Interestingly, the Munich setting contains a textual variant specifically documented in Flanders. As Felix Diergarten has pointed out, ‘Gaude humilis beata’, the initial verse of the sixth stanza, is found in the version of the prayer transmitted in books of hours belonging to Charles the Bold and to Jacques de Luxembourg, a knight of the Order of the Golden Fleece during Charles’s rule.^[37] This finding clearly places the Munich cycle in relation with the Franco-Flemish environment. If this suggests that the composer was Franco-Flemish, it becomes more difficult to maintain that the work had a Milanese ‘provenance’; at the very least, the connection needs to be plausibly explained.

There is thus a discrepancy between the textual evidence, which suggests a Flemish origin, and the musical evidence, in particular the Elevation section and the loco-rubrics, which connect the cycle to Milan. To explain this discrepancy, Diergarten has considered the idea that Gaude flore virginali might have been originated outside of Milan without the Elevation passage. He suggested that the Elevation music, considered as a ‘Milanese’ phenomenon, could have been added later to adapt the motet cycle for use at Mass in the manner of the Milanese motetti missales.^[38] However, neither the transmission nor compositional details provide any support for this tentative idea.^[39] Instead, recent research has shown that, within the motetti missales repertory in the Libroni, a substantial number of poems are directly related to the Franco-Flemish region.^[40] This gives greater credibility to the view that Northern practices of singing at mass, involving a polyphonic or multi-voice setting of various prayer texts, might have been important antecedents for the motet cycles by Franco-Flemish composers in the Milanese Libroni. This should also be taken into account when considering the provenance of Gaude flore virginale. Recent research has also revealed practices of singing at the Elevation in Franco-Flemish areas,^[41] allowing one to reasonably assume that the work might have been composed outside of Milan with the Elevation section.

Recent studies have underlined the textual and stylistic connections between the motetti missales by Compère and Weerbeke and genres such as sequences and hymns, associated with simple forms of polyphony.^[42] In this context, a setting such as Gaude flore virginali acquires a unique position: it provides an important example of the use of rhythmic prayers in longer multi-voice compositions sung during Mass, especially during the canon of the Mass, during which the words of the priest could not be heard by the congregation. The polyphonic setting of a sequence-like poem, in fact, suggests that, within the large repertory of sacred poetry broadly cultivated and employed in various functions, some prayers – detached from their original context, liturgical or otherwise – might have been used in Mass services long before the period in which the Milanese and Munich motetti missales originated.^[43] Pending further research, both Gaude flore virginali and the other Munich cycle witness an interaction between devotional and liturgical practices during mass that encourages an investigation of the phenomenon of the motetti missales from a broader, not strictly local perspective – especially when one considers the enormous musical heritage that has been lost, in Flanders in particular.^[44]

As discussed elsewhere,^[45] in the cyclic works of the Franco-Flemish composers active in Milan we find rhythmic patterns and melodic turns belonging to a long tradition of simple polyphony, associated with genres of rhythmic prayer such as hymns or sequences.^[46] Even though Gaude flore set a rhythmic prayer and is functionally similar to the Milanese motet cycles by Compère and Weerbeke, it embodies a very different compositional approach.^[47] Prominent features of the motet cycles by the aforementioned Franco-Flemish composers, such as phrases marked by clear-cut cadences, paired voices, extended homophonic texture with sesquialtera passages, and use of imitation are not characteristic of the Munich cycle, where a fully polyphonic, non-imitative texture prevails. Despite the rhythm of the poem, the cycle avoids a clear declamation of the text, and the musical flow is only rarely interrupted by full cadences as points of articulation. The polyphonic setting appears to be intentionally conceived in an ‘elevated’ style, turning away from the declamatory mode of polyphony echoing the recitation patterns of prayers. By contrast, the motet cycles by Compère and Weerbeke are more refined in their contrapuntal writing but at the same time purposely oriented to or inspired by recitation formulae and the forms of simple polyphony associated with their texts. The results of a stylistic comparison, then, call into question the idea of the motetti missales as a corpus defined by function and style. As Diergarten pointed out in his close and insightful examination of Gaude flore virginali, the concept of a ‘Milanese style’, generally used to refer to the core group of the motetti missales, is problematic in general and especially inappropriate for the Munich cycle.^[48] Not only is it inadequate to describe the cycles by Compère and Weerbeke, but it is also unhelpful to describe the broader Milanese musical tradition, which cannot be defined concretely or univocally.^[49]

In his dissertation, Noblitt provided an effective description of the stylistic features of Gaude flore virginali,^[50] pointing out the most relevant aspects mentioned above and highlighting the means used by the composer – in addition to the clefs, modus and time signature^[51] – to create relationships between the individual motets: for example, the long sustained sonorities in the openings and the common use of some motivic material. Even if no cantus firmus has yet been identified, Noblitt recognized the presence of at least one pre-existing melody, on which the composer based the Elevation section and which he reused in the last motet of the cycle.^[52] By arguing that the Elevation section was possibly added at a later point, Diergarten had to explain why the same melody was used in a different motet. He suggested the composer might have decided to employ a melody already included in the cycle in order to create a link with the music already written. Even if this cannot be ruled out, analysis of the cycle suggests a different explanation: that the last two motets were intentionally connected by this pre-existing melodic material for some special reason. Here the composer probably aimed to build stronger unity between the motets, in particular for the only motet that begins with a duo. More importantly, it seems intentional that the composer used this melody specifically for the two textual interpolations to the poem Gaude flore virginali, thus musically highlighting the more personal character of those short prayers.^[53]

Alongsite the stylistic features underlined by Noblitt and Diergarten, Gaude flore virginali is characterized by sometimes inelegant contrapuntal writing, with dissonant passing notes (even on a strong beat),^[54] awkward passages with ‘mi contra fa’ collisions,^[55] irregular or clumsy leaps^[56] as well as parallel motion in fifths and octaves.^[57] Such ‘irregularities’ occurring in the manuscript cannot be merely explained as copying errors; instead, they should be considered compositional features and are strongly related to a linear contrapuntal conception.^[58] In performance, this linear conception is compelling in its own fashion.^[59]

As a whole, Gaude flore virginali shows peculiarities and features that distinguish it from the adjacent motet cycle Natus sapientia, the compositional style of which reveals a more careful hand.^[60] The proximity of the two cycles in the manuscript, provided with rubrics referring to sections of the mass, and some additional common stylistic features led Noblitt and some subsequent scholars to believe that both motet cycles must be the creation of the same composer, but significant stylistic differences can be detected. The results of an experiment carried out by Cory McKay with Jsymbolic, involving the computational analysis of many compositional parameters, suggest that the two Munich cycles are the work of two different composers.^[61] Interestingly, the computational analysis suggests that Martini might in fact be the composer of Gaude flore virginali; the other cycle is classified as stylistically closer to the music of Josquin des Prez.^[62] While these attributions remain conjectural, it is intriguing that the former coincides with Noblitt’s earlier proposal. If his intuition was based mostly on biographical details, this result suggests it might also have more stylistic support than one might have thought. In the absence of firm data, the question of attribution will have to remain open, but the computational experiment warns us against assuming uncritically that both Munich cycles must be by a single composer. It should also be taken into account that Scribe A may have intended to arrange the contents of the manuscript systematically, which could also explain the juxtaposition of similar works by different composers.

The present edition differs from the previous one published by Noblitt in many aspects. In addition to maintaining the original note values and mensuration signs, this edition also retains the overall formal design consisting of six motets, as presented in the Munich source, without adding any loco-rubrics absent in that source. In the manuscript the full text is given mostly just in the Superius and in the Tenor voices, while the other voices have only incipits (with the exception of the full-texted Elevation passage). While Noblitt limited his edition to including the text present in the source, this edition provides all voices with text underlay.^[63] This decision was taken in order to facilitate a four-voice vocal performance and in consideration of the habits of the scribe, who often left the text setting incomplete in the manuscript. In the first section of Munich 3154 he mostly notated the text just in the Superius, in the Tenor in some cases, with the other voices receiving only incipits.

As stated in the Editorial Methods, the text underlay is to be regarded more as a suggestion than a prescriptive solution. Additional accidents have been added sparingly where considered necessary. Because of the peculiar writing habits of the composer, false relations occurring at many places are intentionally retained. Editorial interventions in the music are restricted to corrections of errors and listed in the critical apparatus.^[64]