Differently from most of the motetti missales cycles, the anonymous cycle Natus sapientia was not entered in the Gaffurian Libroni, but in the so-called ‘Leopold Codex’ (D-Mbs Mus. Ms. 3154), immediately following another missales cycle, Gaude flore virginali.^[1] The Leopold Codex, currently preserved in the Bayerische Staatsbibliothek, Munich, was apparently compiled in Innsbruck and is likely connected to the parish church of St Jacob.^[2] On the basis of an accurate study of the forty-five different watermarks appearing in the manuscript’s paper, Thomas Noblitt established that its compilation had stretched over a few decades, from around 1466 to around 1511.^[3] Conceived probably as a repository,^[4] this manuscript contains mixed repertory (mainly motets, masses, and Magnificats) entered by forty-three scribes.^[5] Its two sections were originally copied and foliated separately (respectively 1–200, of which 1–19 are now lost, and 1–297) and then bound together. In the nineteenth century, during the compilation of a catalogue of the music manuscripts held by the Bayerische Staatsbibliothek, Julius Joseph Maier added in blue pencil the piece numbering and a continuous foliation, to which we refer in the present edition.^[6] Only 26 out of 174 compositions included in the manuscript carry an attribution (to Guillaume Du Fay, Heinrich Isaac, and Johannes Martini among others),^[7] but a few more can be added through concordances in coeval sources.

The cycles Gaude flore virginali and Natus sapientia, both of which are unica and presumably to be ascribed to a single composer,^[8] are quite different in many respects from the missales entered in the Gaffurian Libroni;^[9] therefore it is not clear what kind of connection they might have with Milan.^[10] The Leopold Codex gatherings containing both cycles were dated through their paperʼs watermarks to 1476 and Gaude flore virginali and Natus sapientia were probably copied around that date.^[11] The Milanese missales cycles by Loyset Compère and Gaspar van Weerbeke were also composed during the 1470s: in fact, Compère’s cycles date around 1474–76 (at latest 1477 in the case of the Galeazescha),^[12] while those of Weerbeke are probably connected with his first stay in Milan from 1471–72 to 1481.^[13] It is unclear whether the anonymous composer of the Munich missales had a good knowledge of the Milanese ones, as surmised by Noblitt, or not: in fact, no objective element allows to establish a direct link between them.^[14]

Concerning their possible composer, Noblitt first proposed to ascribe the Munich cycles to Johannes Martini, who spent a few months in Milan in 1474 and was a singer in the Sforza chapel.^[15] This hypothesis was based on a few – and, admittedly, insubstantial – stylistic similarities between these cycles and Martini’s works. Moreover, Noblitt explained the presence of motetti missales in a manuscript compiled in Innsbruck with one or more possible trips to this city that Martini might have made to meet one of his friends, the organist Paul Hofhaimer.^[16] A few years later, however, Noblitt himself withdrew this conjecture without providing any explanation,^[17] and no other attribution has been advanced so far.

As the other missales by Loyset Compère and Gaspar van Weerbeke contained in the Libroni, Natus sapientia consists of eight motets:

Scholars usually agree on the eight-motet structure, with the only exception being Maier, who only counted seven motets, probably considering motet 2 as the second part of motet 1, because they both are composed on a single stanza of the text.^[18]

As the table shows, the logical sequence of the motets does not match their order of appearance in the manuscript: in fact, the motet ‘post Elevationem’ (no. 6) was copied before its correct position, on a folio previously left blank (f. 45r). This awkward placement is simply due to the fact that page layout is not consistent through the cycle but rather changes according to the length of each single motet.^[19] Indeed, longer motets require an entire opening, where two voices are accommodated on the verso of a folio and two on the recto of the following one, while shorter motets have all four voices one after the other on a single page. After copying motets 3 (on a single page) and 4 (on an entire opening), a single page was left over between them, and the scribe decided to fill it with motet 6, which precisely required that space.

As for loco rubrics,^[20] during the copying of the cycle, the scribe inserted only ‘loco Introitus’ and ‘loco Patrem’; ‘loco Offertorii’ and ‘post Elevationem’ were added later by another hand. According to Gerhard Croll, who considers this cycle specifically associated with the liturgy of the Passion, the second motet probably has to be considered loco Kyrie, since during the Lenten season no Gloria could be sung.^[21] However, this hypothesis has to be discarded, since no missales cycle with written down rubrics contains such an indication.^[22] In any case, according to the latest research, the correspondence between loco rubrics and liturgical moments was not meant to be precise: the music superimposed on the spoken mass was probably a continuous flow and the Elevation was the only moment in which the two layers were synchronized.^[23] Loco rubrics were thus only useful to orient oneself approximately within the mass; therefore, whatever label would have been implied for this motet, it makes as good as no difference from the point of view of performance. One could also wonder why the scribe, during the copying, inserted just two rubrics. In fact, the first, ‘loco Introitus’, is essential to immediately recognize that this cycle belongs to the category of motetti missales; the insertion of ‘loco Patrem’ in fact does not seem to have a particular meaning. The later addition of ‘loco Offertorii’ and ‘post Elevationem’ rubrics can instead easily be explained as a reminder, taking into account that the position of those motets in the manuscript does not correspond to their logical order within the cycle. Concerning the remaining four motets without loco indications, their function is clear from their position, and thus rubrics are unnecessary.

Differently from the texts of the missales by Compère and Weerbeke, which draw from multiple sources, Natus sapientia is entirely based on a hymn for the Short Office of the Cross, with the only interpolation being the antiphon ‘Adoramus te Christe’ at the end of motet 5.^[24] The hymn, presumably written in the first third of the fourteenth century and usually ascribed to Aegidius Romanus or Pope John XXII,^[25] was widespread, and is also attested in Milanese sources.^[26] Its text, written in Goliardic metre,^[27] contains hypermetric and hypometric verses. Since hymn melodies are usually strophic and are not very suitable for verses of different lengths, this hymn was probably more often read than sung, although recent research provides evidence of a few monodic intonations.^[28] In all the extant sources, the hymn begins with the slightly different incipit ‘Patris sapientia’, but ‘Natus’ was apparently attested in a fourteenth-century manuscript from Vienna, which unfortunately was stolen in the 1920s.^[29] However, it can not be excluded that, both in the Leopold Codex and in the Vienna manuscript, ‘Natus’ might result from a misreading of the initial capital ‘P’.^[30] The hymn Patris sapientia also constitutes the bulk of Loyset Compère’s Officium de cruce, a nine-motet cycle first published by Ottaviano Petrucci in Motetti B and then transmitted by other sources.^[31] Differently from the setting in the Leopold Codex, in Compère’s cycle the seven motets on the Patris sapientia hymn are preceded by two additional motets, of which the second was composed on the above-mentioned antiphon ‘Adoramus te Christe’.^[32]

Two major changes have been made in setting this text to music in the missales cycle: first, the initial stanza of the hymn is divided between motets 1 and 2, causing an awkward change in the original meaning of the text.^[33] Second, only the first seven stanzas, corresponding to the seven canonical hours, were set to music, while the last one, which contains an invocation to Christ, was left out. Moreover, motets 4, 7, and 8 lack a half or an entire verse at the beginning, which contains precise references to the hours, and in motet 5 the section ‘Adoramus te Christe’ was added to accompany the Elevation.

According to Noblitt, this cycle was performed on Maundy Thursday and probably also in votive masses during the year.^[34] However, given that the hymn Patris sapientia is attested not only in liturgical books, but also in prayer books, and that it was recited by confraternities during their gatherings, this motet cycle was probably also performed on different occasions in devotional contexts, for purposes of anamnesis and recollection.^[35]

Beside the text, the unity of the cycle is also achieved through the same head-motif (c''-b'-g'-a'-g') in motets 1, 3, 5, and 7, and similar ones in nos. 2, 4, and 6, and by recurring harmonic and/or melodic patterns.^[36] No cantus prius factus, however, has been detected so far.^[37] With respect to the style of the Milanese missales, Natus sapientia presents several differences: imitation is more rarely used, and mostly the pieces are built on full-voice counterpoint, with fewer changes in texture and a reduced number of duos.^[38] Moreover, while the other missales that include sesquialtera sections always end this way, motet 3 of Natus sapientia begins in a ternary meter but finishes in a duple one.^[39]

The cycle Natus sapientia was entered by the anonymous scribe labelled by Noblitt as ‘A’, who is also the main scribe responsible for the first part of the manuscript; recent research suggests that he has possibly to be identified with Wolfgang Unterstetter, who worked as choirmaster in St Jacob from 1466 to 1478. ^[40]  As for most of the pieces contained in this section, Scribe A provided the text only in the Cantus and left the other voices textless. The copyist’s work is quite inaccurate, as shown mainly by the frequent mistakes and, especially, by the absence of clefs in the whole cycle, for which no space has been left;^[41] their position can be reconstructed with some difficulty through the awkward placement of the key signatures, which sometimes varies from stave to stave even within a single voice.^[42] Even the positioning of the key signatures with respect to the calligraphic initials (also written by Scribe A) is sometimes wrongly reversed and not consistent within single folios.^[43] The space left at the beginning of each motet to insert the initial in the Cantus voice sometimes remained empty; however, even in these cases, the first word of each motet never lacks the first letter.^[44] Final barlines are also missing. All these inaccuracies suggest that this cycle was probably never performed directly from the Leopold Codex. It is not clear whether the numerous mistakes in both music and text are due to the copyist or were already present in the model that he used. However, it is all the more strange that Scribe A probably referred to the same manuscript for copying the above-mentioned missales cycle Gaude flore virginali, in which clefs and initials are always written properly. Surely missing clefs in Natus sapientia would not be a problem for singers if the key signatures were correctly positioned; thus it is possible that clefs were also missing in the model. The random placement of key signatures, instead, is probably due to Scribe A’s negligence, since they change from one stave to another, and therefore he would have needed to maintain exactly the same layout of the ancestor to reproduce them properly.

The present edition deviates in some respects from that of Noblitt.^[45] He provided the text only in the Cantus voice, as in the manuscript, but in two forms: first in a diplomatic transcription, respecting both the words’ spelling and their original position under the notes, and then in an editorial one, with normalized spelling and the text placed according to the editor’s choice. In this edition, instead, no diplomatic text is given and an editorial text has been added to all voices to facilitate performance. In fact, although the cycle was probably never sung directly from the Leopold Codex, it might have been performed in this way from other sources now lost. The addition of the missing text in the bottom voices was not always straightforward and a few ligatures were ignored in order to accommodate all the syllables. However, this solution seems preferable to the vocalization without any text of these three voices, if account is taken of the frequent homorhythmic passages and repeated notes,^[46] and, especially, of the long Altus–Tenor duo in motet 4, measures 7–19 (with twelve breve rests in the Cantus and a complete verse of the hymn missing if the other voices simply vocalize).

Editorial interventions in the music – slightly more numerous than in Noblittʼs edition^[47] – are limited to corrections of mistakes and listed in the Critical apparatus. Dissonances produced by passing notes, as awkward as they may sound, were not modified to avoid altering the single voices’ melodies and the overall effect, since they seem to be compositional choices rather than simple mistakes that need to be fixed. All clefs are editorial and wrong key signatures were moved to their right position.

Basel, November 2020
Updated July 2021