Overview of contents and physical description

Ohio 220 is a collection of four hitherto unknown polyphonic songs; titles are provided in Table 1, editions in Examples 1–4, and the texts and translations in the Appendix. As Paweł Gancarczyk has described, compositions of this type belong to a central European tradition of polyphonic cantiones. Footnote ² These songs are characteristically bipartite in structure: the first section of a cantio is called the versus and the second is the repeticio. Footnote ³ Items 1 and 4 are dedicated to St Dorothy and St Margaret, respectively, while items 2 and 3 are both in praise of the Blessed Virgin Mary. Items 3 and 4 – those on the verso – are fully extant, with three and two voice parts, respectively,Footnote ⁴ but item 2 on the recto is missing an essential contratenor; this may also be the case for item 1.Footnote ⁵ Further musical and formal details to support these structural observations are discussed later.

Table 1. Contents of Ohio 220

Example 1. Ave gemma virtuosa / Tenor.

Example 2. Virgo decus castitatis / Tenor.

Example 3. Virginis ascensus divinitus / Contratenor / Tenor.

Example 4. O preclara Margaritha / Tenor.

Ohio 220’s isolated leaf is parchment, and measures approximately 194 mm at its widest point and 273 mm at its tallest (Figures 1 and 2).Footnote ⁶ The writing block is 155 mm × 239 mm. Both recto and verso contain eleven staves, ranging from 14 to 16 mm in height. Although the writing block was ruled in dry point, suggesting an organised approach, a rastrum was not used, and the staves do not consistently match up in both the left and the right margins, which range from 15 to 20 mm on the recto and 17 to 24 mm on the verso. The bottom margins measure between 15 and 17 mm, while those of the top are approximately 15 mm. The top of the fragment has presumably been trimmed – the initial at the top of the recto would not otherwise have been decorated to the very top edge – while the bottom edge of the fragment is likely unaltered. There is evidence of sewing stations (see especially the bottom left of the recto), of which there were probably five. Given the more extensive wear and glue stains on the recto (and the annotation in the top margin of this page to be discussed presently), it is likely that the polyphonic leaf served at some point as either the front of a parchment wrapper or the front flyleaf to another codex or libellus (with the recto facing frontwards), and – at some point – the recto was glued to the cover of the book’s binding.

Figure 1. Ohio 220, recto (reproduced with permission).

Figure 2. Ohio 220, verso (reproduced with permission).

The manuscript fragment is distinctive in its appearance, though not atypical of other central European manuscripts of the period. Ohio 220 features large black initials on both recto and verso, with decorative knots and filled spaces, including an informal human face in the lower initial of the verso; the initials on the recto are further edged in red ink.Footnote ⁷ The lowest voice in all four songs is labelled ‘tenor huius’, that is, ‘tenor of this [song]’ (and the second low voice of Virginus ascensus ‘contratenor huius’). In terms of these features of decoration and tenor labelling, a plausible comparand is CZ-Pn II C 7 (the Jistebnice Cantional), a source from the 1420s whose precise provenance is unknown.Footnote ⁸ CZ-Pn II C 7 includes black initials with red penwork decoration in various styles throughout the book, comparable to those found in Ohio 220 (see Figure 3, discussed further later). The same approach to labelling lower vocal parts (‘tenor huius’) is also present. Examination of central European sources suggests that these features are too common to indicate a single location for their origin, but do hint at the copyists’ awareness of contemporary scribal norms.

Figure 3. Quem elegit in CZ-Pn II C 7, fol. 56r (99a) (Creative Commons BY-NC-SA 4.0).

The time spent on decorative initials suggests that the scribe responsible for them was experienced and took care in their tasks. The mise-en-page, however, demonstrates poor planning: the scribe ran out of space at the bottom of the recto (Figure 1) and had to conclude the tenor voice of Virgo decus on the right-hand side of stave 6; the contratenor of Virginis ascensus on the verso (Figure 2) had to be finished on stave 8. There are a handful of basic copying errors as well, the most egregious of which is that the texted voice – the cantus – of O preclara Margaritha was notated a third too low (compare Figure 2 with Example 4). A level of roughness in the copying process is therefore evident in this leaf, which may well have originated from what was once a larger, semi-formal collection.Footnote ⁹ Despite the fact that only a single leaf survives, the full text of each cantus voice is extant, and seems reasonably accurate, though alignment of pitches and syllables is occasionally difficult to discern.

The characteristic bipartite structure of each song is signalled visually on the page in two ways, which is most easily observed on the less-worn verso (Figure 2). First, hastily drawn single or double bar lines divide all voice parts into two sections (the versus and the repeticio). In the cantus of the St Margaret song, for example, see the end of stave 9, and in its tenor, about two-thirds of the way across the bottom stave. Second, the tenor of each song is also labelled at the beginning of each second section: see ‘R^o tenoris’ (R^o = repeticio) on the bottom stave.Footnote ¹⁰

A short phrase at top of the recto, some of which is illegible, appears to bear no relation to the musical contents:Footnote ¹¹

This annotation references the incipit of Alexandre de Villedieu’s well-known Doctrinale puerorum (written at the turn of the thirteenth century into the fourteenth), a grammar text in verse that begins ‘Scribere clericulis paro doctrinale novellis’. The reason for this annotation is unclear, but it might have a connection to a now lost host volume, if this music fragment was reused as binding material (or as a wrapper) for a manuscript or printed book that began with the Doctrinale puerorum. Footnote ¹² An allusion to Villedieu’s well-distributed book may not help in establishing a specific provenance for the music, but it does add weight to the likelihood of the manuscript source having emanated from an institution such as a song school in a cathedral or monastery, where Latin would be taught to young learners.

The music and text were entered by two scribes, in at least two stages (see Table 1). On both the recto and verso there is some ambiguity to the order of working for scribes, and it may be that amendments were made in several stages. In general, it appears that the text was written before the music notation, and the initials appear to have been drawn before the notation was added, however, using the same dark black ink as the music notation. The compressed minim with a right flag at the beginning of the tenth stave on the verso indicates that the notation was entered after the initial, although the red ink decoration on the recto was added after the notation (the red ink is now visible on top of the black ink). The music hand (‘B’) on the verso is certainly different to the music hand (‘A’) on the recto: scribe B draws thicker, blacker stems, occasionally with a slight flick of the pen stroke at the top of the stem, and breves are drawn with slightly protruding vertical strokes on each side. The minim stems of music scribe A, on the other hand, are drawn with a much lighter brown ink, although it is possible that another hand (possibly scribe B, or the individual who added the words ‘Ave gemma’ and ‘munde’ in another text hand under stave five of the recto) went over some of these stems on the first two and half staves of the recto with the blacker ink. The text scribe of the recto (‘a’) uses this light brown ink, and apparently also added the clefs, which look to have the same light brown ink composition, and the filigree decoration on the second initial on the ‘A’ of ‘Ave’. The text, clefs and initial decoration on the verso are all in the same black ink as the music notation, and it is likely the case that music scribe B and text scribe b are the same individual. Similarly, given the light brown ink recipe used for the text, clefs and minim stems on the recto, it seems that music scribe A and text scribe a are another individual.

Notational features and musical style

To some extent, the notation of Ohio 220 reflects core practices of the notational system associated with the French Ars Nova, whose fullest theorisation occurred in the first half of the fourteenth century.Footnote ¹³ The mensuration of all four compositions is imperfect tempus, major prolation: note values range from imperfect longs to minims, with trochaic motion predominantly in semibreves and minims, although with stretches of breves in the lowest voice(s). Minim stems are mostly ascending, but at times are descending, perhaps reflecting an exemplar or simply scribal inconsistency.Footnote ¹⁴ Rests are infrequent, with three semibreve rests in the cantus in Ave gemma and two breve rests in the upper voice of Virgo decus. There are no minim rests. Ligatures are mostly found in the lower voice(s): in Ave gemma the tenor’s activity approaches that of the cantus, with cum opposita proprietate (or c.o.p.) ligatures (two semibreves) the only type of ligature in the tenor. In the other three compositions, a greater variety of ligatures is used in the tenor and contratenor voices, with up to fifteen notes joined in ligature. The most advanced notational feature in these compositions is the occasional instance of remote imperfection: a breve is imperfected by a minim in the cantus at bars 28 and 35 of Virginis ascensus (Example 3) and at bars 22 and 36 of O preclara (Example 4), and an imperfect long is imperfected by a minim in the cantus at bars 18–19 of O preclara. In each case, the remote imperfection is highlighted with a flag added to the right of each minim stem and followed by a stroke of division, and in the instance of the imperfected imperfect long, a descending stem is added to the left of the long to indicate that this note is a long. It is possible that this figuration is used here to indicate a particular vocal delivery that in Ars Antiqua and early Ars Nova sources could have been notated with a plicated long or breve, and in fact that is how this figuration is notated in the tenor voice of O preclara, where a plicated long is found, with descending stems on the left and the right of the square note head, and a longer stem on the left, to sound simultaneously with the remote imperfection in the cantus at bars 18–19.Footnote ¹⁵

This straightforward rhythmic language, and the relative lack of notational complexity, speaks to the copying or genesis of Ohio 220 during the very late fourteenth century or the first few decades of the fifteenth century. Its notational features can be readily compared with central European sources of the same decades, and this date range concurs with our observations on style and genre. Although French-style notational treatises were found in central Europe, notably in Bohemia, several central European musical sources of the same period use an adapted and simplified system, which is not represented in any extant contemporaneous theoretical work.Footnote ¹⁶ Martin Horyna has argued that the notational method found in such sources did not require its own theorisation, being in essence a pared-down version of earlier French models.Footnote ¹⁷ In particular, this notation employs a simplified system for ligatures: breves and longs, for example, are typically notated in the same way in ligatures as they are individually (see especially the descending ligatures in the lower voice(s) in Figures 1 and 2, as well as in Figure 3 discussed later). Styling this type of writing ‘Bohemian mensural notation’, Horyna proposes that its earliest examples date from no earlier than 1390.Footnote ¹⁸

Table 2 summarises the most salient musical and formal features of the four Ohio 220 songs. Each presents the characteristic binary structure of a cantio with uneven section lengths, and is relatively short, ranging from 39 to 65 breves. In actual performance, however – as demonstrated shortly – these songs may have been longer. The four songs favour C, D and to a lesser extent E as tonal centres; Virgo decus stands out from the others with different tonal foci for its two sections (E, then D). All four exhibit similar ranges for their respective voice parts: the cantus always falls within G–aa and the tenor/contratenor within C–d. Perhaps the most striking melodic feature is the use of lengthy melismas; see, for instance, the opening of Ave gemma (Example 1) or the beginning of the second section of O preclara (Example 4), where the melodic device is used to highlight the song’s subject, ‘Margaretha’. In general, the cantus of each song moves predominantly in stepwise motion, with occasional leaps up to a fifth. The tenors (and contratenor of Virginis ascensus) are for the most part much more disjunct; peculiar here are the ascending sevenths in Ave gemma (Example 1, bars 1–2 and 16–17).Footnote ¹⁹

Table 2. Structural and stylistic features of the songs in Ohio 220

Regular small-scale dissonances between the melody and supporting line(s) typify the harmonic language; see, for example, the ninths on the second semibreves of bars 11 and 12 of Ave gemma, and the seventh in bar 13. This leads to a curious moment in the song: there is hocket on a single syllable from bars 14 to 16, with two semibreve rests; this moment suggests that the text may be a contrafact, since there is no obvious reason why a hocket would be meaningful here. Moreover, the a/d fourth in bar 15 might indicate a missing lower voice, but this is the only moment in the song that really calls for it (though the second half of bar 31 would also benefit from a contratenor). In contrast, unsupported fourths abound in Virgo decus, as shown in the table, thereby indicating a missing contratenor; since these two songs appear together on the recto, perhaps indeed Ave gemma is also missing its contratenor, with both contratenor voices possibly occupying part of the opening’s missing facing verso (though such a mise-en-page is speculative). All four songs include examples of parallel fifths and octaves. Towards the end of Ave gemma’s first section, the name ‘Dorothea’ is highlighted with a series of homophonic fifths (the second half of bar 22 to bar 24); the same song includes one example of parallel unisons (bar 2).

There are textual concordances for three of the items, but no discernible musical relationship between the Latin-texted polyphonic songs in Ohio 220 and the melodies found with the same texts in other sources. Most significantly for tracing the longer heritage of the materials in Ohio 220, the text of Virgo decus is transmitted within a motet family witnessed in many of the well-known sources of Ars Antiqua polyphony and is cited as an example in two Ars Antiqua theoretical treatises (Table 3).Footnote ²⁰ Within the surviving sources there are four main musical presentations of this text: as the motetus in the three-voice Marian motet Res nova mirabilis / Virgo decus castitatis / ALLELUYA; as the motetus of a two-voice motet; as a three-voice conductus motet (where both triplum and motetus sing the same text Virgo decus castitatis); and the three-voice devotional song transmitted in Ohio 220. The first three of these are closely related musically, with the same music for the motetus and tenor voices. The earliest sources for this Ars Antiqua motet family are associated with mid-thirteenth-century France (one of the earliest French sources, F-BSM 119 (148) V, has a colophon directly next to, and in the same hand as the motet tenor, with a date of 1264), but examples are also found from the royal nunnery of Las Huelgas and, in the conductus-motet version, in a source associated with Ely Cathedral in East Anglia (GB-Ctc O.2.1).Footnote ²¹

Table 3. Concordances for the text Virgo decus castitatis

^a F-Pa 3517 transmits a two-voice version of the motet on fol. 118r (motetus and tenor), however, the triplum Res nova mirabilis is copied independently on fol. 2r.

^b Discantus positio vulgaris in Hieronymus de Moravia, Tractatus de musica, ed. S. M. Cserba, Freiburger Studien zur Musikwissenschaft 2 (Regensburg, 1935), 189–94, at 193; Heinrich Sowa, ed., Ein anonymer glossierter Mensuraltraktat 1279, Königsberger Studien zur Musikwissenschaft 9 (Kassel, 1930), 1–132, at 80.

The evergreen and simple poetic style of Virgo decus – essentially a list of epithets appropriate to the Virgin Mary as Queen of Heaven (virgo regia), with its lilting accentual verse with regular syllable counts and rhyme scheme – likely contributed to its wide dissemination.Footnote ²² The use of linguistic tropes and refrains was a fundamental part of composing both devotional and secular lyrics, which were very often deliberately intertextual; it is not unusual to find such phrases in texts with very different functions and social contexts.Footnote ²³ Given the compositional strategies of other items in Ohio 220, it is possible that the text Virgo decus had an independent transmission for use within a Marian liturgical service; indeed, one of the sources for Virgo decus in Table 3 (GB-Ob Lyell 72) is a north Italian processional. The motet is copied within the sequentiary portion of this chant manuscript in the midst of a handful of polyphonic sequences and motets. The text that is transmitted with Virgo decus in the triplum of the Ars Antiqua three-voice Latin double motet – Res nova mirabilis – is found independently as a sequence in a German manuscript dated to the final third of the fifteenth century, D-Mbs g716 (fol. 34r).Footnote ²⁴

What sets the copy of Virgo decus in Ohio 220 apart, however, is that it appears to be completely unrelated musically to the Ars Antiqua motet. The Ars Antiqua motet follows the accentual pattern and regular structure of the text, setting it syllabically in first mode to phrases of 7 longs that subdivide into subphrases of 4 longs+3 longs, accommodating the 8- and 5-syllable lines of verse. The tenor matches the first-mode rhythms of the motetus.Footnote ²⁵ Yet the setting in Ohio 220 completely disregards the patterns of syllable count and rhyme, with a more melismatic setting of irregular phrase lengths that includes lengthy virtuosic melismas, most dramatically at the beginning of the ‘B’ section on ‘Nos’ (Example 2, bars 37–43). In general, when the setting is syllabic, the rhythms mostly reflect the Latin accents, with accented syllables set to semibreves and unaccented syllables set to minims (see, for example, bars 3–4 or bars 17–18).

Comparisons with central European examples

Comparison with the object of the aforementioned study by Horyna – Que est ista – is instructive, since there is much in common between his example and the four pieces in Ohio 220 from notational and stylistic perspectives.Footnote ²⁶ Que est ista is preserved securely in his main source (CZ-Pn X A 6, front pastedown), from Česky Krumlov, and is also found more imperfectly in a second.Footnote ²⁷ Its three voice parts, as in Virginis ascensus, comprise a texted melody, underpinned by an untexted tenor and contratenor; the lower voices do not appear to derive from chant.Footnote ²⁸ As in all four pieces within Ohio 220, regular small-scale dissonances between the melody and supporting lines in Que est ista typify the harmonic language, as do occasional parallel octaves, and melismas in the cantus feature prominently, most notably on the first syllables of each of the song’s two sections. The syllabic text-setting is sometimes not entirely satisfactory however it is arranged, suggesting either a flexible attitude or perhaps that the well-known text of the Assumption antiphon is a contrafact. That these stylistic features can be detected across all four pieces in Ohio 220 strongly suggests – if not confirms – that the new source was part of a comparable tradition in a similar location.

Two additional comparands selected from the extant central European polyphonic cantiones can further contextualise our four songs. Flos florum, uniquely preserved as the lone musical item in the Czech source CZ-Pu V H 31, exemplifies those features just mentioned, but, additionally, the text written below its tenor voice provides a tantalising piece of information: ‘Tenor huius pulcerrimi rundelli ach du getruys blut von alden soln’, that is, the ‘beautiful rondellus, Ach du getruys blut von alden soln’ is concealed in ‘this tenor’.Footnote ²⁹ As mentioned, the contours of the lower voices in the four songs in Ohio 220 certainly do not suggest an origin in chant, and although we have yet to identify any musical relationship between their tenor melodies and any pre-existent songs, it seems plausible that they are likewise drawn from secular song.

A second comparand, Quem elegit, is preserved in at least four manuscripts – now all housed in Czech libraries and presumed to be of Czech origin – and sheds light on two aspects of performance practice. First, in CZ-Pn II C 7 (the Jistebnice Cantional from the 1420s discussed briefly earlier), the tenor moves along in rhythms similar to the four songs in Ohio 220 (Figure 3).Footnote ³⁰ In the other three manuscripts, however, the tenor might be described as ‘riffed on’, often with lilting trochaic semibreve–minim motion, and there are even occasional semiminims; see, for example, Figure 4, from CZ-HKm Hr-6 (Codex Franus, olim II A 6).Footnote ³¹ The repeated pitches in the final section might suggest that the second manuscript is an attempt at notating how Quem elegit would have been rendered in performance, perhaps even an artefact of a particular instrumental style, where the repetition of individual notes is driven by performance on an instrument less capable of sustaining a long note easily, such as a keyboard instrument or a psaltery. Regardless, the lively tenor line suggests the influence of vernacular music on cantiones, and Quem elegit must have been a popular song due to its preservation in four separate extant manuscripts, three of which have decorated rhythms in the tenor voice.Footnote ³²

Figure 4. Quem elegit in CZ-HKm Hr-6, fol. 344r (Creative Commons BY-NC-SA 4.0).

A second aspect of performance practice relates to form. In CZ-HKm Hr-6 (Figure 4), two additional verses for the ‘A’ music are added as residuum texts after the music, signalled by two Vs in red, indicating that the repetition of the A section is sung with different texts (perhaps AaB AaB). However, in CZ-Pn II C 7 (Figure 3), where there is underlay of the second versus directly under the first versus in the cantus (only one additional verse is provided in this source), the written-out repetition of the pitches of the ‘A’ music in the tenor at the end of this voice part (final stave in the figure) indicates that another statement of the ‘A’ music would be sung after the ‘B’ section (R^o) is sung. Thus, given the tenor’s first ‘A’ being enclosed in red ‘bar lines’, the song’s form as performed would be AaBA, similar to the form of some virelais.Footnote ³³ In light of the remarkable stylistic and formal similarities of the four Ohio 220 songs with these comparands, it is entirely possible that the four Ohio 220 songs would be sung with a repeat of their first sections even though in the source no additional text is provided for the return to the versus. Footnote ³⁴

Virginis ascensus stands apart from the other songs in Ohio 220, not only because it has an extant contratenor, but also because its second section is essentially the same music stated twice, but with different text the second time, and with a four-bar cauda to conclude. As shown in Example 3, bars 51–59 are an exact repetition of 41–49. In the cantus, the repetition is signalled in the manuscript (Figure 2) with a faint single vertical line and with a larger E initial for Et. And while the repetition of the contratenor in the song’s second section is written out in full – see the right-hand portions of staves 7 and 8 – it is not in the tenor: at the end of stave 6, the label indicating the second section, R^o, is followed by the text iterum incipias (start again) and then at the beginning of stave 7 is the word finalis (final; that is, what we would call the second ending).Footnote ³⁵ The musical structure of the song as a whole, as notated, is ABb. If the versus were then sung again in performance, then its form would be ABbA, resembling that of a typical virelai. This structure supports the idea that the song may have originated as a secular virelai and was then repurposed by the substitution of a Latin, devotional text.

Devotional context of texts and music in Ohio 220

Institutions most likely to own a group of devotional pieces include churches, monastic foundations or secular cathedrals, and we have already identified a possible connection with an institution of this type that maintained its own song school. That the four pieces honour the Blessed Virgin Mary and two other women – St Dorothy and St Margaret – arguably suggests either that the section of the original book contained a group of pieces relevant to the sanctorale (though there is no ostensible liturgical ordering), or that these female saints held some special meaning for the source’s copyists or users. It is possible to speculate further: the clustering of only female devotional figures would be appropriate for a book used in a nunnery, perhaps one dedicated to one of these women. Certainly, the tone of each lyric takes the women’s virginity as its central theme, and indicates the role of Mary, Dorothy and Margaret in intercession (see the texts and translations in the Appendix). Although some books from female houses incorporated vernacular items in response to uneven Latin literacy, the use of Latin text remained commonplace.Footnote ³⁶ We can also look to the compilation of female saints’ lives as exempla more widely: it was typical for high status women and nuns to receive books focused on virgin martyrs (such as St Catherine of Alexandria), and women were involved in the production, dissemination and reception of female-centred hagiography.Footnote ³⁷

Marian song was so ubiquitous as to make establishing a provenance based on dedication alone impossible. Yet praise of other holy figures can sometimes be more informative. The first item on Ohio 220, Ave gemma, represents one of very few extant polyphonic songs in honour of Dorothy.Footnote ³⁸ St Dorothy was martyred on 6 February 287 or 304 in Caesarea, Cappadocia, though the details of her life remain obscure.Footnote ³⁹ Her cult was disseminated widely, especially from the eleventh century. The oldest surviving version of her legend is that by St Aldhelm (639–709), whose De laudibus virginitatis (‘On the praise of virginity’) was written for the Abbess of Barking Abbey in England.Footnote ⁴⁰ The earliest vernacular historiae are from Germany: the Dorotheen Passie originated in Ostfalen c.1400, although its earliest copy dates to 1476, and a fragmentary life Von sent Dorothea may have been created in early fourteenth-century Bavaria.Footnote ⁴¹ St Dorothy’s Office is preserved in various manuscripts across Europe, all of which indicate provenance from central Europe.Footnote ⁴² The prominence of St Dorothy’s cult in Silesia, which lies across parts of modern Poland, Germany and Czechia, would offer a useful geographical context for the song. The church of Saints Dorothy, Wenceslaus and Stanislaus (Wrocław/Warsaw) once held an important relic of the saint’s skull and would therefore have been a focus of her cult.Footnote ⁴³

A Prussian anchoress is also a potentially attractive, if secondary, proposition for the use of the song: St Dorothy of Montau (hereafter Dorothea von Montau, feast day 25 June), was born on St Dorothy’s feast day, 6 February 1347 and died in 1394.Footnote ⁴⁴ The proliferation of hagiographical materials in honour of the more widely known St Dorothy in fourteenth- and fifteenth-century central Europe sometimes makes distinguishing between references to the two women challenging, though their vitae could not have been more different. Where there are ample sources for the Office of St Dorothy, no office has survived for Dorothea von Montau.Footnote ⁴⁵ Dorothea was married at a young age and bore nine children, only one of whom survived childhood; she experienced visions during her marriage, travelled with her husband to holy sites such as Cologne, and was eventually widowed during her husband’s pilgrimage to Rome c.1390. In 1391, Dorothea moved to Marienwerder (Kwidzyn in modern-day Poland) and was installed as anchoress in the cathedral in 1393. With the assistance of her confessor, Johannes of Marienwerder, Dorothea produced many Latin and vernacular books based on her life. Her personal piety was focused on Christ’s Passion, and she received the stigmata. She was quickly venerated after her death as a patron saint of Prussia, and within the decade a large collection of evidence had been compiled for her canonisation, though this did not take place fully until 1976. Archbishop Jan of Jenštejn (d. 1400), well known for his role in promoting the adoption of the Visitation of the Blessed Virgin Mary, drove the initial campaign for Dorothea’s canonisation.Footnote ⁴⁶ Though Ave gemma was most likely created in honour of St Dorothy, about whom the epithets of youth and beauty were more appropriate, it could have been repurposed for Dorothea in places where her cult flourished.

The text of Ave gemma, which lacks direct mention of St Dorothy’s martyrdom, is constructed from two extracts from the Office of St Dorothy, found across several sources (Table 4).Footnote ⁴⁷ Newly composed local office materials regularly drew on more established figures of veneration, deliberately using imitatio to strengthen the authenticity of an emergent cult, so it is possible that St Dorothy’s Office drew on that of other saints. The composite nature of the lyric in Ohio 220 argues against its use as a direct replacement for liturgical plainchant, but there would be nothing to prevent the song being used in broader devotional celebrations.

Table 4. Relationship of Office materials for St Dorothy to the polyphonic song Ave gemma virtuosa

Identifying the Margaret to whom O preclara Margaritha might be dedicated is also challenging, on account of multiple candidates and the shortness of the lyric. Musically, the song appears to be complete in its two extant voice parts. The text is similar to one found in CZ-Pn XIII A 2, fol. 202r–v, a chant book with fifty-six added items of polyphony, whose copying in Kolin – by Martinus Baccalarius Vyskytna – has been dated to 1512 (Table 5).Footnote ⁴⁸ This strongly suggests that the text of the two-voice piece was adapted from a pre-existent office text for St Margaret. There is a structural mismatch between the lyrics associated with the repeticio in each version, where the material that forms the repeticio in Ohio 220 is not signalled as such in the office source (CZ-Pn XIII A 2). This sort of treatment is not necessarily significant, given the adaptive practices that were fundamental to the creation of devotional poetry at the time, and the necessity of adaptation when working with contrafaction, as seems likely in all the cantiones in Ohio 220.

Table 5. Relationship of office text for St Margaret to the polyphonic song O preclara Margaritha

There are two possible candidates for the Margaret specifically referenced in the polyphonic song, and both enjoyed enthusiastic veneration in central Europe. The orthography of the name in Ohio 220, given as both Margaritha and Margaretha, accords with the flexible spellings of her name common across central Europe. The office materials in CZ-Pn XIII A 2 are preserved as a (monophonic) cantio in honour of St Margaret of Antioch (13 July), the most ubiquitous holy figure of that name; the identity of Margaret in this chant source is not in doubt, since the cantio is immediately followed by the feast of the Dispersion of the Apostles (15 July). That said, neither the chant source nor the abridged text of Ohio 220 mentions the best-known aspects of Margaret of Antioch’s vita: her martyrdom and the dragon whom she dramatically defeated, and thus the lyric could be easily appropriated for eponymous saints. The poem praises Margaret’s virginity and virtue, shining through ‘the interweaving of a glistening lily’, her ‘garland of holiness’; the Antioch chant text names her as a ‘bride of Christ’, but this is not transferred to the polyphonic song. The lily was commonly representative of virginity and purity, and found across diverse texts and visual representations; poet John Lydgate (d. c.1451), for example, named St Margaret of Antioch ‘The chaste lely of whos maydenhede / Thorugh martyrdam was spreynt [covered] with roses rede.’Footnote ⁴⁹ However, it may be significant that the Dominican nun Margaret of Hungary (d. 1270, feast day 18 January) is traditionally depicted holding a lily and a book, and the text would therefore be appropriate to her veneration; nuns were also considered brides of Christ. Margaret of Hungary was not officially made a saint until 1943, but she was considered a holy figure from her own lifetime, remained one of the most celebrated figures in central Europe, and there were several unsuccessful attempts to canonise her in the intervening centuries.Footnote ⁵⁰ If in honour of St Margaret of Hungary, which is certainly plausible if difficult to explore with a lack of comparable office materials, the polyphonic song O preclara Margaritha would be the sole example of a piece of music so dedicated. As with Ave gemma virtuosa, the more widely known saint is the more likely dedicatee, but it is important to consider alternatives in the light of the complex and intertwined devotional contexts of these texts.