Earlier Use of Hufnagelschrift and Square Notation

A strong indication of these sources’ diverse origins is their notation. The Sint-Catharinadal manuscripts were originally notated in Hufnagelschrift or square notation. Most older manuscripts deploy Hufnagelschrift, identified with lands north and east of the Rhine, as shown in NL-OHnp 76 (Figure 3). This form of notation, defined by rhomboid noteheads written with inclined nibs, contains features also present in Messine notation.Footnote ⁴⁹ In her analysis of three fifteenth-century graduals from the diocese of Utrecht, Ike de Loos identified a similar style in antiphoners of the northern Netherlands. Such notation was, de Loos argued, typical of an East-Frankish style of notating chant that became exclusive of northern territories from the thirteenth century onwards.Footnote ⁵⁰ This form appears in no fewer than five of the twelve Sint-Catharinadal sources from the sixteenth and early seventeenth centuries, and all the sources pre-date the 1620s. These manuscripts may represent the earliest sources that arrived at the priory after their compilation in the sixteenth century, probably before the initial period of iconoclastic unrest of the 1560s, after which Catholic influence from the north was limited.

Figure 3 Hufnagelschrift: NL-Ohnp 76, fols. 46^v–47^r [37^v–38^r]

Most of the later manuscripts are written in a form of square notation, identified with the Southern Netherlands, as seen in NL-OHnp 93 (Figure 4). This square form, de Loos argued, is more typical of Brabant and lands south and west of the Rhine. This group also includes NL-OHnp 81 and 92 and the latter part of NL-OHnp 97a, all of which are estimated to have been made in the sixteenth and early seventeenth centuries. There is therefore a point of change within the original notation of the Sint-Catharinadal chant sources, from Messine/Hufnagelschrift to square, suggesting a shift of geographical influence from the Northern to the Southern Netherlands, with a putative date for the shift falling around the 1610s and ’20s.Footnote ⁵¹ This southern influence after the 1610s is all too understandable, given that access to land immediately to the north of Breda would have been almost impossible during periods of Spanish occupation.Footnote ⁵² A significant influx of sources around the 1620s coincides with van Dunne’s reform of the priory, suggesting that these sources came from the Spanish Netherlands as part of Sint-Catharinadal’s revival.

Figure 4 Older square notation: NL-OHnp 93, fols. 4^v–5^r [pp. 8–9]

Gathering Structures and Indications of Ownership

The present gathering structures of some of these books are not original. Some, such as NL-OHnp 76 and 76a, are modern compilations of different chant books, possibly from outside of the priory, which were later bound into new volumes and the foliation or pagination sequences adapted.Footnote ⁵³ Although the origins of these manuscripts are uncertain, areas of influence have been inferred from the use of specific forms of notation within subsections of a source’s gathering structure, as demonstrated in NL-OHNp 73, an antiphoner that contains several chants for the divine Office for various feast days. Like NL-OHnp 76 and 76a, this source comprises numerous fragments that were probably rebound sometime during the seventeenth century. That earlier parts of this source were made for Premonstratensian use is without doubt. The manuscript has music for the feast of St Norbert in two separate gatherings (see Table 3).Footnote ⁵⁴ These two sections have differing numbers of staves per page: nine in gathering 2 and eleven in gathering 3, as well as different scribes (see two versions of the name ‘Norbertus’ in Figure 5) and an entirely different style of initials and page decoration, all of which suggests that these were gatherings from diverse Premonstratensian communities that found their way into the later assemblage. Parts of this source came from outside the order, however. The final page in gathering 2 (fol. 16^v [p. 36]) includes a notice clarifying its date of compilation in 1684 (Figure 6). Accompanying this note is a Christogram accompanied by three nails piercing a heart, suggesting that Jesuits were responsible for the compilation of this ostensibly Norbertine gathering of chants.Footnote ⁵⁵ This manuscript gathering, despite postdating the 1680 revision, was itself revised, as is shown by the painted-over staves. This gathering may have entered NL-OHNp 73 in the early sixteenth century as a former Jesuit antiphoner. Jesuits were, after all, not denied the right to celebrate the feast of St Norbert.Footnote ⁵⁶

Table 3 Summary of chants for two versions of the office for St Norbert as given in NL-OHnp MS 73

Figure 5 Comparison of the name ‘Norbert’ written by two scribes in different gatherings within NL-OHnp 73: (a) fol. 4^r [p. 13], in gathering 2 (i); (b) fol. 19^v [fol. 3^v; p. 40], in gathering 3 (ii)

Figure 6 Date of 1684 in NL-OHnp 73, fol. 16^v [p. 36], in gathering 2

Alternatively, Jesuits may have made this gathering specifically for a Norbertine community.Footnote ⁵⁷ Given that Breda’s Jesuit house was located less than a kilometre from Sint-Catharinadal, the sharing of musical sources was feasible.Footnote ⁵⁸ Even by the mid 1680s, long after the conclusion of the Eighty Year’s War, it is possible that Sint-Catharinadal was sourcing its chant repertoire from outside and rebinding fragments into serviceable manuscripts, a practice known to have existed elsewhere, not just within Breda and its environs but throughout Europe.Footnote ⁵⁹ These fragments were ones that were well used, as shown by marginal notes and revisions entered on small scraps of paper bound into the gathering.Footnote ⁶⁰ It is therefore reasonable to assume that NL-OHNp 73 represents an assemblage of useful chants for a Norbertine community, some borrowed or acquired elsewhere. Given the cost of acquiring new manuscripts – or even new leaves of parchment or paper – it is understandable that religious houses would have economised by repurposing old sources into new ones.

Only one source, NL-OHnp 77, shows precise traces of its former ownership in an opening note:

Not only does this manuscript indicate a precise date (1676), but it also indicates an individual owner: canoness Maria Margaret Brouwers. From this indication, this source was not intended to be read around a lectern by an assembled group of canonesses, but was one woman’s personal book intended either for her participation in the liturgy, or for her private devotions. This canoness with the surname Brouwers was almost certainly a Premonstratensian resident at the priory itself, because this surname was that of an important local family that sent several female members to be canonesses at Sint-Catharinadal over the course of the sixteenth to eighteenth centuries. In the eighteenth century, there is mention of one Maria Anna Brouwers and her sister Elisabeth, both of whom were sent to Sint-Catharinadal to boost numbers after a period of low recruitment.Footnote ⁶² Given the appearance of multiple Brouwerses at Sint-Catharinadal prior to this period, it is likely that NL-OHnp 77’s owner was one such family member.

This book therefore shows that in addition to manuscripts that were read by multiple canonesses, there were also books owned and used by individual women.Footnote ⁶³ NL-OHnp 77 offers a brief conspectus of an individual’s expected knowledge of and participation in the Premonstratensian liturgy, as well as the impact individuals had on a book’s contents and presentation. The manuscript is an antiphoner with Temporale and Sanctorale chants that form part of the winter cycle, from Advent to Easter. NL-OHnp 77’s contents are relatively complete and cohesive, containing a systematic ordering of chants for the entire liturgical year, unlike NL-OHnp 76 and 76a. For a source presumably intended for personal use, probably in the choir of the church, it is also quite elaborate, with frequent use of coloured initials and recurring, coloured marginal drawings of animals, commonly birds. This book was therefore one that held personal value, both as an ordered compendium of chant repertoire and as a decorated object. NL-OHnp 77 was not merely the property of Maria Margaret Brouwers: her ownership would have been recognised by the community of sisters around her, who may well have witnessed her using it in the choir, and who indicated her ownership in the note on the front page, which was entered after her death.

Revised Square Notation

A common feature in the Sint-Catharinadal sources is revision, with most changes to the original chant enacted to conform to Nivers’s reform. Nivers was an influential organist and chant reformer, who had enjoyed close access to Louis XIV and the French court since at least the 1670s.Footnote ⁶⁴ France had a long interest in plainchant as a living tradition during the early modern era. During the time of Louis XIV, chant displaced the king’s favourite grand motet on important feast days at the royal chapel at Versailles.Footnote ⁶⁵ Revision was as important a priority in France as anywhere else, with long melismas and ‘defective’ accentuation unsuited to contemporary humanistic literary tastes. Revisionists called for melodic simplification, adapted to principles of tonal music, and set chant to modern rhythms (i.e., plain-chant mesuré).Footnote ⁶⁶ In this environment, a committee of senior canons commissioned Nivers around 1677 to revise the Premonstratensian antiphoner and gradual. The Premonstratensians had expressed the wish as early as 1660 for someone to ‘remove the useless protractions of chant, correct their accents and eliminate all forms of dissonances’.Footnote ⁶⁷ Nivers’s efforts were emblematic of the Counter-Reformation spirit, directed towards an extensive revision of sung liturgy and driven by the wish to elevate chant to what many perceived to be the glories of an earlier repertoire.Footnote ⁶⁸ His motivations were also political, echoing the priorities of an increasingly vocal Gallican church in Louis XIV’s France, which challenged Roman influence.Footnote ⁶⁹

Although newer printing technologies meant that buying revised official liturgical books had become cheaper, the relative price of paper meant that volumes such as Nivers’s 1680 antiphoner and gradual were still an extravagant purchase.Footnote ⁷⁰ Therefore, chant communities often updated their older manuscripts to reflect the latest, correct version of a chant instead. The older chant books of Sint-Catharinadal represent this priority to update even older sources. The priory’s books are written with melodies that were deemed by the mid seventeenth century to be too florid. They were therefore corrected to accord with the less melismatic melodies of Nivers’s Antiphonarium Praemonstratense.Footnote ⁷¹ It is uncertain whether Sint-Catharinadal ever had a copy of the new 1680 revision, as records of its presence in the region survive only in the abbeys of Park, Averbode, Tongerlo and Grimbergen.Footnote ⁷² The revisions to the Sint-Catharinadal sources may therefore have been implemented based upon a temporary period of access to the new edition. Sint-Catharinadal was still, after all, in a period of change, having moved to Oosterhout permanently just the year before the revision was published. There were doubtless greater priorities than purchasing the newest antiphoner, whereas relying on older sources that were updated may have been a more manageable solution. Other manuscripts from neighbouring houses (for example, Tongerlo and Grimbergen) were both retained and updated in the wake of Nivers’s 1680 edition, and many of these sources are centuries older than the earliest of the Sint-Catharinadal manuscripts.Footnote ⁷³

The Sint-Catharinadal sources incorporate multiple notational styles simultaneously within the same manuscript. The revised square form of notation used by later scribes, seen in the corrections to NL-OHnp 76a (Figure 7), differs from the earlier square script through a more distinctly seriffed form, the regular replacement of oblique pairs by square notes and the occasional presence of semibreves (all characteristics of seventeenth-century printed chant). That the revised square script appears consistently across Sint-Catharinadal’s musical sources suggests that sometime after the 1680 revision to the Premonstratensian rite an individual, or a small team of people following the same notational standard, were responsible for revising the house’s musical repertoire. That staves as well as notes were painted over is observed in NL-OHnp 76a. This source revises its Hufnagelschrift chants by painting over the original layer entirely, and on some pages scribes inserted newly ruled lines in red ink, which stand out from the usual black staves. The canoness-scribes then wrote new melodies and texts over the top in the revised square form.

Figure 7 Revised square notation: NL-OHnp 76a, fols. 17^v–18^r (corrections)

This approach appears to have been adopted throughout all sources at the priory, and partially in the hymnary NL-OHnp 80.Footnote ⁷⁴ Antiphoners intended to be read by multiple canonesses were revised, but so were those intended for individual women. As shown in Figure 8, NL-OHnp 77 appears to have been edited with the revised square notation as the new standard; however, unlike NL-OHnp 76 and 76a, the revised square notation is not entirely dissimilar to NL-OHnp 77’s earlier script, with the only discernable difference being the use of serifs on both upper and lower sides of each neume, rather than the exclusive use of descending serifs for the earlier layer of certain manuscripts.Footnote ⁷⁵ What this conformity of revision across the manuscripts shows is that a small number of scribes appears to have updated the manuscripts at Sint-Catharinadal, and that this revision was probably part of the house’s official policy, enacted to conform with Nivers’s new standard.

Figure 8 Revised square notation: NL-OHnp 77, fols. 66^v–67^r [pp. 126–7]

Most examples of erasure from houses to the south in the dioceses of Antwerp and Mechelen show revision at the same points of the manuscript. The Magnificat antiphon of First Vespers from the office of the Finding of the True Cross, O crux gloriosa (Cantus ID 004018), was revised in B-Gu BKT.006 of Tongerlo Abbey, and these revisions accord with Nivers’s 1680 edition.Footnote ⁷⁶ There are extensive points of erasure and occasional corrections written by a different scribe. Older chant retained in Nivers’s revision is kept. The same technique of erasing individual notes was used in B-Br 210 from Grimbergen Abbey, north of Brussels (fol. 116^r–v). While both sources from Tongerlo and Grimbergen write out the newly revised chant melody, NL-OHnp 76 from Sint-Catharinadal only writes out the notation for the first system (Figure 9). That NL-OHnp 76 only enters an incipit of the Niversian revision offers clues as to how Sint-Catharinadal’s canonesses read their chant manuscripts: here notation served as a prompt to the revised melody, with only the chant’s opening melody and the remainder of the chant text required to summon the rest of the chant from the singer’s memory.Footnote ⁷⁷ Partially notating portions of chant may also indicate that part of the melody was sung by a particular group of women, or even played by an organ; however, since the pattern of partially notating is inconsistent throughout the manuscript, and in the absence of more precise information on performance circumstances, this theory is unsupported by further evidence.

Figure 9 Incomplete revision of O crux gloriosa: NL-OHnp 76, fols. 106^v–107^r [89^v–90^r]

There are also inconsistencies that point to a less organised approach within Sint-Catharinadal itself. Occasionally there are revisions made on newly inserted leaves or smaller slips of paper, as seen in Figure 10 from NL-OHnp 76, where two sections of chant for Easter Sunday appear on separate slips: an antiphon for Matins, Alleluia, alleluia, alleluia, alleluia (Cantus ID 001329), and a detail in the first responsory for Sext, Alleluia. Angelus domini (Cantus ID 006093). On these slips are written the chant melody in square notation, accompanied by a new text script. The revised square notation given in an initial revision was either written down incorrectly and had to be corrected, or a later revision occurred that superseded the earlier revision. In any case, these inserted sheets, placed amongst revised text, point towards an inconsistency of the repertoire: either as something that was adapted in multiple stages and so required multiple stages of correction, or was not stable enough to have been understood in the first place and so risked introducing errors that then had to be corrected through the insertion of new slips of paper.

Figure 10 Revision on attached slips: NL-OHnp 76, fols. 10^v–11^r [4^v–5^r]

A point of interest in both NL-OHnp 76 and 76a is the approach towards editing some chants while leaving others in their original state. While these two sources contain chants for a wide variety of offices, the large majority only provide the notated Magnificat antiphons for First or Second Vespers and the Benedictus antiphon at Lauds. Figure 11 shows various chants for Sundays after Pentecost. Every Magnificat antiphon has been edited to accord with Nivers’s 1680 revision, with each following Benedictus antiphon appearing in its original, unedited Hugnagelschrift. There is a direct correspondence here to Nivers’s revision, since only the Magnificat antiphons that appear in the 1680 edition were updated, while those absent from the edition (i.e. the antiphons for the Benedictus) were left unedited. The canonesses would presumably have ignored these unaltered chants if they were no longer a part of the Premonstratensian service, so reminders of the earlier tradition remained alongside the updated liturgy. The unedited chants in the Hufnagelschrift may have still been recognisable to the canonesses, who could have recalled the earlier liturgy even if it was no longer permitted to be sung. That the O crux gloriosa chant in Figure 9 is only partially notated suggests memory played a crucial role within Sint-Catharinadal’s community, with entire melodies recalled from a melodic incipit or from the text alone.Footnote ⁷⁸

Figure 11 Revised Magnificat antiphons and unrevised Benedictus antiphons: NL-OHnp 76, fols. 79^v–80^r [67^v–68^r]

Leaving parts of the revised sources in the earlier Hufnagelschrift did not necessarily mean that older melodies could not have been recognised. Hufnagelschrift was not altogether inaccessible to the manuscripts’ readership, given Sint-Catharinadal’s location close to where two notational systems coexisted. The porous boundaries between Hufnagelschrift and square forms is apparent in contemporaneous sources such as B-Br 4826, a sixteenth-century Vesperale made for the church and attached college of the Jesuits in Leuven (now the Sint-Michielskerk).Footnote ⁷⁹ This repository of office chants, written on paper by what appears to be a variety of scribes, is a clean copy that did not go through subsequent editing. Here, square notation demarcates beginnings and ends of chants, while Hufnagelschrift is used almost consistently elsewhere. The square notation of B-Br 4826 was practical for incipits and cadential formulae, offering easy readability to its audience. But while Hufnagelschrift can identify discrete pitches, it was avoided at moments of intonation and final cadences. Unlike the chants in this Jesuit source, entire Hufnagelschrift chants in the Sint-Catharinadal antiphoners were often erased. While Hufnagelschrift was readable, the fact that few Sint-Catharinadal sources use it directly alongside revised square notation within the same chant suggests that the canonesses did not accept the close juxtaposition of notational styles. The Sint-Catharinadal sources indicate a preference for the clarity of square notation for chants that the canonesses sang more frequently, and this revised square notation mirrors the simplicity of notational style in Nivers’s 1680 antiphoner.

A New Source of Stroke Notation

Like many scribes tasked with updating their religious books elsewhere, those at Sint Catharinadal sometimes deviated from editorial conventions, and such a deviation appears in NL-OHnp 76. Many pages in the manuscript appear with stroke notation, in a script that cannot be dated accurately, but which nevertheless provides information about a melody’s rhythm through a series of vertical strokes. Table 4 summarises the complete chants with stroke notation in NL-OHnp 76, many of which are Magnificat antiphons for liturgical feasts throughout the year. Figures 12 and 13 present several examples of these strokes from the source.

Table 4 Summary of chants with stroke notation in NL-OHnp 76

Figure 12 Stroke notation: NL-OHnp 76, fols. 66^v–68^r [57^v–58^r]

Figure 13 Stroke notation: NL-OHnp 76, fols. 18^v–20^r [11^v–13^r]

Just as in normal square notation or Hufnagelschrift, the placing of the strokes indicates the pitches that should be sung. The number of strokes per note indicates how long each tone should last. A single stroke represents one tactus, two represents two and so on.Footnote ⁸⁰

Using stroke notation for Latin texts and in liturgical books used for singing is unusual. Stroke notation exists in mostly non-liturgical sources, particularly those for instrumental music, throughout the Netherlands and northern Germany, with a smaller number found in England and Italy.Footnote ⁸¹ They point to an alternative form of musical notation and means of musical literacy, which was used throughout the fourteenth and fifteenth centuries.Footnote ⁸² Barbara Haggh-Huglo has drawn attention to the Helmond manuscript, NL-HELga inv. no. 215, which is a register mostly containing contracts and acknowledgements of debts and transfers by the aldermen of Helmond.Footnote ⁸³ Bound into this volume is an oblong leaf (fol. 99^r), with two entries dated to 1416, containing three secular songs written in simple stroke notation indicating semibreves and minims. A further significant source is the Gruuthuse manuscript (NL-DHk 79.K.10).Footnote ⁸⁴ While this manuscript does contain some liturgical prayers set to music, they are few, and strokes are not used for anything other than secular repertoire. Also worth mentioning is B-Br 15589–15623, a songbook that dates from around the fourteenth to fifteenth centuries, filled with secular songs in the Dutch language.Footnote ⁸⁵ While most of this songbook has texts lacking notation, two folios include stroke notation. The first, fol. 23^v,Footnote ⁸⁶ once contained four pieces but has had a significant portion of the page cut out. Remaining are over 200 secular songs, with the collection headed ‘dits een rondeel’ (this is a rondeau). A further page in the manuscript, fol. 157^r, comprises six musical phrases in stroke notation for a simple secular song, with text divided into strophes of varying lengths. The rubric indicates that it is a song for St Peter’s eve (28 June).Footnote ⁸⁷ This song, ‘Wech op wech op dat herte mijn’, is not liturgical in nature but is part of the secular celebrations associated with Midsummer Eve. These sources demonstrate a consistent link between stroke notation and secular songs, particularly those of a popular or folkloric quality.

Stroke notation has been associated with a simpler form of representing music for those less versed in notational literacy, or for instrumental genres.Footnote ⁸⁸ The rhythmic system depicted by mensural notation, on the other hand, is less literal, since its method of depicting increased rhythmic values is non-figurative. The most common mensural notes, from shortest to longest, are the semiminima (filled stemmed rhombus), minima (void stemmed rhombus), semibrevis (void rhombus), brevis (void square) and longa (void square with stem). The sequence of visual cues in mensural notation – from filled to void, from stemmed to stemless, from rhombus to square – does not self-evidently indicate the increase in duration. Stroke notation, on the other hand, shows an accumulation of strokes that directly corresponds to a note’s greater time value. Stroke notation has the disadvantage of overwhelming the reader. When faced with the pressures of performance, it is challenging to keep track of a very long sequence of strokes. Further compounding the system’s limitations are the impracticalities of space for writing out longer notes in mensural notation, like the longa or maxima, which can despite being one notational grapheme represents many more durational beats. Stroke notation’s visually literal nature nevertheless makes it an accessible aid and, in the case of NL-OHnp 76, the durations required are both short and within a narrow range of the longest and shortest notes (one or two strokes), meaning that practicalities of both reading and writing the notation are not complicated.

Such strokes do not appear in any other Sint-Catharinadal source, so it is likely that a small group of women carried out these changes on NL-OHnp 76 at its initial stage of revision. Given that strokes are also sometimes associated with instrumental music, particularly that played by the organ or another keyboard instrument, an organist may have used this source for Magnificat antiphons and other chants that were often performed with organ. While it would be tempting to compare the strokes to Nivers’s revisions to see if both can be interpreted rhythmically, none of the cases in NL-OHnp 76 investigated so far include a comparable passage in the 1680 edition.Footnote ⁸⁹ Most instances in this manuscript appear to be the first revised entry after the older staves and notation were painted over. See, for instance, Figure 13 above, where the Magnificat antiphon for Vespers on Easter Wednesday, Dixit Jesus discipulis (Cantus ID 002295; fol. 19^r–v [12^r–v]), appears partly in the revised square form and partly in stroke notation. The sections that are in strokes on ‘quos predidistis nunc ascendit’ are the initial layer of revision, just like the surrounding square form. Here, stroke notation clarifies the rhythm of the revised chant. The scribe was comfortable with reading square notation, but at moments of uncertainty a literal depiction of tactus through accumulation of strokes was deemed helpful. Overleaf is one moment where the same scribe is identifiable. Observe the Magnificat antiphon for [Second] Vespers on Easter Tuesday, Videte manus meas (Cantus ID 005400; fol. 19^v [12^v]), written in strokes on a separate slip of paper pasted onto the original page, with a new text script in block letters. The only notes not in stroke notation here are the final two tones in void breves, at the end of the ‘Alleluia’: presumably, for the untexted final syllables, which are easier to sing without strokes on account of their falling on unison tones. The reason that this scribe decided to correct not just a chant but its text too may have been to account for new distribution of space with the stroke form.

Multiple musical scribes writing in stroke notation suggests that a small community of canonesses at Sint-Catharinadal thought of chant through this non-standard medium. See, for instance, the inserted part-leaf (fol. 67^r–v) in Figure 12 above, pasted within an earlier opening (fols. 66^v, 68^r [57^v–58^r]). On the recto side of the leaf is the invitatory antiphon for Matins at Pentecost, Alleluia spiritus domini (Cantus ID 001034). This invitatory on the recto appears in revised square notation, whereas the chant on the verso, the Magnificat antiphon for [Second] Vespers in the summer histories office De Regum, Quomodo ceciderunt fortes in bello (Cantus ID 006487za), is in stroke notation. Both revisions differ in their textual script from that of the inserted leaf on fol. 19^v [12^v] in Figure 13. The script here is cruder in its form, normally not cursive save for the textually ligated ‘g’ on ‘ego’. Meanwhile, on fol. 67^r in Figure 12, the script is also largely in block letters, but has distinctive features such as the open majuscule form for lower-case ‘e’. Both scripts have a closed miniscule lower-case ‘e’ form, but they differ in that fol. 67^v in Figure 12 shows greater tendency towards cursive writing. The melody, however, is in stroke notation. The stroke thickness and hue of ink in both cases are identical to their corresponding text scripts. Like Figure 13, the first scribe of Figure 12 uses strokes throughout, save for a punctum and virga on ‘ceciderunt’; however, this scribe also deploys diagonal strokes that appear elsewhere throughout the manuscript. Such strokes appear to be restricted to penultimate (and unaccented) syllables of longer words; they may therefore signal stress patterns in performance, which may correspond to rhythmic difference. For this second scribe, stroke notation offers a more literal representation of rhythmicised chant, yet strokes also bridge the gap between indicating stress patterns and specifying exact duration. Here, the first scribe of Figure 12 deploys strokes to indicate subtleties of performance practice and not just to demarcate the melody into units of temporal space.

Despite the interest in sources with stroke notation, NL-OHnp 76 has entirely escaped scrutiny. Yet this Sint-Catharinadal source is of profound significance, since it is the only Premonstratensian chant manuscript in current knowledge that contains stroke notation in the Netherlands.Footnote ⁹⁰ So why was stroke notation used at all, particularly for liturgical books for music that was ostensibly sung? Based upon the difference in penmanship between the two scribes, there were probably between three and five canonesses notating chant in this way, suggesting that more than one individual was comfortable using this alternative form and at times conceived of rhythm literally, as an accumulation of strokes, alongside the more common, abstract conventions of square notation.Footnote ⁹¹ That the canonesses should be using a form of notation commonly used for both secular and instrumental genres, in what has perhaps unfairly been considered a ‘lower’ form of notation, suggests that musical literacy at Sint-Catharinadal drew from a wide variety of sources, not just the square or mensural systems that proliferated in early-modern chant books. The presence of stroke notation also suggests that its associations with the secular and instrumental must be reconsidered, and that the boundaries between notational systems were porous. The precise circumstances under which canonesses updated the Sint-Catharinadal sources are unknown. There is no precise indication of when stroke notation appeared in NL-OHnp 76, nor for what reason. But given the practice of editing chant incipits, alternative chants and specific chants like Magnificat antiphons that were commonly played on an organ, there is some evidence to suggest that Sint-Catharinadal’s canonesses edited parts of NL-OHnp 76 to facilitate chant performance on an organ. Given the records of music making at the priory, including the ownership of keyboard instruments, it is not entirely inconceivable that the canonesses were accustomed to thinking about liturgical chant through a notational medium associated with non-liturgical and non-vocal music.

* * *

The chant manuscripts of Sint-Catharinadal’s library offer a case study of musical understanding within a female monastic community of the early modern era. The features of the antiphoners discussed suggest a localised response to liturgical reform during the sixteenth and seventeenth centuries, where almost continual upheaval at the priory inhibited effective transmission and understanding of the revised liturgy. As this bird’s-eye view of Sint-Catharinadal shows, the priory’s canonesses carried out these reforms inconsistently, and neither the dissemination of newly printed works nor the updating of older sources entirely satisfied the ambitions of revisers. The Sint-Catharinadal sources show that alongside newly printed antiphoners, scribes updated older sources and revised them to reflect liturgical changes in the chant texts and melodies. Often scribes tried to carry out these reforms systematically, as seen in NL-OHnp 77. Other sources point to a less comprehensive process of reform, such as NL-OHnp 76 and 76a, where only older chants in Hufnagelschrift that were in regular use appear erased and replaced. In these latter two sources, revised chants sit alongside those that the canonesses left unchanged, or which had fallen out of use. The existing sources of Sint-Catharinadal therefore point to a world where old and new chant repertoires existed in a continuum. Reminders of older chant texts and melodies were still very much available to manuscript readers, both to guide the reader’s oral memory of the correct, revised chants or to recreate earlier repertoire when required.

An unexpected example of non-liturgical and non-vocal influence – stroke notation – offers insights into monastic communities of the early modern era. Since this notation has few if any associations with sung liturgical music, these sources suggest that sacred vocal, secular vocal and instrumental repertoires operated within close spheres of influence. Sint-Catharinadal’s manuscripts show that rather than being the preserve of purely instrumental genres, stroke notation interacted with and potentially informed the performance of the sung liturgy. A number of the canonesses of Sint-Catharinadal were familiar with stroke notation and no doubt were aware of wider repertoires associated with it. That much is clear from the ownership of secular music books and instruments. It is not unreasonable to think that the canonesses, too, existed in a sound world that also included secular song alongside liturgical chant. Exposed to the sound of profane music, certain scribes deployed a variety of notational styles, whether that be the Hufnagelschrift of the north, square notation or the stroke notation common to popular and instrumental genres. For the canonesses of Sint-Catharinadal, the domain of non-liturgical music thereby formed some influence as they updated their chant and make it worthier of Counter-Reformation sonic ideals.

These sources therefore show the tension between reform and its implementation. Monastic houses across Europe implemented changes to chant based upon local practices, making Sint-Catharinadal’s chant holdings typical in their lack of consistency. What is remarkable, however, is that so many of Sint-Catharinadal’s earlier sources survive in their partially edited forms. These manuscripts document the struggles of the priory’s canonesses, as they grappled with the editorial requirements of Nivers’s 1680 revision. Expectations to update chant to Premontré’s standards ran counter to practical solutions. Compromises emerged, so that those singing the divine liturgy could recreate newly approved chants as easily as possible. This flexibility between ideals and practice resulted in the convergence of different notational systems. In sum, this conspectus of Sint-Catharinadal provides a better picture of the canonesses as they sought to update their liturgy. It is remarkable that having a properly updated chant library and a thriving musical tradition still held a certain priority, especially given successive conflicts in Breda and the priory’s gradual decline. Yet despite their attempts to update the liturgy, it is understandable that the few remaining sisters of the seventeenth century failed to adhere to the new plainchant ideals that emanated from further afield. Sint-Catharinadal’s sources point to individual canonesses resorting to non-standardised ways to depict the sound of revised chants, recalled from prior experience. Such recollection was informed by models from liturgical contexts, as well as from beyond the cloister.